WEBVTT
00:00:00.000 --> 00:00:02.350
So I don't know how we haven't already used this prompt mic.
00:00:03.129 --> 00:00:07.690
Perhaps we should be feeding all of the podcasts into chat GPT to, to tell us.
00:00:08.369 --> 00:00:13.640
But, um, I'm going to ask you today, the prompt that you get on chat GPT when you go onto the site.
00:00:14.060 --> 00:00:15.369
And that is what can I help with?
00:00:17.850 --> 00:00:23.350
Can you find a thousand or so people who would like to purchase my social media card game?
00:00:25.269 --> 00:00:25.600
What?
00:00:25.629 --> 00:00:25.890
Yeah.
00:00:25.890 --> 00:00:26.960
1 billion users.
00:00:26.960 --> 00:00:27.839
Let's get the plugin.
00:00:28.000 --> 00:00:37.012
Yeah, 1 billion users, which I will note this week OpenAI announced that they are aiming to get 1 billion users on chat GPT.
00:00:37.023 --> 00:00:38.383
So this all ties together
00:00:38.512 --> 00:00:39.203
Oh my God.
00:00:39.203 --> 00:00:42.332
It's, do you think they, they took inspiration from the Kickstarter?
00:00:42.573 --> 00:00:46.740
They must have, but I would like them to send 1 billion users to the Kickstarter.
00:00:46.770 --> 00:00:52.520
We are not quite at our threshold of support to actually make the game go into production.
00:00:52.820 --> 00:01:00.939
But if you are listening to this, I know you will like this game because it is all about social media and it is fun for the whole family to play.
00:01:01.420 --> 00:01:06.840
So if you're listening to this and you have not backed our 1 million users, 1 billion users.
00:01:07.355 --> 00:01:08.325
Game on Kickstarter.
00:01:08.325 --> 00:01:10.974
You can just go to Kickstarter and search 1 billion users.
00:01:11.194 --> 00:01:12.495
Please, please check it out.
00:01:12.504 --> 00:01:13.715
We need more backers.
00:01:13.754 --> 00:01:25.075
And if you know other people who are interested in online speech and trust and safety, and you want to have a fun game to play that will also help you explain trust and safety and social media and all this kind of stuff.
00:01:25.234 --> 00:01:26.194
Please, please do it.
00:01:26.224 --> 00:01:27.974
I hate to beg, but I need to beg.
00:01:29.109 --> 00:01:31.230
I'm, I'm going to buy five more.
00:01:32.265 --> 00:01:32.844
There we go.
00:01:32.864 --> 00:01:33.715
We have a discount.
00:01:33.724 --> 00:01:35.594
If you buy five, we have a discount,
00:01:35.844 --> 00:01:36.584
I'm going to do it.
00:01:36.775 --> 00:01:39.724
but anyways, Ben, what can I help you with today?
00:01:39.965 --> 00:01:40.584
What can you help with?
00:01:40.594 --> 00:01:43.355
I am incredibly jet lagged, Mike.
00:01:43.385 --> 00:01:47.795
I've just landed back from your fair nation after a week of work there.
00:01:47.864 --> 00:01:50.265
And I do not know what day or time it is.
00:01:50.334 --> 00:01:56.871
Um, I have not figured out how, To, yeah, to actually speak as you can tell.
00:01:57.141 --> 00:02:03.341
Um, so any, any advice, any help on how to get back to a normal rhythm of sleeping would be much appreciated.
00:02:03.941 --> 00:02:04.902
Let's see how this goes.
00:02:05.272 --> 00:02:07.561
Yes, yes, this, this may be an interesting one.
00:02:07.621 --> 00:02:09.181
Uh, I can confirm that.
00:02:09.472 --> 00:02:14.861
Well, one, I was up late last night working on stuff and I was texting Ben who was in Chicago.
00:02:15.337 --> 00:02:17.098
this was not that many hours ago
00:02:17.617 --> 00:02:17.967
No,
00:02:18.448 --> 00:02:19.693
and now he is.
00:02:19.853 --> 00:02:28.902
In London, and he showed me his, bag that he had just thrown in the corner after getting back home in time to record this.
00:02:28.953 --> 00:02:33.932
So we are both working on quite little sleep and, uh, Ben even less than I.
00:02:33.932 --> 00:02:39.893
So this may be a very fun or very strange episode, depending on where things go.
00:02:39.893 --> 00:02:40.828
Bye.
00:02:40.927 --> 00:02:41.388
everyone.
00:02:49.512 --> 00:02:54.913
Hello and welcome to control all speech, your weekly roundup of the major stories about online speech.
00:02:55.173 --> 00:02:57.402
Content moderation and internet regulation.
00:02:57.883 --> 00:02:59.832
It's December the 6th, 2024.
00:03:00.152 --> 00:03:04.462
And this week's episode is brought to you with financial support from the future of online trust and safety fund.
00:03:04.812 --> 00:03:14.310
And by today's sponsor internet society, a global nonprofit that advocates for an open globally connected, secure, and trustworthy internet for everyone.
00:03:14.956 --> 00:03:15.597
I'm Ben Whitelaw.
00:03:15.597 --> 00:03:17.597
I'm the founder and editor of everything in moderation.
00:03:18.012 --> 00:03:27.701
at least I, I am, or was the last time I checked, um, and I'm with an equally kind of bleary eyed Mike Masnick, for altogether different reasons.
00:03:28.551 --> 00:03:29.361
It's good to be here, Mike.
00:03:29.382 --> 00:03:32.231
We haven't been in this, in these respective chairs for a few weeks.
00:03:32.627 --> 00:03:33.497
Yeah, that's right.
00:03:33.497 --> 00:03:34.986
We took off last week.
00:03:35.016 --> 00:03:41.189
We missed all the big stories, uh, during Thanksgiving week and then the week before that you were traveling.
00:03:41.189 --> 00:03:42.310
So you were not here.
00:03:42.340 --> 00:03:47.009
And so, uh, it's been, been a little while since, since we've, been able to do this together.
00:03:48.210 --> 00:03:52.847
Um, we, we've got a lot to cover today, partly because yeah, we were, we took a break last week.
00:03:53.073 --> 00:03:55.560
happy Thanksgiving again for, to you and to our listeners.
00:03:55.860 --> 00:04:07.032
And we have a really good, bonus chat at the end of today's episode with some super smart folks from the internet society, around social media bands in Canada and Australia and age verification in general.
00:04:07.062 --> 00:04:08.995
So, stick around for that at the end.
00:04:08.995 --> 00:04:09.159
Thanks.
00:04:09.300 --> 00:04:14.823
Today's show, and we won't cover those kinds of stories today, just because we do that really well with, with those folks.
00:04:15.099 --> 00:04:19.103
we're going to dive straight in to, um, what is essentially Mike a breaking story.
00:04:19.432 --> 00:04:31.466
Um, it's, it's a live story that you essentially woke up, rubbed your eyes and, uh, have dived into, it's the, uh, the tick tock case it's back on the agenda.
00:04:32.067 --> 00:04:32.646
Yeah.
00:04:32.757 --> 00:04:33.247
So.
00:04:33.637 --> 00:04:36.857
We knew that there was supposed to be a ruling coming down basically any day.
00:04:36.857 --> 00:04:51.637
This is the, US law that in theory forces ByteDance to divest of TikTok in the US or forces, the app stores and ISPs to block access to it.
00:04:52.107 --> 00:04:55.060
And they are supposed to either divest.
00:04:55.250 --> 00:04:58.867
or we'll have access blocked on January the 19th.
00:04:58.908 --> 00:05:16.964
And there's a whole sort of political side story to this, which is that our incoming returning president, uh, Donald Trump, who originally was the one who proposed the TikTok ban and tried to do so in a very ham fisted way that failed, four years ago.
00:05:17.459 --> 00:05:29.120
Has now completely flipped positions, perhaps because he was backed heavily by one of, ByteDance's biggest investors, but has insisted that he's going to protect TikTok.
00:05:29.279 --> 00:05:35.845
But he comes into office, I think a day or two after this law goes into effect.
00:05:35.894 --> 00:05:41.172
And so TikTok had challenged the law and it went to the DC circuit, which is in the appeals court.
00:05:41.701 --> 00:05:48.041
And they came down with their ruling today, basically saying that the law is.
00:05:48.442 --> 00:05:49.072
Just fine.
00:05:49.122 --> 00:05:53.062
There is no problems with the law and TikTok must obey it.
00:05:53.062 --> 00:05:54.641
And ByteDance must obey it.
00:05:54.968 --> 00:06:00.007
this ruling did come out literally, about an hour before we were recording.
00:06:00.158 --> 00:06:04.997
I have read what I think are the most salient points in it.
00:06:05.269 --> 00:06:11.927
however, given the timing, I may make some mistakes in terms of this because I have not had a chance to go through it in great detail.
00:06:11.927 --> 00:06:13.266
Have you even had a coffee Mike?
00:06:13.357 --> 00:06:13.447
Are
00:06:13.516 --> 00:06:25.040
I have not actually, I am, I am, uh, yeah, anyways, so, I think they're wrong is, is, is my quick summary.
00:06:25.663 --> 00:06:31.112
it's interesting because the government tried to argue that there were no first amendment issues at all.
00:06:31.132 --> 00:06:34.752
And that the case could be decided without even considering the first amendment.
00:06:35.533 --> 00:06:39.709
The court went in a different direction and said that government was wrong, but that the first amendment.
00:06:40.149 --> 00:06:41.269
Does apply here.
00:06:41.730 --> 00:06:52.259
And then once you say that the First Amendment does apply, then there's a question of which level of scrutiny, and there are different levels, you know have to be passed in order for the law to be considered constitutional.
00:06:52.550 --> 00:06:55.214
The highest bar is is strict scrutiny.
00:06:55.574 --> 00:06:56.995
And then there are, there are lower bars.
00:06:56.995 --> 00:07:01.894
And the government had argued that if the First Amendment does apply, you should go for one of the lower bars.
00:07:02.675 --> 00:07:06.824
ByteDance had argued that it should be strict scrutiny, which is the hardest to pass.
00:07:07.555 --> 00:07:14.848
The court notes that there isn't a comparable situation that says which of the levels of scrutiny should apply.
00:07:15.197 --> 00:07:19.168
And so they actually choose not to, say which level should apply.
00:07:19.168 --> 00:07:33.084
Though there is a concurring, uh, opinion by one of the judges suggesting, which level should apply, but the court says it doesn't matter because we think that this law passes even the highest bar strict scrutiny.
00:07:33.495 --> 00:07:36.035
And so then they just analyze it based on strict scrutiny.
00:07:36.584 --> 00:07:37.665
And I think they do it.
00:07:37.894 --> 00:07:41.675
in my one read through it, I think they do a terrible job of it.
00:07:42.115 --> 00:07:47.170
They basically buy into strongly buy into the claims by the U.
00:07:47.170 --> 00:07:47.329
S.
00:07:47.329 --> 00:07:57.100
government of the national security concerns, even though they admit that no actual evidence is presented for those, but they basically say, well, the U.
00:07:57.100 --> 00:07:57.240
S.
00:07:57.240 --> 00:07:58.129
government, a lot of people in the U.
00:07:58.129 --> 00:07:58.259
S.
00:07:58.259 --> 00:08:03.389
government say that we should be concerned and therefore we're just going to take that as fact.
00:08:03.550 --> 00:08:06.040
This is not uncommon in the U.
00:08:06.040 --> 00:08:06.300
S.
00:08:06.779 --> 00:08:12.122
court system when the government says, well, we have national security and blah, blah, blah.
00:08:12.122 --> 00:08:13.302
We can't tell you what they are.
00:08:13.302 --> 00:08:14.572
We're just really concerned.
00:08:14.920 --> 00:08:17.689
the courts are often willing to go along with that.
00:08:17.870 --> 00:08:25.427
That has led to a whole bunch of really terrible things having to do with civil liberties and civil rights and all sorts of stuff going back decades.
00:08:25.826 --> 00:08:43.325
But that seems to happen again here where they're just like national security concerns seem totally legit and you know ByteDance hasn't given us anything to Reject that they admit multiple times that a lot of these concerns are totally speculative And yet because ByteDance can't respond to the questions Speculative concerns.
00:08:43.764 --> 00:08:45.304
Therefore, this is all okay.
00:08:45.884 --> 00:08:57.111
There's a whole bunch of other like little stuff in here that again, it's just kind of like, well, you know, by chance I said, we have all these other, less restrictive means of doing this, which is part of the strict scrutiny test.
00:08:57.432 --> 00:09:00.845
And the court is like, nah, you know, we don't think any of those good enough.
00:09:00.845 --> 00:09:07.065
I mean, you know, Bytance really pushed this idea that, they would separate out all the operations, which they've mostly done.
00:09:07.294 --> 00:09:09.384
Everything is hosted in the U S by Oracle.
00:09:09.384 --> 00:09:11.504
They've given Oracle the power to audit stuff.
00:09:11.788 --> 00:09:25.977
they even offered the U S government, like an off switch, like that they could, you know, push a button and turn off TikTok, you know, in the case that something really bad happened and the government was like, nah, you know, that doesn't really satisfy what the government is doing here.
00:09:26.268 --> 00:09:28.048
There are a bunch of other oddities in here.
00:09:28.077 --> 00:09:36.231
You know, there were issues around like specifically targeting, like the bill actually names Tik TOK, which seems like it's singling them out for different treatment.
00:09:36.261 --> 00:09:43.298
And the court kind of waves that off and says, well, no, that's not really true, even though it does mention it.
00:09:43.409 --> 00:09:45.631
it, they said at one point, it's not punishment.
00:09:45.631 --> 00:09:54.292
If it was a bill of attainder, which is a specific thing where it's like just targeting someone for punishment, they said, yes, it only names TikTok, but not for punishment.
00:09:54.701 --> 00:09:56.631
And therefore it's not a bill of attainder.
00:09:56.861 --> 00:09:59.562
There are all of these like really odd things.
00:09:59.922 --> 00:10:02.682
So the big question now is kind of like what happens next?
00:10:02.682 --> 00:10:04.701
Like does TikTok.
00:10:04.917 --> 00:10:08.996
get shut off on January 19th and be gone for two days.
00:10:08.996 --> 00:10:14.352
And then Trump comes into office and suddenly changes position and allows to talk to come back.
00:10:14.523 --> 00:10:15.702
I don't know.
00:10:15.702 --> 00:10:28.513
I mean, I, I think, and I could be, there may be, there's some procedural weirdness in terms of how this law was written in terms of like forcing it to go straight to the circuit court rather than a district court, which would be the normal thing.
00:10:28.942 --> 00:10:35.518
I assume that they can now go to the Supreme Court and ask for a stay on this ruling, in order to appeal it.
00:10:35.518 --> 00:10:47.868
And maybe that puts the decision off past, January 19th and allows Trump to get into office and then sort of, you know, do something to keep TikTok, unless he decides that he, he maybe doesn't want TikTok.
00:10:47.937 --> 00:10:49.758
We, we don't really know at this point.
00:10:50.488 --> 00:10:53.538
Who knows what side of the bed Donald Trump will, will wake up on.
00:10:53.817 --> 00:10:56.927
And so it's really interesting to me, Mike, thanks for unpacking that.
00:10:56.927 --> 00:11:03.607
It's really interesting to me that all of the work done by TikTok to essentially kind of separate it out as a, as a.
00:11:04.187 --> 00:11:12.341
distinct legal entity, under this kind of project, Texas, umbrella that it, that it came up with a few years back actually has seemingly not worked at all.
00:11:12.575 --> 00:11:19.361
it has not proven to be persuasive to, in this case at all as to the national security, threat still.
00:11:19.918 --> 00:11:25.552
Why do you think that is, why do you think that the, the kind of, All of that efforts has been, has been kind of fallen on deaf ears,
00:11:25.971 --> 00:11:51.774
I mean, there is some element of it, which is just, honestly, it feels like general fear of China and that comes through in the ruling to like you see repeatedly, just talk of the PRC, the PRC, this, the PRC, that, and even if they say they're not going to, if the PRC comes down and demands that they do this or that, they will have to obey and therefore, that is the overriding concern.
00:11:52.113 --> 00:12:05.482
And, you know, there were stories that came out in the press from like former TikTok employees saying that, employees in China still had access to data and that that's noted in the decision.
00:12:05.482 --> 00:12:06.503
They, they point that out.
00:12:06.503 --> 00:12:10.222
And so they say, you know, we just, basically they just don't trust that this is real.
00:12:10.232 --> 00:12:10.893
They, it's sort of.
00:12:11.448 --> 00:12:25.865
the judges seem to feel that this was kind of a fictional setup, um, and they don't really trust that Oracle's ability to audit it as actually meaningful because again, they, they feel that, the PRC could come down and, do all of this.
00:12:26.275 --> 00:12:31.697
You know, the other argument that is made too, is that, It's really just focused on the PRC.
00:12:31.748 --> 00:12:45.494
I mean, because they say, you know, one of the reasons why they claim that this isn't a first amendment violation is because if the company does divest, they say all of the same content and all of the same moderation could still occur.
00:12:45.683 --> 00:12:48.557
There would be no change to the, expressive nature of things.
00:12:48.557 --> 00:12:52.280
The only thing we are doing is trying to disconnect the PRC from that.
00:12:52.620 --> 00:12:53.821
From tick tock.
00:12:54.191 --> 00:13:03.803
And so they feel that, the project Texas set up and the Oracle audits and the U S government like off button don't separate the company from the PRC.
00:13:03.854 --> 00:13:05.683
And so that's where the court comes down.
00:13:05.994 --> 00:13:12.677
They'd only be happy if, the PRC sold by dance or by dance, changed owners or changed hands.
00:13:12.677 --> 00:13:16.732
That's, that's the only way this would actually kind of, uh, change the, change the outcome in, in many sense.
00:13:17.197 --> 00:13:17.498
Yeah.
00:13:17.498 --> 00:13:21.398
And, and, you know, I think for a variety of reasons, I think, I think it's wrong.
00:13:21.437 --> 00:13:25.878
I think it's, there are other precedents on this that they sort of poo poo.
00:13:25.878 --> 00:13:28.388
And we're just like, the hat doesn't, doesn't really apply here.
00:13:28.725 --> 00:13:30.163
but it is something that happens.
00:13:30.163 --> 00:13:31.653
I think, I think it's a, bad look.
00:13:31.653 --> 00:13:37.196
I think it's a bad look for, you know, a supposedly free country in the U S to sort of take this viewpoint.
00:13:37.265 --> 00:13:44.778
And I think it will justify all sorts of other bad stuff from elsewhere as well, based on things like this, uh, it's, I think it's a bad look.
00:13:45.019 --> 00:13:46.739
I know some people really.
00:13:46.864 --> 00:14:09.956
Are concerned and there may be legitimate reasons to be concerned about the Chinese government and their connections here But I just feel that this law goes against, Basic american principles on on things and and i'm disappointed by by the ruling which you know again I've only read it once and really focused on the first amendment section, which is sort of the second half of the ruling And, um, I just, uh, it strikes me as very unconvincing.
00:14:10.291 --> 00:14:10.650
yeah.
00:14:10.780 --> 00:14:17.390
And that national security concern has yet to be proven as real or anything beyond really hypothetical at this stage.
00:14:17.725 --> 00:14:18.046
Yep.
00:14:18.086 --> 00:14:18.395
Yep.
00:14:18.426 --> 00:14:20.212
And, you know, it, it, that frustrates me.
00:14:20.212 --> 00:14:35.302
I mean, I have this come up in all sorts of, other cases around like Fourth Amendment stuff and encryption and just, you know, the willingness of the courts to accept the government just sort of giving this blanket statement that this is a national security concern without proof is, is just kind of frustrating.
00:14:35.648 --> 00:14:35.988
Yeah.
00:14:36.048 --> 00:14:36.368
Okay.
00:14:36.778 --> 00:14:48.758
it links, um, nicely to, to our next story, Mike, which you have, Figured out, um, it's, it's a TikTok story, but it's about a country that we don't often talk about in controlled speech, um, Romania.
00:14:49.038 --> 00:14:57.140
and I guess this is kind of really what the, in some senses, the threat that the, courts, and the government in the U S are trying to kind of mitigate against.
00:14:57.140 --> 00:14:57.390
Right.
00:14:57.666 --> 00:14:58.306
Yeah.
00:14:58.505 --> 00:14:58.855
Yeah.
00:14:58.855 --> 00:15:01.936
It's, it's kind of an interesting, interestingly related.
00:15:02.005 --> 00:15:24.128
Um, there's an election going on this coming Sunday, in Romania, but in the first round of the election, that's sort of a, you know, a two round election, the first round to figure out who are the top two candidates and then the, final election is you know, with just those two top two candidates, there was a surprise, you know, second place finish that knocked out the incumbent.
00:15:24.138 --> 00:15:28.927
The incumbent, I believe, came in third and so is not taking part in the election this weekend.
00:15:29.427 --> 00:15:38.918
there was a candidate, and I may pronounce his name wrong, but Kalin Georgescu, who was considered a sort of mostly unknown far right.
00:15:39.138 --> 00:15:44.508
populist candidate who sort of came out of nowhere with a big TikTok following.
00:15:44.927 --> 00:15:50.217
Um, and he has been sort of very pro Russia, pro Putin.
00:15:50.227 --> 00:15:56.368
And in fact, a lot of his successful TikToks are sort of reminiscent of, Putin.
00:15:57.097 --> 00:16:06.490
Style, uh, press appearances, riding a horse, doing judo, apparently running on a track without breaking a sweat.
00:16:06.791 --> 00:16:09.140
It's just sort of this like physical prowess.
00:16:09.591 --> 00:16:11.980
nothing, nothing to, you know, who doesn't like that?
00:16:11.980 --> 00:16:14.441
That sounds like a great, a great TikTok experience.
00:16:14.701 --> 00:16:15.140
yes.
00:16:15.181 --> 00:16:15.551
Yeah.
00:16:15.831 --> 00:16:18.754
And so he built up a really big following on, Tik TOK.
00:16:18.754 --> 00:16:23.313
The polls in the country had suggested that he was, not going to get that many votes.
00:16:23.677 --> 00:16:26.717
and a lot of people, it appears sort of chalked up his.
00:16:27.008 --> 00:16:30.878
Big Tik TOK following to foreign influence campaigns.
00:16:30.878 --> 00:16:34.898
They also noted that those videos were like highly stylized and produced.
00:16:35.197 --> 00:16:40.638
And there were questions of who was doing some of the producing and the stylizing behind it.
00:16:41.168 --> 00:16:47.004
There were questions about, whether or not the followers were real, were they foreign efforts?
00:16:47.004 --> 00:16:47.864
Were they bots?
00:16:48.227 --> 00:16:50.148
there were questions around.
00:16:50.523 --> 00:16:55.133
Influencers who were promoting the videos and were they paid or not?
00:16:55.712 --> 00:17:13.624
I, I, I found this really interesting point around, like, whether this candidate was, marked in the same way as other political candidates, basically suggesting that, they weren't kind of tagged as a political candidate running an election and therefore gotten larger reach than the other candidates, which is a really interesting kind of element to this, isn't there?
00:17:13.953 --> 00:17:14.463
Yeah.
00:17:14.663 --> 00:17:17.173
And that's unclear because tick tock denies that.
00:17:17.449 --> 00:17:30.144
but deal is because there are restrictions on political advertising within Romania, If all the candidates are treated as candidates, then in theory, there were limitations on kind of what sorts of promotions they could do.
00:17:30.144 --> 00:17:34.444
And if they were paying influencers to promote, is that a political ad?
00:17:34.500 --> 00:17:35.781
and how does that apply?
00:17:36.061 --> 00:17:51.817
And so there are all these questions and the suggestion from many, including some government officials is that TikTok didn't designate him as a, politician, as a candidate, and therefore didn't have these restrictions and that allowed his videos to go much further and build up a bigger following.
00:17:52.376 --> 00:17:53.686
Again, TikTok denies it.
00:17:53.686 --> 00:17:56.356
So it's a little unclear whether or not that is true.
00:17:56.527 --> 00:17:59.535
There is though, that, the things that TikTok.
00:17:59.815 --> 00:18:08.271
Did do it was at one point they did remove what they refer to as a covert network that was trying to boost his videos.
00:18:08.461 --> 00:18:10.434
And so, they did.
00:18:10.711 --> 00:18:20.241
Something, you know, they certainly, paid attention to something and found some inauthentic behavior behind him and that they claim that they stopped.
00:18:20.241 --> 00:18:25.932
And they also said that they found some similar types of inauthentic behavior for some of the other candidates.
00:18:26.311 --> 00:18:28.662
Um, so it wasn't that they did nothing.
00:18:29.008 --> 00:18:37.739
they were willing to step up and do something, but Romanian officials, feel pretty sketched out by this and they've asked the EU to investigate.
00:18:38.204 --> 00:18:53.174
and you know, it, it brings up all the questions that happen when, whenever there are unexpected results in elections and everyone's looking for explanations and, and these days, oftentimes they quickly jump to, well, it was this internet platform that caused the problem.
00:18:53.174 --> 00:18:53.394
Yeah.
00:18:53.525 --> 00:18:53.924
Yeah.
00:18:53.974 --> 00:19:02.825
I mean, there's a couple of interesting points here is in some ways this story is, one we've seen many times before, which is platform being used by political candidate.
00:19:03.138 --> 00:19:06.259
with the kind of looming threat of foreign interference in the background.
00:19:06.548 --> 00:19:11.848
And, you know, we've there's countless examples of that, Analytica being the kind of most, most famous one, I guess.
00:19:12.249 --> 00:19:16.420
And, It's nice to see that things don't change in the best part of a decade.
00:19:17.400 --> 00:19:22.926
But I was really interested in, you know, if that is the case, if there is a kind of lack of oversight here, Romania is pretty small.
00:19:23.017 --> 00:19:25.134
It has 20 million citizens.
00:19:25.564 --> 00:19:29.223
8 million of those apparently use TikTok or have a kind of TikTok account.
00:19:29.223 --> 00:19:31.186
So it's a significant number.
00:19:31.406 --> 00:19:39.194
You know, why was it that there wasn't really, I guess, more attention paid to Making sure that candidates were, given the same kind of platform.
00:19:39.865 --> 00:19:48.734
And I decided to kind of go into the DSA data, my find out how many, how many moderators Romania has, um, or how many are Romanian speaking.
00:19:49.144 --> 00:19:58.845
And, and so there are, according to the DSA report from January to June, 95 Romanian speaking moderators for a country of 20 million.
00:19:59.459 --> 00:20:01.308
And we don't know if that's good or bad.
00:20:01.308 --> 00:20:04.788
It doesn't necessarily take into account, I guess, people setting policy either.
00:20:04.788 --> 00:20:08.845
That's probably just the folks who are, looking at reports and appeals and those kinds of things.
00:20:09.224 --> 00:20:11.335
But it doesn't seem a lot compared to other countries.
00:20:11.481 --> 00:20:15.128
so the Netherlands, which has roughly the same number of.
00:20:15.294 --> 00:20:22.693
citizens, we don't know exactly how many users, of TikTok, but they have 160 moderators, speaking Dutch.
00:20:22.693 --> 00:20:23.713
roughly double,
00:20:23.903 --> 00:20:30.940
So roughly double, um, Sweden has, roughly the same number of moderators, 99, but has half the population.
00:20:31.329 --> 00:20:35.210
And again, we don't, we don't know how many, Of those people use, use Tik TOK.
00:20:35.619 --> 00:20:41.289
So again, you know, it's interesting to me that kind of how resources are applied.
00:20:41.660 --> 00:20:55.650
And again, this is a tale as old as time, how platforms, in the kind of non core markets, set policy and operationalized policy is something that we, we see time and time again as, as a sticking point, do you think that's kind of relevant here?
00:20:55.650 --> 00:20:58.913
Does that feel like a, a pathway trod a little bit?
00:20:59.263 --> 00:20:59.544
Yeah.
00:20:59.544 --> 00:21:24.611
I mean, you know, it is one of these, big questions that comes up all the time, so many of these discussions, you know, and we try to, you know, one of the reasons you and I are always looking for stories outside of, the U S and, The big countries that everybody talks about is because there is an important story there about how these companies handle it and the, you know, less followed countries where, where less attention is, paid to them.
00:21:25.080 --> 00:21:28.000
And so I think, I think it is a, it's a huge story.
00:21:28.330 --> 00:21:32.540
And, you know, there is this element of this one where I sort of feel like, I think.
00:21:32.885 --> 00:21:41.685
A lot of people really believe that so much of his following, in this case, were bots, that they didn't think that the votes would follow, and yet the votes did.
00:21:41.695 --> 00:21:49.732
And so, there was this part of me when I first story where I was like, and, you know, the first thing I was reading was like, oh, you know, it's all fake followers and bots.
00:21:49.732 --> 00:21:52.383
I was like, yeah, but it wasn't fake voters, you know.
00:21:52.772 --> 00:21:55.022
So, so something is happening here.
00:21:55.490 --> 00:22:03.914
but to be honest too, like, I don't know that, you can say necessarily that like, Oh, it's 95 or whatever moderators too little.
00:22:04.397 --> 00:22:05.198
I don't know.
00:22:05.258 --> 00:22:15.758
And we, we don't know, like, was that not enough where the policy's not in place again, there's the whole thing where like TikTok has denied this and they did take down some inauthentic behavior.
00:22:16.065 --> 00:22:19.766
so it, we, we have, you know, Partial information, not full information.
00:22:19.955 --> 00:22:30.400
If the EU does an investigation, it would be really interesting to find out if more details come out of it, but it is, it is sort of an interesting story pay attention to and to see what comes of this.
00:22:30.795 --> 00:22:31.464
yeah, definitely.
00:22:31.464 --> 00:22:36.365
And I mean, I think the other part of this story, something that kind of is a thread that runs through U.
00:22:36.365 --> 00:22:36.555
S.
00:22:36.565 --> 00:22:40.285
TikTok ban as well, is just the, is a media story, really.
00:22:40.285 --> 00:22:48.911
It's about the fact that there is so many people kind of consuming TikTok and the shift away from traditional media to, new forms of media.
00:22:48.911 --> 00:22:52.121
The fact that politicians can bypass traditional forms of media.
00:22:52.122 --> 00:22:52.296
Yeah.
00:22:52.506 --> 00:22:56.930
as this candidate has done and still perform very well in an early round.
00:22:56.930 --> 00:22:58.509
So we'll keep a tabs.
00:22:58.690 --> 00:23:01.980
I've never said this before, Mike, but I'll be keeping tabs on the Romanian election this Sunday.
00:23:02.059 --> 00:23:02.559
Uh,
00:23:03.744 --> 00:23:04.315
There you go.
00:23:04.545 --> 00:23:06.714
This, this, this coming Sunday is my birthday.
00:23:06.714 --> 00:23:10.964
So I will celebrate it by paying attention to the Romanian election as well.
00:23:11.349 --> 00:23:12.470
happy birthday in advance.
00:23:12.470 --> 00:23:13.029
And, uh,
00:23:13.184 --> 00:23:13.325
you.
00:23:13.559 --> 00:23:14.759
What, what a way to celebrate
00:23:15.924 --> 00:23:16.625
Yes, yes.
00:23:16.989 --> 00:23:25.890
so if those kind of two stories paired together, I guess, represent in some ways the kind of ghost of, antitrust past Mike, you know, the way that the,
00:23:26.144 --> 00:23:26.835
a segue.
00:23:26.865 --> 00:23:27.954
What a segue.
00:23:28.349 --> 00:23:31.813
you know, I think the next story is the ghost of, of antitrust future.
00:23:32.240 --> 00:23:38.101
next one you picked out and it's about a man that I actually didn't know about, uh, up until, an hour or so ago.
00:23:38.631 --> 00:23:41.201
Um, but he's an incredibly frightening looking man.
00:23:41.201 --> 00:23:44.250
I'll have to say, and I'm not sure if I'm ever going to forget his face.
00:23:45.161 --> 00:23:49.270
Um, so tell us about Andrew Ferguson and, and what he's come out with this week.
00:23:49.441 --> 00:23:50.111
Oh my goodness.
00:23:50.121 --> 00:23:52.590
So Andrew Ferguson is an FTC commissioner.
00:23:52.810 --> 00:24:04.184
and as a Republican, the FTC has five commissioners, three are appointed by whichever party has the presidency and two are the, other party.
00:24:04.204 --> 00:24:06.880
So right now there are two Republican commissioners.
00:24:06.932 --> 00:24:11.133
FTC commissioners and three Democratic ones, that will flip in January.
00:24:11.782 --> 00:24:15.715
And, Andrew Ferguson is one of the Republican commissioners.
00:24:16.006 --> 00:24:25.079
He is vying for very clearly vying for the chair, taking over what Lena Khan's position is, is right now under a Trump administration.
00:24:25.675 --> 00:24:37.403
and there was a New York Post article Today or yesterday that sort of detailed, there are three leading candidates, the two current FTC commissioners, which is Ferguson and the other one's Melissa Holyoke.
00:24:37.759 --> 00:24:44.220
and then there's, there's a third person, who, worked for Senator Mike Lee that he's really pushing for to be chair.
00:24:44.539 --> 00:24:53.200
And the, the question that people around Trump are apparently asking is like, which one of these is going to be toughest on, they say, big tech.
00:24:53.441 --> 00:24:54.950
but they really mean Trump's enemies.
00:24:55.180 --> 00:25:03.029
Especially considering how much support he got from certain tech sectors this time around, it'll be, you know, the companies that they don't like.
00:25:03.569 --> 00:25:10.619
And so the FTC took a fairly typical FTC action this week against an e commerce platform called Goat.
00:25:11.224 --> 00:25:16.795
The details aren't even that interesting in the case, but basically they lied about shipping times.
00:25:16.835 --> 00:25:19.884
People were paying for premium shipping and not getting it in time.
00:25:20.424 --> 00:25:25.508
And then also they had a, buyer protection, thing where they're saying like, if anything goes wrong, we'll protect you.
00:25:25.508 --> 00:25:27.327
And then they weren't, they weren't living up to that.
00:25:28.028 --> 00:25:32.795
The FTC took action on them and basically saying, you know, they were making promises and they weren't living up to it.
00:25:32.795 --> 00:25:34.644
That's an unfair and deceptive practice.
00:25:35.015 --> 00:25:37.194
Very typical standard FTC.
00:25:37.829 --> 00:25:39.680
Nothing at all interesting about that.
00:25:40.202 --> 00:25:56.915
The other Republican commissioner, Holyoke, put out a concurring statement on this that basically just said, this also proves that, we can use the same authorities to go after big tech companies for unfair moderation decisions.
00:25:57.411 --> 00:25:57.750
Okay.
00:25:58.226 --> 00:25:59.266
Which is nonsense.
00:26:00.191 --> 00:26:00.550
Yeah.
00:26:00.962 --> 00:26:12.455
but it was like a one paragraph thing and it looks like to me, at least Andrew Ferguson then said, okay, I see that I got a one up it because we're in, we're in a fight here for the chair position.
00:26:12.611 --> 00:26:13.411
I raise you.
00:26:13.682 --> 00:26:16.199
I raised you, I raised you crazy.
00:26:16.262 --> 00:26:19.208
and put out this like four page concurring thing.
00:26:19.278 --> 00:26:26.214
Again, none of this has anything to do with the ruling on goat, which is the company that this is ostensibly about, saying like, yes.
00:26:26.234 --> 00:26:28.164
And I agree with Holyoke that we can.
00:26:28.412 --> 00:26:44.762
Use this power to go after unfair moderation, but also we, we can use our powers to go after advertisers who stopped advertising on X because that must be antitrust collusion to censor free speech.
00:26:44.883 --> 00:26:46.772
And we have to support free speech.
00:26:46.903 --> 00:26:54.103
And right now there's only one free speech platform Elon Musk, the brilliant, wonderful free speech supporter.
00:26:54.482 --> 00:27:03.347
And, you know, How dare, any other platform censor American speech and that must be illegal and how dare they not advertise.
00:27:03.718 --> 00:27:13.046
so he goes after, advertisers who stopped, he goes after Garm, which we've talked about in the past saying, you know, that was, clear evidence of, of collusion.
00:27:13.276 --> 00:27:25.395
He goes after NewsGuard, saying that, NewsGuard, who I've written about a few times that the Republicans have gone crazy about, NewsGuard, all they do is just say which news sources are trustworthy and which are not.
00:27:25.776 --> 00:27:30.913
And he sort of admits when he's talking about NewsGuard, like, yes, NewsGuard can have its own opinions.
00:27:31.282 --> 00:27:37.903
But if multiple companies are basing decisions on those opinions, that is antitrust collusion.
00:27:38.303 --> 00:27:51.039
It, it is four pages of crazy indicating once again, like with Brendan Carr, we talked about a few weeks ago, like with Brendan Carr that, these.
00:27:51.242 --> 00:28:01.057
Bureaucrats really intend to use the powers of government to attack speech online, and they're framing it all within the language of free speech.
00:28:01.387 --> 00:28:05.351
The whole thing over and over again, he talks about, how important free speech is.
00:28:05.371 --> 00:28:08.740
And he does the Elon Musk is the only believer in free speech.
00:28:08.951 --> 00:28:13.964
And every platform has to use the same policies as, Elon does.
00:28:14.174 --> 00:28:15.625
And how dare they not do that?
00:28:15.867 --> 00:28:32.414
so just to clarify, so we have a situation where we're antitrust as we've talked about in the last couple of weeks of the podcast which is going to be a big theme within the Trump administration, and, you know, he's run out of the FTC has a guy in Brendan Carr who doesn't know much.
00:28:33.420 --> 00:28:35.200
Brendan Carr's the FCC, not the
00:28:35.464 --> 00:28:36.394
Sorry, sorry.
00:28:36.394 --> 00:28:43.488
Yeah, so, so the FTC commissioners also don't know very much according to this letter, at least this particular guy.
00:28:43.698 --> 00:28:48.290
And so we have a situation where like maybe nobody knows anything about
00:28:49.346 --> 00:28:55.326
it's, I mean, the question is, do they know or are they just like putting on a show for Trump?
00:28:55.355 --> 00:28:55.655
Right.
00:28:55.685 --> 00:28:58.415
And, and it's just this sort of like populist thing.
00:28:58.806 --> 00:29:02.205
I don't know Andrew Ferguson that well, Brendan Carr, I know a little bit.
00:29:02.205 --> 00:29:04.115
And so I know he knows that he's.
00:29:04.570 --> 00:29:08.201
Lying, like Brendan Carr is smart enough to know what the law is.
00:29:08.607 --> 00:29:12.345
I don't know enough about Ferguson to know whether or not he knows this is crazy.
00:29:12.845 --> 00:29:32.381
You know, one of the lines that really got me in this letter was like, he claims at one point that the proof of collusion among big tech companies to censor content in an illegal manner is that simultaneously, he specifically says simultaneously, all of the big tech platforms blocked.
00:29:32.615 --> 00:29:39.006
All discussion and reporting on the Hunter Biden laptop in 2020.
00:29:39.340 --> 00:29:40.790
Which just isn't the case.
00:29:41.175 --> 00:29:42.996
None of that happened, right?
00:29:43.006 --> 00:29:48.500
The only thing that happened was two companies took some action, Twitter and Facebook.
00:29:49.046 --> 00:29:51.705
The action that Twitter took was it blocked the link.
00:29:51.786 --> 00:29:53.996
It did not block any other reporting on it.
00:29:54.006 --> 00:29:55.445
There was other reporting on it.
00:29:55.586 --> 00:29:57.526
It did not block any discussion of it.
00:29:57.526 --> 00:29:59.256
There was lots of other discussion on it.
00:29:59.256 --> 00:30:01.705
In fact, like it was like a trending topic.
00:30:02.185 --> 00:30:04.175
The only thing they did was they blocked the link.
00:30:04.316 --> 00:30:08.355
To a single New York post story for 24 hours.
00:30:08.395 --> 00:30:12.026
Then they reversed their policy and allowed that link to be shared.
00:30:12.615 --> 00:30:18.435
The only thing that Facebook did was it said, well, there's some questions about the story, so we're going to keep it out of the trending.
00:30:18.586 --> 00:30:20.685
It won't go into the trending topics.
00:30:21.115 --> 00:30:23.486
and they reversed that policy relatively quickly.
00:30:23.945 --> 00:30:28.195
That was the only thing it did, but he declares unequivocally that the entire.
00:30:28.421 --> 00:30:32.601
Tech industry simultaneously blocked all discussion of this.
00:30:33.174 --> 00:30:51.637
And I, you know, one of the things that gets me is I pointed out earlier this year that Elon Musk did all that and more when apparently Iranian hackers got access to the Trump campaigns dossier on JD Vance and they passed around and most, most media sources didn't bite on it.
00:30:51.647 --> 00:30:59.557
Finally, Ken Klippenstein, who has a substack posted it and Elon banned Ken.
00:30:59.788 --> 00:31:01.087
He blocked all.
00:31:01.512 --> 00:31:06.309
Links to, any part of Ken's sub stack, not just that one article.
00:31:06.700 --> 00:31:08.730
He, you know, pulled down all sorts of stuff.
00:31:08.730 --> 00:31:11.049
And to this day, I don't think you can share that.
00:31:11.059 --> 00:31:22.826
He did let Ken back on the platform after like two weeks, So everything that, they have accused Twitter of doing to the Hunter Biden laptop story, Elon has done and more and gone much further.
00:31:23.076 --> 00:31:24.195
And yet in this.
00:31:24.477 --> 00:31:39.397
comment from this FTC commissioner, he claims that Elon is the big free speech supporter and the actions taken on the, the Hunter laptop, which didn't happen, prove that they're illegal censoring, collusion, antitrust.
00:31:39.788 --> 00:31:41.798
Not everything about this.
00:31:41.887 --> 00:31:43.387
You've got your hands on your head, Mike.
00:31:43.387 --> 00:31:44.667
It's
00:31:45.107 --> 00:31:49.196
it's so wrong, but this is, this is unfortunately the world that we're living in.
00:31:49.196 --> 00:31:56.170
and it, gives a sense of how the incoming administration is going to attack content moderation.
00:31:56.259 --> 00:31:57.779
They're going to make these claims.
00:31:57.960 --> 00:32:04.869
They are going to try and use every legal lever they have, even as they are crazy and totally counterfactual to reality.
00:32:05.119 --> 00:32:05.369
Yeah.
00:32:05.680 --> 00:32:15.170
There's no I mean, the idea that FTC could prove that platforms coordinated on policy changes or anything like that would be so difficult to do, right?
00:32:15.170 --> 00:32:26.463
If this actually was, you know, is a route you want to go down, how do you go about saying that this company over here would Has done the same thing as this company over here a way that amounts to collusion.
00:32:26.747 --> 00:32:42.990
Yeah, well, the thing that they can do, and they probably will do, is that they can conduct investigations, and they can demand to see all sorts of internal files, and that is what is going to happen, almost certainly, and then, I would guess what would come out of it is, probably a, really misleading investigation.
00:32:43.335 --> 00:32:50.419
findings and they'll release things selectively that take things out of context and make, make claims that are just not accurate.
00:32:50.772 --> 00:32:52.449
and it's going to be a mess.
00:32:52.582 --> 00:33:04.634
and this is why we're seeing, tech companies, you know, trying to kiss the ring of Donald Trump and, and try and make nice because they know if they don't, they're going to face all of this kind of, authoritarian nonsense.
00:33:04.839 --> 00:33:06.099
yeah, no, indeed.
00:33:06.460 --> 00:33:13.047
And, and, you know, currying favor, it's actually, uh, again, a very good segue onto our next story, um, is, is fast becoming
00:33:13.051 --> 00:33:14.231
did it on purpose, Ben.
00:33:15.666 --> 00:33:16.794
The theme of this, episode.
00:33:16.844 --> 00:33:23.676
So, you know, we have a situation where, FTC commissioners are, are, cozying up to, the new administration.
00:33:23.777 --> 00:33:29.467
We also have a situation where Meta in a very coordinated way is doing the same thing.
00:33:29.676 --> 00:33:47.487
and so this week, a number of different outlets, including the financial times and the verge reported on, Nick Clegg, the, president of global affairs at Metta, talking about how, Metta essentially overstepped the line when it came to content moderation during COVID, the comments were made in a, reporter.
00:33:48.092 --> 00:33:58.902
Briefing, which is, uh, you know, does happen, but it's a very kind of coordinated, very kind of controlled environment, for a very senior person within Metta to kind of make these statements.
00:33:59.281 --> 00:34:10.454
And, yeah, the, coverage has been essentially that Metta is not apologizing, but admitting that it, it overstepped the line in terms of, COVID information controlled during that period.
00:34:10.849 --> 00:34:11.068
And.
00:34:11.449 --> 00:34:20.298
This comes on the back of, you'll remember Mike, that letter by Mark Zuckerberg, uh, to Jim Jordan, which we had a good laugh about, um,
00:34:21.173 --> 00:34:22.094
I thought it was a cry.
00:34:23.418 --> 00:34:47.704
we cried a bit, um, it, it, it felt like a letter that had been, written at gunpoint, um, I remember, I remember saying, and, uh, almost kind of made, made to write that, and it was clearly in lieu of this situation, right, that where Donald Trump becomes president again, and, you have a situation as was reported this week, where Zuckerberg is invited to Mar a Lago to talk about the future of tech policy, out of the U S.
00:34:47.713 --> 00:34:53.347
So in a number of different ways we have, you know, in here, we have meta kind of cozying up to the new administration.
00:34:53.347 --> 00:34:55.628
We have the FTC commissioners doing the same.
00:34:56.121 --> 00:34:58.780
Is there anybody that has any dignity left?
00:34:59.550 --> 00:35:02.391
You know, I think this stuff is so obvious, right?
00:35:02.391 --> 00:35:14.963
It's, you know, it's so obvious and in some respects, I'm frustrated about the way that it's reported, in this way, because apart from the line that says This is a briefing with Nick Clegg and some journalists who've been invited there.
00:35:15.978 --> 00:35:24.213
quite clearly, you know, designed to be a signal to the Trump administration of, we know what you're going to ask us to do and we're happy to do it.
00:35:24.603 --> 00:35:24.974
Yeah.
00:35:25.233 --> 00:35:25.543
Yeah.
00:35:25.594 --> 00:35:26.554
That's exactly what it was.
00:35:26.614 --> 00:35:28.403
this was totally a messaging thing.
00:35:28.733 --> 00:35:35.083
It was, you know, a coordinated attempt by the company to lay out this message that will be embraced by.
00:35:35.391 --> 00:35:40.418
the sort of MAGA faithful, to insist that it proves, I mean, this will be extended, right?
00:35:40.418 --> 00:35:54.521
They'll say it proves that, not only that MEDA was overly, willing to suppress speech, but it will just be reinforced with the claim that, the demands for that came from the government, which is the part that he didn't say that.
00:35:54.764 --> 00:35:57.181
But that is, claim that lots of people are making.
00:35:57.664 --> 00:36:08.840
and so, This is, it's a spineless capitulation to this argument and basically it's giving the Republicans ammo to claim that we were right all along.
00:36:09.061 --> 00:36:10.570
We were unfairly targeted.
00:36:10.721 --> 00:36:14.744
We were unfairly censored and, even meta admits it.
00:36:14.900 --> 00:36:16.541
and we'll see that over and over again.
00:36:16.561 --> 00:36:18.530
And people will point to this as if it's proof.
00:36:18.791 --> 00:36:21.021
I had somebody yell at me this week,
00:36:21.577 --> 00:36:22.277
only one.
00:36:22.557 --> 00:36:29.168
well, there were a few people, but someone was yelling at me about this, where I was talking about some of this and they were saying, well, you know.
00:36:29.518 --> 00:36:34.458
Zuckerberg admitted under oath that the US government pressured him to take down content.
00:36:34.487 --> 00:36:37.402
He didn't want to take which is not What actually happened,
00:36:37.507 --> 00:36:37.907
Right.
00:36:38.452 --> 00:36:43.882
but like the message gets out there and kind of knows what they're doing when they, when they say this.
00:36:44.282 --> 00:36:46.422
and you know, I understand why they're doing it.
00:36:46.443 --> 00:36:55.353
You know, they feel like they need to do it to avoid, to hopefully avoid costly stuff, but it is, shows a real lack of principles as far as I'm concerned.
00:36:55.867 --> 00:36:56.197
Yeah.
00:36:56.197 --> 00:37:14.523
And if you're a trust and safety professional working in meta who, probably spent countless hours trying to figure out what the policy should be during an evolving situation that no one has ever seen before that was COVID and no one will ever see, you know, for a long time, that's going to be really, really tough to take.
00:37:14.748 --> 00:37:16.057
it's demoralizing, right?
00:37:16.057 --> 00:37:23.938
I mean, you know, the reality is what these companies should be doing and they don't do is saying, like, these are really, really, really difficult decisions.
00:37:24.108 --> 00:37:25.728
And there was no way to get it right.
00:37:25.797 --> 00:37:27.038
There was simply no way.
00:37:27.038 --> 00:37:30.018
I mean, I talk about sort of impossibility theorem here.
00:37:30.208 --> 00:37:31.487
There is no way to get it right.
00:37:31.677 --> 00:37:32.273
And that is it.
00:37:32.483 --> 00:37:35.103
Extra true in a case where you have something that is brand new.
00:37:35.112 --> 00:37:38.393
Nobody understands, you know, nobody understood the details of COVID.
00:37:38.592 --> 00:37:39.913
We didn't know what was right.
00:37:39.913 --> 00:37:41.123
We didn't know what was wrong.
00:37:41.393 --> 00:37:44.452
And people made choices and lots of people made wrong choices.
00:37:44.672 --> 00:37:49.822
Some of them made wrong choices because they just didn't have enough information and there's more information came out, they adjusted.
00:37:50.043 --> 00:37:54.079
Some people made wrong choices because they, you had crazy ideas in their head.
00:37:54.298 --> 00:37:57.148
There were all sorts of wrong choices that were made along the way.
00:37:57.268 --> 00:38:01.889
And it wasn't because of like any, you know, in many cases.
00:38:02.443 --> 00:38:04.914
It wasn't because of bad actions or bad ideas.
00:38:05.248 --> 00:38:06.748
I think, you know, the companies.
00:38:07.047 --> 00:38:08.768
Try to put forth their best effort.
00:38:08.867 --> 00:38:10.547
That's what Meta should be saying.
00:38:10.797 --> 00:38:16.101
You know, yes, we may have made mistakes, but we made best efforts based on what kind of information we had.
00:38:16.271 --> 00:38:17.240
We took this seriously.
00:38:17.240 --> 00:38:18.681
We wanted to keep people safe.
00:38:18.951 --> 00:38:24.291
And because of the changing nature of the information environment, we had to make decisions on the fly.
00:38:24.641 --> 00:38:30.831
And instead he comes out and he gives this statement, which is basically like, Oh, you know, we took down too much content because there was too much pressure on us.
00:38:30.831 --> 00:38:33.380
Like, you know, come on, stand up, have a spine.
00:38:33.411 --> 00:38:33.960
I, it's.
00:38:34.891 --> 00:38:36.721
It's really, really frustrating to me.
00:38:36.940 --> 00:38:37.690
This was a chance.
00:38:37.690 --> 00:38:44.001
This was an opportunity for them to educate people on how trust and safety works and what the real purpose of trust and safety is.
00:38:44.311 --> 00:38:48.791
And instead he's feeding into the narrative that it's this awful censorship machine.
00:38:48.880 --> 00:38:51.231
And it's, it's really, really frustrating.
00:38:51.411 --> 00:38:51.701
Yeah.
00:38:51.701 --> 00:38:56.764
I mean, I've talked to a few folks recently about trust and safety is kind of marketing problem.
00:38:56.994 --> 00:39:12.880
The fact that, you know, it needs to kind of present itself continually as that difficult challenge, those impossible trade offs, and it needs to kind of completely an ongoing message that out, um, it doesn't help when Nick comes out and gives a, a soft briefing to those journalists that the opposite is true.
00:39:13.271 --> 00:39:28.713
You know, he does say, interestingly, just before we move on about how, AI has all this potential, but right now there are some really pissed off people on, on Metas platforms for the fact that, they make mistakes and remove innocuous or innocent content.
00:39:28.713 --> 00:39:30.074
And, again, you know, just.
00:39:30.541 --> 00:39:33.612
it's kind of completely crazy because it was literally two weeks ago.
00:39:33.612 --> 00:39:42.099
We were talking about how, how threads was an absolute mess of a, of a moderation process and you had all of these kind of, terms that should have been moderated being taken down.
00:39:42.099 --> 00:39:54.132
So I'd love for media to be a bit more critical and, and, uh, clear on, on what it is that senior people like Clegg are saying, and I think that's, that's really part of our job, um, to, to, to do that too.
00:39:54.498 --> 00:40:01.818
they, they, you know, put this in context, like put, put his statements in context and I didn't feel like, like the media was really doing that.
00:40:02.072 --> 00:40:03.152
no, no, indeed.
00:40:03.501 --> 00:40:06.219
Mike, I'm gonna, um, we can't finish on that low note.
00:40:06.268 --> 00:40:08.429
we've got a couple of lighter stories.
00:40:08.509 --> 00:40:16.949
Um, but I'm gonna, throw to you to pick, you had, end with ChatGPT since that's where we started episode.
00:40:17.338 --> 00:40:20.188
Um, tell us about this kind of fun lighthearted story.
00:40:20.603 --> 00:40:20.804
Yeah.
00:40:20.804 --> 00:40:30.358
This was kind of an interesting story where suddenly it started spreading wide that, chat GPT would break if you tried to get it to say anything The name David Mayer.
00:40:31.085 --> 00:40:39.956
I saw it first in a very funny post on blue sky from Andy Bayo, who's is really interesting guy runs, uh, has the site waxy.
00:40:39.956 --> 00:40:40.266
org.
00:40:40.266 --> 00:40:43.525
And he does lots of really interesting stuff, but he had heard that.
00:40:43.525 --> 00:40:57.885
And so he, he created this question for chat GPT, which was combined the first name of the artist who recorded Ziggy Stardust and the last name of the artist who recorded your body is a wonderland into a single name and chat GPT starts.
00:40:58.065 --> 00:41:01.726
And says the artist who recorded Ziggy Sardust is David Bowie.
00:41:01.916 --> 00:41:05.385
And the artist who recorded Your Body is a Wonderland is John Mayer.
00:41:05.585 --> 00:41:11.266
Combining their names and it says David, and then it breaks and it says, I am unable to produce a response.
00:41:11.976 --> 00:41:14.576
It refuses to produce the name David Mayer.
00:41:15.286 --> 00:41:25.925
And what people, yeah, what people then quickly discovered was that there was a short list of names that chat GPT will break if you try and get it to produce those names.
00:41:26.646 --> 00:41:30.036
Then the hunt was on to sort of figure out who they were and why
00:41:31.596 --> 00:41:31.976
odd.
00:41:32.175 --> 00:41:37.286
So, and did they find out what the cause was that, what, what, why doesn't ChatGPT like these names?
00:41:37.335 --> 00:41:48.489
so as far as anyone can tell, and as I don't think open AI has come out and said anything yet, and I will note that they have fixed the David Mayer one, so that it now does work.
00:41:48.489 --> 00:41:51.998
But the other names on the list do not work, still do not work.
00:41:52.059 --> 00:41:52.449
Okay.
00:41:53.039 --> 00:41:57.219
It appears that OpenAI just created a block list.
00:41:58.309 --> 00:41:59.969
It appears to be about six names.
00:41:59.969 --> 00:42:06.159
There might be more, um, but there's a block list that if you try and get it to produce that name, it will break.
00:42:07.378 --> 00:42:10.782
And it just not, you know, it doesn't break in a nice way.
00:42:10.802 --> 00:42:15.083
It just, you know, we'll go halfway through a question and then say, I am unable to produce a response.
00:42:16.242 --> 00:42:35.126
I really love the idea that there's some incredibly well played, you know, some of the smartest kind of machine learning engineers, you know, sat next to some of the world's, brightest trust and safety experts, some of whom we know probably, um, you know, sat together being like, what are the list of six names that we've gotta add to a block list?
00:42:35.391 --> 00:42:35.760
Yeah.
00:42:36.101 --> 00:42:45.351
And it's just, it's so obviously just a straight up lock list that says we will not produce these names and people have gone through and reasoned out some of them and they're all slightly different.
00:42:45.771 --> 00:42:58.538
There's a few that we don't quite know why there was one person, Guido Scorza, who is a data protection expert in Italy, and he had posted that he had used the GDPRs.
00:42:58.746 --> 00:43:05.260
right to erasure process against open AI and said, basically you have to forget everything about me.
00:43:05.681 --> 00:43:13.420
And the only way that they could figure out to do so was to put him on this block list Because they don't actually remember anything about him.
00:43:13.431 --> 00:43:19.143
Like there's no this is the thing that a lot of people don't Understand about these systems is it's not a big database.
00:43:19.143 --> 00:43:20.784
It's not going in and collecting stuff.
00:43:21.003 --> 00:43:22.673
It is training and just thinking about stuff.
00:43:22.673 --> 00:43:33.190
So the only way to say like, never produce any information on Guido Scorza is to put them on a, I mean, there are better ways, but this was the sort of fast and quick way.
00:43:33.510 --> 00:43:43.333
There are other people, you know, the, Brian Hood was a mayor in Australia who claimed that it produced defamatory content about him and he threatened to sue.
00:43:43.583 --> 00:43:43.943
And so.
00:43:44.318 --> 00:43:46.007
that one probably came up first.
00:43:46.018 --> 00:43:50.608
Somebody had actually told me about that one, like a year and a half ago when it, when it came up and I had meant to investigate it.
00:43:50.608 --> 00:43:51.797
And then I just never got to it.
00:43:52.168 --> 00:43:58.947
Jonathan Turley, who's a law professor and sort of a famously sort of Trumpian law professor.
00:43:59.228 --> 00:44:03.858
Went on this big rant of claiming also that chat GPT had defamed him.
00:44:03.867 --> 00:44:11.398
And so he's on the The weird one is Jonathan Zatrain, who I know, and asked like, what the hell, what
00:44:11.498 --> 00:44:12.047
yeah,
00:44:12.378 --> 00:44:13.327
to get off this list?
00:44:13.557 --> 00:44:15.097
And he has no idea.
00:44:15.467 --> 00:44:16.768
right, right.
00:44:17.010 --> 00:44:18.693
he doesn't know what's going on.
00:44:18.724 --> 00:44:25.264
I joked with him that like, Oh, you know, are you trying to come up with something to prove a point or something about all this?
00:44:25.264 --> 00:44:26.204
He's like, I did nothing.
00:44:28.103 --> 00:44:28.603
So,
00:44:28.813 --> 00:44:31.574
And he didn't know he was on ChachiPT's naughty step.
00:44:31.594 --> 00:44:33.713
he, he had found it, he had actually discovered it.
00:44:33.724 --> 00:44:42.411
So he, he added himself when everyone was pointing out to this list of names, he pointed out, like he had discovered a few months ago that chat GPT breaks on his name.
00:44:42.773 --> 00:44:44.304
and so, you know, it's just.
00:44:44.548 --> 00:44:46.688
Basically, it's like this really simple way.
00:44:46.739 --> 00:44:52.998
I understand probably why it happened because all these people, except for Jonathan, as far as we know, got really mad.
00:44:53.009 --> 00:44:54.898
We still don't know which David Mayer there is.
00:44:54.898 --> 00:44:55.699
There's speculation.
00:44:55.708 --> 00:45:00.068
There's a few different, there's also a David Farber, Faber, and nobody's sure which one that is.
00:45:00.409 --> 00:45:05.865
but the other ones, like people got mad and open AI probably was like, you know what, this guy is a pain.
00:45:06.016 --> 00:45:06.775
Fucking headache.
00:45:07.666 --> 00:45:10.206
And therefore just put them on the ban list.
00:45:10.206 --> 00:45:14.735
Like, let's not deal with this because we don't want to go through a lawsuit and have to do this.
00:45:14.735 --> 00:45:19.295
And there's no other way to like effectively, because they're, you know, they don't understand.
00:45:19.295 --> 00:45:21.759
It's not that chat GPT is defamatory.
00:45:21.759 --> 00:45:24.708
It's whatever the prompt is led to defamation.
00:45:25.039 --> 00:45:31.389
And it's not that we're collecting data on this guy, but you know, he thinks we are, and do we really want to fight it in court?
00:45:31.429 --> 00:45:32.539
I just put them on the ban
00:45:32.753 --> 00:45:33.664
Put him on the list.
00:45:33.793 --> 00:45:36.364
I'm pretty sure that's the kind of thing that happened.
00:45:36.744 --> 00:45:46.617
You know, you would hope that a company with as many resources, and as you noted, as many, smart engineers would come up with a more sophisticated way than this, but they haven't.
00:45:46.668 --> 00:45:53.547
And so, you know, I know how things like that happen and I'm sure it was just kind of like, ah, just make this headache go away in the easiest way possible.
00:45:53.827 --> 00:45:56.760
And that means every time you try and produce this name, we're going to break it.
00:45:57.030 --> 00:45:57.550
Yeah.
00:45:57.601 --> 00:45:58.170
Interesting.
00:45:58.673 --> 00:46:05.951
just as an aside, there's some, there's the OpenAI put out some research in the past week around kind of red teaming in new, new and novel ways.
00:46:06.286 --> 00:46:08.135
I wonder if they should be reading their own research.
00:46:08.971 --> 00:46:12.121
Yeah, yeah, absolutely, absolutely.
00:46:12.465 --> 00:46:23.539
so yeah, if you, if David May is listening, to the podcast, if he's, if he's a regular listener, um, make yourself known, um, we would love to know what you did to be put on the, uh, on the shit list of, uh,
00:46:23.753 --> 00:46:24.153
Yeah.
00:46:24.184 --> 00:46:25.324
And then you got taken off.
00:46:25.324 --> 00:46:28.023
So if you have a legal complaint, get it, get it going
00:46:28.393 --> 00:46:30.474
Yeah, we know some lawyers.
00:46:30.713 --> 00:46:31.903
Um, great.
00:46:31.903 --> 00:46:32.443
Thanks Mike.
00:46:32.614 --> 00:46:36.177
Um, that's not the, you know, that's not the kind of end of today's episode.
00:46:36.177 --> 00:46:37.641
It's, we've got a great bonus chat.
00:46:37.641 --> 00:46:40.110
I don't know if you want to kind of give us a preview of that.
00:46:40.481 --> 00:46:40.780
Yeah.
00:46:40.780 --> 00:46:41.708
So this was really great.
00:46:41.708 --> 00:46:43.438
I had this discussion just yesterday.
00:46:43.514 --> 00:46:45.088
it's a topic that we've talked about a lot.
00:46:45.298 --> 00:46:50.159
The social media age bands in Australia and how they're popping up elsewhere.
00:46:50.338 --> 00:46:58.469
and so this is a discussion I had with, two folks from the internet society, Natalie Campbell, who's the senior director of North American government and regulatory affairs.
00:46:58.509 --> 00:47:02.438
And John Perrino, who's a senior policy and advocacy expert.
00:47:02.806 --> 00:47:08.135
And they're concerned about these age restrictions and age verification.
00:47:08.416 --> 00:47:27.146
And, we talked about Australia and also Canada has a bill that is, moving forward that one of the incredible things in the chat was that, it would require places like Starbucks to verify the age of anyone who wants to use the wifi in Starbucks and, you know, all sorts of stuff.
00:47:27.146 --> 00:47:29.536
And so they, they're sort of concerned about.
00:47:29.847 --> 00:47:36.887
these laws, the proliferation of these laws, what it means for the open internet and what age verification requirements would mean.
00:47:37.117 --> 00:47:40.434
The really fascinating discussion and we'll go to that right now.
00:47:51.025 --> 00:47:52.762
Natalie, John, welcome to the podcast.
00:47:52.802 --> 00:47:53.911
Glad to have you here.
00:47:54.061 --> 00:48:01.422
wanted to start by talking about the law that just passed in Australia, which effectively bans those under 16 from social media.
00:48:01.827 --> 00:48:12.271
John, can you talk through the details of that law, including sort of which sites are impacted and how do you, Does the Australian government expect sites to know the age of their visitors?
00:48:14.041 --> 00:48:30.293
great question, Mike, and, you know, for, for quick background for listeners, this legislation moved through in about less than two weeks and there was a consultation period open less than 24 hours.
00:48:30.804 --> 00:48:35.324
Um, the Australian social media age verification stuff really, really flew through.
00:48:35.786 --> 00:48:48.009
And there's a lot that we still don't know, even though the legislation passed, especially on what social media platforms would be required to do in order to comply with this.
00:48:48.648 --> 00:48:55.719
So right now, essentially all we know is that government ID would not be required.
00:48:56.268 --> 00:49:02.148
That should include things like a passport or government issued driver's license, also a digital ID.
00:49:02.748 --> 00:49:09.820
And we know that the social media is for Australians under the age of 16.
00:49:10.550 --> 00:49:18.501
and then the final thing that was kind of added on late was a digital duty of care, which again is to be determined.
00:49:18.940 --> 00:49:23.451
So the bottom line on this is much of the bill is to be determined.
00:49:23.690 --> 00:49:32.393
They don't know what age verification would be, as they say, reasonable for the social media platforms to use.
00:49:32.873 --> 00:49:47.996
And as a lot of the comments pointed out, there was actually already an age verification study ongoing, and that's still going on, which tools, maybe work best, what the trade offs are in different age verification methods.
00:49:48.237 --> 00:49:50.146
The Australian government was already doing this.
00:49:50.527 --> 00:49:51.856
They don't have the results.
00:49:51.916 --> 00:49:53.396
They probably won't have.
00:49:53.681 --> 00:49:59.512
The results until about six months before the social media platforms have to comply with this.
00:50:00.295 --> 00:50:05.795
I know that we've seen a bunch of other countries also exploring similar ideas around age verification.
00:50:06.101 --> 00:50:17.561
on the podcast, we've talked about it in the UK and somewhat in the US to a lesser extent, there are a couple of different issues in the US, but also now Canada is exploring a similar issue.
00:50:17.661 --> 00:50:22.168
So, Natalie, can you talk through what the proposal in Canada is about?
00:50:22.876 --> 00:50:33.052
Sure, so, Senate Bill S 210, this is an act to restrict young persons online access to sexually explicit material, is very close to becoming a law.
00:50:33.532 --> 00:50:36.802
This bill has been flying under the radar for quite a long time.
00:50:36.822 --> 00:50:41.233
I don't think that people are taking it very seriously because most Senate bills.
00:50:41.628 --> 00:50:42.726
Don't make it to law.
00:50:43.190 --> 00:50:55.175
but this has found its way into the very late stages of our parliamentary process, and it's probably the most dangerous bill to the Internet, in Canada right now.
00:50:55.434 --> 00:51:00.559
and one of the main reasons for that is because of its age verification mandates.
00:51:00.733 --> 00:51:07.474
That would apply to virtually every intermediary on the internet and some that are not on the internet as well.
00:51:08.054 --> 00:51:17.083
So essentially what the bill tries to do is to prevent young people from access to sexually explicit material, which is defined extremely broadly.
00:51:17.409 --> 00:51:31.268
and makes it so that every intermediary that would play a part in facilitating sexually explicit material for profit online would have to verify a user's age or face very high fines.
00:51:31.757 --> 00:51:51.088
So, this, you know, is this, is this, Probably the most extreme age verification proposal we've seen so far, because it's not just targeting websites, it's targeting internet service providers, content delivery networks, search engines, email services, even a Starbucks location that's providing access to Wi Fi.
00:51:51.509 --> 00:51:58.128
Because what content is flowing through this pipes could be sexually explicit material.
00:51:58.545 --> 00:52:05.248
they now have a duty to verify users age and in doing so like it's, it's different.
00:52:05.579 --> 00:52:33.739
When we're talking about websites doing this, but when we're thinking about every single intermediary on the Internet having a duty to verify users age, that gets really problematic when we're thinking about an open Internet first, because most traffic is encrypted, and it's not possible for most infrastructure intermediaries to know what's flowing through the pipes, and even if they could, content doesn't flow through the Internet as a whole piece of content.
00:52:34.108 --> 00:52:35.039
It's packets.
00:52:35.509 --> 00:52:46.329
And so it becomes extremely difficult to, one, identify, what is sexually explicit material and, you know, is Canada's definition of this type of content.
00:52:46.679 --> 00:52:59.846
And then, second of all, having to, decide to try and identify that stuff means you can't use things like encryption, which is the foundation of security for every service and user on the Internet.
00:53:00.585 --> 00:53:13.255
So it's very concerning in that, the implications for security online, but also the fact that you're placing huge barriers to access because now is an Internet user, because a lot of intermediaries.
00:53:13.550 --> 00:53:16.724
Won't know what's sexually explicit material.
00:53:16.733 --> 00:53:40.297
They might just start doing age verification to Everyone and that means I now, you know as somebody who's based in canada would have to trust a whole lot of entities with Very personal information whether it's government id or you know biometrics i'm having to trust a lot of third parties Who I might not have?
00:53:40.297 --> 00:53:48.077
You Any direct relationship with with my personal information, which is a huge barrier to privacy and anonymity online.
00:53:48.621 --> 00:53:55.054
so this is, from an open Internet aspect, this is super problematic because you're creating huge barriers to access.
00:53:55.539 --> 00:53:58.528
for people who might not be able to get government issued I.
00:53:58.528 --> 00:53:58.778
D.
00:53:58.778 --> 00:54:07.048
or might not be able to use the Internet without the promise of anonymity, which is a huge hurdle for a lot of people in marginalized communities and young people as well.
00:54:07.599 --> 00:54:12.108
But also the fact that this has huge implications for security online.
00:54:12.626 --> 00:54:28.398
and you know, not Enabling intermediaries to use encryption that could be extremely devastating to people's security online and making people vulnerable to a whole range of bad stuff that I don't think was the intention of this bill.
00:54:28.768 --> 00:54:31.139
Yeah, I mean you, you mentioned just really quickly.
00:54:31.139 --> 00:54:32.349
I wanted to follow up on one point.
00:54:32.509 --> 00:54:45.846
You mentioned the fact that since it hits every intermediary, the potential that, like a Starbucks, if you wanted internet in a Starbucks, is the idea there that, Starbucks would have to not just check your ID, but like record it somehow?
00:54:45.846 --> 00:54:51.135
Is that part of the fear that if you want to get on the Wi Fi at Starbucks, you have to first prove how old you are?
00:54:51.659 --> 00:55:09.699
Well, this is the problem is that the way that it defines who this bill applies to is internet service provider, but the definition it uses for that is a person who provides access to the Internet Internet content hosting or electronic mail.
00:55:10.588 --> 00:55:24.949
The person who provides internet access could mean Starbucks and could mean that, you Starbucks has to know what are you looking at when you're accessing the internet or just say, just show us your ID if you want to use the Wi Fi, right?
00:55:24.949 --> 00:55:30.096
So, problematic in both senses and just a reminder that please use VPNs.
00:55:30.456 --> 00:55:30.485
Okay.
00:55:32.101 --> 00:55:38.650
So, I mean, so we're seeing this all over, obviously, Australia and Canada, as you guys discussed, we've discussed the UK in the past.
00:55:38.670 --> 00:55:42.423
There are various bills in the U S that also touch on this elsewhere around the world.
00:55:42.804 --> 00:55:44.653
What should we make of this trend?
00:55:44.653 --> 00:55:47.244
What, you know, what, why is this all, all of a sudden happening?
00:55:47.244 --> 00:55:49.673
We've had the internet obviously for decades now.
00:55:49.793 --> 00:55:56.947
Why is it suddenly that everyone is trying to pass these kind of age related bans or age verification.
00:55:57.134 --> 00:55:59.824
ideas, John, do you want to start there?
00:56:00.248 --> 00:56:09.239
Yeah, I mean, to that point, it certainly feels like everything is happening all at once everywhere in every corner of the world.
00:56:09.878 --> 00:56:18.568
I mean, we're seeing this, as you mentioned, you know, the UK has been working on this almost a decade, but their age verification guidance comes out in January.
00:56:19.389 --> 00:56:21.088
The European Union is working on this.
00:56:21.583 --> 00:56:24.563
There are almost two dozen U.
00:56:24.563 --> 00:56:24.704
S.
00:56:24.704 --> 00:56:30.693
states that have passed age verification for adult websites, more with social media.
00:56:31.329 --> 00:56:32.597
it really is happening everywhere.
00:56:33.068 --> 00:56:33.557
Why?
00:56:33.648 --> 00:56:35.887
You know, it's a whole bunch of factors.
00:56:35.987 --> 00:56:46.331
Um, one kind of interesting story with the Australia legislation is it seems to have stemmed depending on who you ask, from Jonathan Haidt's book.
00:56:46.489 --> 00:56:52.623
and, you know, there's, has been a lot of discussion, because of recent literature.
00:56:53.338 --> 00:56:58.329
The Anxious Generation is Jonathan Height's book, as I'm sure many of your listeners are familiar with.
00:56:58.759 --> 00:57:00.778
There's just been more discussion on this.
00:57:00.818 --> 00:57:02.478
There's been a lot of discussion in the U.
00:57:02.478 --> 00:57:02.918
S.
00:57:02.938 --> 00:57:08.239
You know, the Kids Online Safety Act, but there are, you know, true age verification bills being introduced.
00:57:08.838 --> 00:57:11.079
There's even a case going to the Supreme Court.
00:57:11.411 --> 00:57:13.402
And that's the Free Speech Coalition v.
00:57:13.443 --> 00:57:14.222
Paxton case.
00:57:14.242 --> 00:57:17.967
And that's something that the Internet Society Weight in on with an amicus.
00:57:17.967 --> 00:57:24.710
We joined with the Center for Democracy and Technology, New America's Open Technology Institute, and some academics on that case.
00:57:25.034 --> 00:57:33.414
And, you know, the thing that we think doesn't get talked about enough is that age verification laws are not just about young people.
00:57:33.902 --> 00:57:34.625
to
00:57:34.824 --> 00:57:38.650
verification laws are about everyone, if not done right.
00:57:39.318 --> 00:57:56.085
You know, this puts everyone's privacy security at risk and like Natalie already said, you know, this can be discriminatory this the marginalized communities and there's really interesting research on this marginalized communities can benefit the most from social media can benefit the most from being online.
00:57:56.735 --> 00:57:58.706
It can be more difficult for them to get online.
00:57:58.905 --> 00:58:01.106
There can be more social factors that make it.
00:58:01.181 --> 00:58:08.670
Um, you know, generally their worlds, it's more difficult.
00:58:08.864 --> 00:58:10.543
To make those types of connections.
00:58:10.893 --> 00:58:16.873
so what we're really focused on is that everyone can get access to the Internet that is not.
00:58:17.434 --> 00:58:23.443
Gating off access to the Internet, to news, information, health, entertainment, right?
00:58:23.800 --> 00:58:32.201
so this is, this is really going to be a challenge and we see so many pieces of regulation that are being implemented and introduced right now.
00:58:32.543 --> 00:58:35.481
luckily at the Internet Society, we have an incredible community.
00:58:36.018 --> 00:58:38.277
for instance, our Australia chapter.
00:58:38.757 --> 00:58:39.637
Jumps to action.
00:58:39.637 --> 00:58:48.233
There was less than 24 hours to file, comments, they made the deadline, got the comments in, you know, and that's so important.
00:58:48.233 --> 00:58:57.561
And so many of our, our local, you know, regional chapters are engaging with governments in their home countries on this issue.
00:58:58.021 --> 00:59:01.891
So on that point, you know, there's really good debate on this.
00:59:02.326 --> 00:59:03.786
really great interaction.
00:59:04.076 --> 00:59:09.195
More and more technologists are getting involved in this, more standards development organizations.
00:59:09.445 --> 00:59:13.985
So the engineers who actually make the internet function, right, are getting involved on this.
00:59:14.266 --> 00:59:17.706
So that's encouraging, but it really is happening everywhere at once.
00:59:18.065 --> 00:59:27.938
And we really need to make sure that those who need the internet the most, can actually be safe online and are not having their privacy and security exposed through all this.
00:59:29.903 --> 00:59:47.173
you've both talked a little bit about sort of the implications of this and the reasons why there are concerns, but Natalie, I want to finish up with you and just say, you know, for policymakers, some of whom might be listening to this, hopefully, um, who are looking at these laws and thinking about them, what kinds of.
00:59:47.454 --> 00:59:53.833
Factors, what should they be considering before, you know, proposing or voting for these kinds of laws?
00:59:54.436 --> 01:00:01.228
So, first, like, it's really important to understand that the Internet Society works to make sure that the Internet is for everyone.
01:00:01.259 --> 01:00:08.489
And when we talk about an open Internet, we're talking about making sure we're lowering barriers to access to the Internet.
01:00:08.994 --> 01:00:11.074
We also want a healthy Internet.
01:00:11.094 --> 01:00:18.833
We are, you know, like John mentioned, we're a huge community, not just the Internet Society, but our chapters and members and organizational members around the world.
01:00:19.193 --> 01:00:23.994
We all want a healthy Internet and care about making sure there are safe spaces for people online.
01:00:24.664 --> 01:00:31.373
But we also want to make sure that, you know, things like encryption and the fundamentals of a secure Internet.
01:00:31.903 --> 01:00:41.273
Are not undermined and that people don't experience barriers to access, that could be complete hurdles to accessing the Internet in the first place.
01:00:41.813 --> 01:00:50.831
So, I mean, it's not to say that there might never be a solution for age verification that could not hinder things like security and open Internet.
01:00:51.400 --> 01:00:56.951
But policymakers really do have to be thinking through how could their proposals.
01:00:57.177 --> 01:01:00.927
Impact, people's access to the Internet and their safety online.
01:01:01.438 --> 01:01:07.128
And we have a tool that helps us analyze these proposals is called the Internet Impact Assessment Toolkit.
01:01:07.458 --> 01:01:11.403
We think of it like an environmental impact assessment for the Internet.
01:01:11.824 --> 01:01:21.240
And so what we do, is, um, Offer ourselves to work with governments who are thinking about, whatever issues are working out that may relate to the Internet.
01:01:21.251 --> 01:01:32.280
And we'll often use this framework that describes what the Internet needs to exist in the 1st place and to be more open, global connected, secure and trustworthy and we help them think through how might.
01:01:32.583 --> 01:01:36.893
Particular proposal impact these goals for a healthy Internet.
01:01:37.623 --> 01:01:45.278
And so always are available to work with governments to think through these aspects, but we have a toolkit that policymakers can use themselves.
01:01:45.858 --> 01:02:03.483
And we think that it's really important that, We just don't jump to law proposals that don't consider those impacts on the internet because such as a case in Canada, there can be very extreme consequences for people's access to the internet and their very safety and security online.
01:02:03.947 --> 01:02:04.447
All right.
01:02:04.478 --> 01:02:10.771
Well, Natalie and John, thank you for coming on the podcast and thank you for all the good work that the Internet Society does.
01:02:11.170 --> 01:02:25.181
And, uh, I hope if anyone's listening to this and you are, working on a bill like this or thinking about these kinds of laws that you, uh, listen closely and, take a look at what the Internet Society is doing on this and, and what, resources they have available.
01:02:25.731 --> 01:02:26.768
Thanks again for joining us.
01:02:27.103 --> 01:02:27.742
Thanks Mike.
01:02:27.891 --> 01:02:28.387
Thanks, Mike.
01:02:31.677 --> 01:02:33.818
Thanks for listening to Ctrl-Alt-Speech.
01:02:34.007 --> 01:02:37.197
Subscribe now to get our weekly episodes as soon as they're released.
01:02:37.998 --> 01:02:44.768
If your company or organization is interested in sponsoring the podcast, contact us by visiting ctrlaltspeech.Com.
01:02:44.958 --> 01:02:47.157
That's C T R L Alt Speech.
01:02:47.518 --> 01:02:47.847
com.
01:02:48.748 --> 01:03:01.818
This podcast is produced with financial support from the Future of Online Trust and Safety Fund, a fiscally sponsored multi donor fund at Global Impact that supports charitable activities to build a more robust, capable, and inclusive trust and safety ecosystem.
00:00:00.000 --> 00:00:02.350
So I don't know how we haven't already used this prompt mic.
00:00:03.129 --> 00:00:07.690
Perhaps we should be feeding all of the podcasts into chat GPT to, to tell us.
00:00:08.369 --> 00:00:13.640
But, um, I'm going to ask you today, the prompt that you get on chat GPT when you go onto the site.
00:00:14.060 --> 00:00:15.369
And that is what can I help with?
00:00:17.850 --> 00:00:23.350
Can you find a thousand or so people who would like to purchase my social media card game?
00:00:25.269 --> 00:00:25.600
What?
00:00:25.629 --> 00:00:25.890
Yeah.
00:00:25.890 --> 00:00:26.960
1 billion users.
00:00:26.960 --> 00:00:27.839
Let's get the plugin.
00:00:28.000 --> 00:00:37.012
Yeah, 1 billion users, which I will note this week OpenAI announced that they are aiming to get 1 billion users on chat GPT.
00:00:37.023 --> 00:00:38.383
So this all ties together
00:00:38.512 --> 00:00:39.203
Oh my God.
00:00:39.203 --> 00:00:42.332
It's, do you think they, they took inspiration from the Kickstarter?
00:00:42.573 --> 00:00:46.740
They must have, but I would like them to send 1 billion users to the Kickstarter.
00:00:46.770 --> 00:00:52.520
We are not quite at our threshold of support to actually make the game go into production.
00:00:52.820 --> 00:01:00.939
But if you are listening to this, I know you will like this game because it is all about social media and it is fun for the whole family to play.
00:01:01.420 --> 00:01:06.840
So if you're listening to this and you have not backed our 1 million users, 1 billion users.
00:01:07.355 --> 00:01:08.325
Game on Kickstarter.
00:01:08.325 --> 00:01:10.974
You can just go to Kickstarter and search 1 billion users.
00:01:11.194 --> 00:01:12.495
Please, please check it out.
00:01:12.504 --> 00:01:13.715
We need more backers.
00:01:13.754 --> 00:01:25.075
And if you know other people who are interested in online speech and trust and safety, and you want to have a fun game to play that will also help you explain trust and safety and social media and all this kind of stuff.
00:01:25.234 --> 00:01:26.194
Please, please do it.
00:01:26.224 --> 00:01:27.974
I hate to beg, but I need to beg.
00:01:29.109 --> 00:01:31.230
I'm, I'm going to buy five more.
00:01:32.265 --> 00:01:32.844
There we go.
00:01:32.864 --> 00:01:33.715
We have a discount.
00:01:33.724 --> 00:01:35.594
If you buy five, we have a discount,
00:01:35.844 --> 00:01:36.584
I'm going to do it.
00:01:36.775 --> 00:01:39.724
but anyways, Ben, what can I help you with today?
00:01:39.965 --> 00:01:40.584
What can you help with?
00:01:40.594 --> 00:01:43.355
I am incredibly jet lagged, Mike.
00:01:43.385 --> 00:01:47.795
I've just landed back from your fair nation after a week of work there.
00:01:47.864 --> 00:01:50.265
And I do not know what day or time it is.
00:01:50.334 --> 00:01:56.871
Um, I have not figured out how, To, yeah, to actually speak as you can tell.
00:01:57.141 --> 00:02:03.341
Um, so any, any advice, any help on how to get back to a normal rhythm of sleeping would be much appreciated.
00:02:03.941 --> 00:02:04.902
Let's see how this goes.
00:02:05.272 --> 00:02:07.561
Yes, yes, this, this may be an interesting one.
00:02:07.621 --> 00:02:09.181
Uh, I can confirm that.
00:02:09.472 --> 00:02:14.861
Well, one, I was up late last night working on stuff and I was texting Ben who was in Chicago.
00:02:15.337 --> 00:02:17.098
this was not that many hours ago
00:02:17.617 --> 00:02:17.967
No,
00:02:18.448 --> 00:02:19.693
and now he is.
00:02:19.853 --> 00:02:28.902
In London, and he showed me his, bag that he had just thrown in the corner after getting back home in time to record this.
00:02:28.953 --> 00:02:33.932
So we are both working on quite little sleep and, uh, Ben even less than I.
00:02:33.932 --> 00:02:39.893
So this may be a very fun or very strange episode, depending on where things go.
00:02:39.893 --> 00:02:40.828
Bye.
00:02:40.927 --> 00:02:41.388
everyone.
00:02:49.512 --> 00:02:54.913
Hello and welcome to control all speech, your weekly roundup of the major stories about online speech.
00:02:55.173 --> 00:02:57.402
Content moderation and internet regulation.
00:02:57.883 --> 00:02:59.832
It's December the 6th, 2024.
00:03:00.152 --> 00:03:04.462
And this week's episode is brought to you with financial support from the future of online trust and safety fund.
00:03:04.812 --> 00:03:14.310
And by today's sponsor internet society, a global nonprofit that advocates for an open globally connected, secure, and trustworthy internet for everyone.
00:03:14.956 --> 00:03:15.597
I'm Ben Whitelaw.
00:03:15.597 --> 00:03:17.597
I'm the founder and editor of everything in moderation.
00:03:18.012 --> 00:03:27.701
at least I, I am, or was the last time I checked, um, and I'm with an equally kind of bleary eyed Mike Masnick, for altogether different reasons.
00:03:28.551 --> 00:03:29.361
It's good to be here, Mike.
00:03:29.382 --> 00:03:32.231
We haven't been in this, in these respective chairs for a few weeks.
00:03:32.627 --> 00:03:33.497
Yeah, that's right.
00:03:33.497 --> 00:03:34.986
We took off last week.
00:03:35.016 --> 00:03:41.189
We missed all the big stories, uh, during Thanksgiving week and then the week before that you were traveling.
00:03:41.189 --> 00:03:42.310
So you were not here.
00:03:42.340 --> 00:03:47.009
And so, uh, it's been, been a little while since, since we've, been able to do this together.
00:03:48.210 --> 00:03:52.847
Um, we, we've got a lot to cover today, partly because yeah, we were, we took a break last week.
00:03:53.073 --> 00:03:55.560
happy Thanksgiving again for, to you and to our listeners.
00:03:55.860 --> 00:04:07.032
And we have a really good, bonus chat at the end of today's episode with some super smart folks from the internet society, around social media bands in Canada and Australia and age verification in general.
00:04:07.062 --> 00:04:08.995
So, stick around for that at the end.
00:04:08.995 --> 00:04:09.159
Thanks.
00:04:09.300 --> 00:04:14.823
Today's show, and we won't cover those kinds of stories today, just because we do that really well with, with those folks.
00:04:15.099 --> 00:04:19.103
we're going to dive straight in to, um, what is essentially Mike a breaking story.
00:04:19.432 --> 00:04:31.466
Um, it's, it's a live story that you essentially woke up, rubbed your eyes and, uh, have dived into, it's the, uh, the tick tock case it's back on the agenda.
00:04:32.067 --> 00:04:32.646
Yeah.
00:04:32.757 --> 00:04:33.247
So.
00:04:33.637 --> 00:04:36.857
We knew that there was supposed to be a ruling coming down basically any day.
00:04:36.857 --> 00:04:51.637
This is the, US law that in theory forces ByteDance to divest of TikTok in the US or forces, the app stores and ISPs to block access to it.
00:04:52.107 --> 00:04:55.060
And they are supposed to either divest.
00:04:55.250 --> 00:04:58.867
or we'll have access blocked on January the 19th.
00:04:58.908 --> 00:05:16.964
And there's a whole sort of political side story to this, which is that our incoming returning president, uh, Donald Trump, who originally was the one who proposed the TikTok ban and tried to do so in a very ham fisted way that failed, four years ago.
00:05:17.459 --> 00:05:29.120
Has now completely flipped positions, perhaps because he was backed heavily by one of, ByteDance's biggest investors, but has insisted that he's going to protect TikTok.
00:05:29.279 --> 00:05:35.845
But he comes into office, I think a day or two after this law goes into effect.
00:05:35.894 --> 00:05:41.172
And so TikTok had challenged the law and it went to the DC circuit, which is in the appeals court.
00:05:41.701 --> 00:05:48.041
And they came down with their ruling today, basically saying that the law is.
00:05:48.442 --> 00:05:49.072
Just fine.
00:05:49.122 --> 00:05:53.062
There is no problems with the law and TikTok must obey it.
00:05:53.062 --> 00:05:54.641
And ByteDance must obey it.
00:05:54.968 --> 00:06:00.007
this ruling did come out literally, about an hour before we were recording.
00:06:00.158 --> 00:06:04.997
I have read what I think are the most salient points in it.
00:06:05.269 --> 00:06:11.927
however, given the timing, I may make some mistakes in terms of this because I have not had a chance to go through it in great detail.
00:06:11.927 --> 00:06:13.266
Have you even had a coffee Mike?
00:06:13.357 --> 00:06:13.447
Are
00:06:13.516 --> 00:06:25.040
I have not actually, I am, I am, uh, yeah, anyways, so, I think they're wrong is, is, is my quick summary.
00:06:25.663 --> 00:06:31.112
it's interesting because the government tried to argue that there were no first amendment issues at all.
00:06:31.132 --> 00:06:34.752
And that the case could be decided without even considering the first amendment.
00:06:35.533 --> 00:06:39.709
The court went in a different direction and said that government was wrong, but that the first amendment.
00:06:40.149 --> 00:06:41.269
Does apply here.
00:06:41.730 --> 00:06:52.259
And then once you say that the First Amendment does apply, then there's a question of which level of scrutiny, and there are different levels, you know have to be passed in order for the law to be considered constitutional.
00:06:52.550 --> 00:06:55.214
The highest bar is is strict scrutiny.
00:06:55.574 --> 00:06:56.995
And then there are, there are lower bars.
00:06:56.995 --> 00:07:01.894
And the government had argued that if the First Amendment does apply, you should go for one of the lower bars.
00:07:02.675 --> 00:07:06.824
ByteDance had argued that it should be strict scrutiny, which is the hardest to pass.
00:07:07.555 --> 00:07:14.848
The court notes that there isn't a comparable situation that says which of the levels of scrutiny should apply.
00:07:15.197 --> 00:07:19.168
And so they actually choose not to, say which level should apply.
00:07:19.168 --> 00:07:33.084
Though there is a concurring, uh, opinion by one of the judges suggesting, which level should apply, but the court says it doesn't matter because we think that this law passes even the highest bar strict scrutiny.
00:07:33.495 --> 00:07:36.035
And so then they just analyze it based on strict scrutiny.
00:07:36.584 --> 00:07:37.665
And I think they do it.
00:07:37.894 --> 00:07:41.675
in my one read through it, I think they do a terrible job of it.
00:07:42.115 --> 00:07:47.170
They basically buy into strongly buy into the claims by the U.
00:07:47.170 --> 00:07:47.329
S.
00:07:47.329 --> 00:07:57.100
government of the national security concerns, even though they admit that no actual evidence is presented for those, but they basically say, well, the U.
00:07:57.100 --> 00:07:57.240
S.
00:07:57.240 --> 00:07:58.129
government, a lot of people in the U.
00:07:58.129 --> 00:07:58.259
S.
00:07:58.259 --> 00:08:03.389
government say that we should be concerned and therefore we're just going to take that as fact.
00:08:03.550 --> 00:08:06.040
This is not uncommon in the U.
00:08:06.040 --> 00:08:06.300
S.
00:08:06.779 --> 00:08:12.122
court system when the government says, well, we have national security and blah, blah, blah.
00:08:12.122 --> 00:08:13.302
We can't tell you what they are.
00:08:13.302 --> 00:08:14.572
We're just really concerned.
00:08:14.920 --> 00:08:17.689
the courts are often willing to go along with that.
00:08:17.870 --> 00:08:25.427
That has led to a whole bunch of really terrible things having to do with civil liberties and civil rights and all sorts of stuff going back decades.
00:08:25.826 --> 00:08:43.325
But that seems to happen again here where they're just like national security concerns seem totally legit and you know ByteDance hasn't given us anything to Reject that they admit multiple times that a lot of these concerns are totally speculative And yet because ByteDance can't respond to the questions Speculative concerns.
00:08:43.764 --> 00:08:45.304
Therefore, this is all okay.
00:08:45.884 --> 00:08:57.111
There's a whole bunch of other like little stuff in here that again, it's just kind of like, well, you know, by chance I said, we have all these other, less restrictive means of doing this, which is part of the strict scrutiny test.
00:08:57.432 --> 00:09:00.845
And the court is like, nah, you know, we don't think any of those good enough.
00:09:00.845 --> 00:09:07.065
I mean, you know, Bytance really pushed this idea that, they would separate out all the operations, which they've mostly done.
00:09:07.294 --> 00:09:09.384
Everything is hosted in the U S by Oracle.
00:09:09.384 --> 00:09:11.504
They've given Oracle the power to audit stuff.
00:09:11.788 --> 00:09:25.977
they even offered the U S government, like an off switch, like that they could, you know, push a button and turn off TikTok, you know, in the case that something really bad happened and the government was like, nah, you know, that doesn't really satisfy what the government is doing here.
00:09:26.268 --> 00:09:28.048
There are a bunch of other oddities in here.
00:09:28.077 --> 00:09:36.231
You know, there were issues around like specifically targeting, like the bill actually names Tik TOK, which seems like it's singling them out for different treatment.
00:09:36.261 --> 00:09:43.298
And the court kind of waves that off and says, well, no, that's not really true, even though it does mention it.
00:09:43.409 --> 00:09:45.631
it, they said at one point, it's not punishment.
00:09:45.631 --> 00:09:54.292
If it was a bill of attainder, which is a specific thing where it's like just targeting someone for punishment, they said, yes, it only names TikTok, but not for punishment.
00:09:54.701 --> 00:09:56.631
And therefore it's not a bill of attainder.
00:09:56.861 --> 00:09:59.562
There are all of these like really odd things.
00:09:59.922 --> 00:10:02.682
So the big question now is kind of like what happens next?
00:10:02.682 --> 00:10:04.701
Like does TikTok.
00:10:04.917 --> 00:10:08.996
get shut off on January 19th and be gone for two days.
00:10:08.996 --> 00:10:14.352
And then Trump comes into office and suddenly changes position and allows to talk to come back.
00:10:14.523 --> 00:10:15.702
I don't know.
00:10:15.702 --> 00:10:28.513
I mean, I, I think, and I could be, there may be, there's some procedural weirdness in terms of how this law was written in terms of like forcing it to go straight to the circuit court rather than a district court, which would be the normal thing.
00:10:28.942 --> 00:10:35.518
I assume that they can now go to the Supreme Court and ask for a stay on this ruling, in order to appeal it.
00:10:35.518 --> 00:10:47.868
And maybe that puts the decision off past, January 19th and allows Trump to get into office and then sort of, you know, do something to keep TikTok, unless he decides that he, he maybe doesn't want TikTok.
00:10:47.937 --> 00:10:49.758
We, we don't really know at this point.
00:10:50.488 --> 00:10:53.538
Who knows what side of the bed Donald Trump will, will wake up on.
00:10:53.817 --> 00:10:56.927
And so it's really interesting to me, Mike, thanks for unpacking that.
00:10:56.927 --> 00:11:03.607
It's really interesting to me that all of the work done by TikTok to essentially kind of separate it out as a, as a.
00:11:04.187 --> 00:11:12.341
distinct legal entity, under this kind of project, Texas, umbrella that it, that it came up with a few years back actually has seemingly not worked at all.
00:11:12.575 --> 00:11:19.361
it has not proven to be persuasive to, in this case at all as to the national security, threat still.
00:11:19.918 --> 00:11:25.552
Why do you think that is, why do you think that the, the kind of, All of that efforts has been, has been kind of fallen on deaf ears,
00:11:25.971 --> 00:11:51.774
I mean, there is some element of it, which is just, honestly, it feels like general fear of China and that comes through in the ruling to like you see repeatedly, just talk of the PRC, the PRC, this, the PRC, that, and even if they say they're not going to, if the PRC comes down and demands that they do this or that, they will have to obey and therefore, that is the overriding concern.
00:11:52.113 --> 00:12:05.482
And, you know, there were stories that came out in the press from like former TikTok employees saying that, employees in China still had access to data and that that's noted in the decision.
00:12:05.482 --> 00:12:06.503
They, they point that out.
00:12:06.503 --> 00:12:10.222
And so they say, you know, we just, basically they just don't trust that this is real.
00:12:10.232 --> 00:12:10.893
They, it's sort of.
00:12:11.448 --> 00:12:25.865
the judges seem to feel that this was kind of a fictional setup, um, and they don't really trust that Oracle's ability to audit it as actually meaningful because again, they, they feel that, the PRC could come down and, do all of this.
00:12:26.275 --> 00:12:31.697
You know, the other argument that is made too, is that, It's really just focused on the PRC.
00:12:31.748 --> 00:12:45.494
I mean, because they say, you know, one of the reasons why they claim that this isn't a first amendment violation is because if the company does divest, they say all of the same content and all of the same moderation could still occur.
00:12:45.683 --> 00:12:48.557
There would be no change to the, expressive nature of things.
00:12:48.557 --> 00:12:52.280
The only thing we are doing is trying to disconnect the PRC from that.
00:12:52.620 --> 00:12:53.821
From tick tock.
00:12:54.191 --> 00:13:03.803
And so they feel that, the project Texas set up and the Oracle audits and the U S government like off button don't separate the company from the PRC.
00:13:03.854 --> 00:13:05.683
And so that's where the court comes down.
00:13:05.994 --> 00:13:12.677
They'd only be happy if, the PRC sold by dance or by dance, changed owners or changed hands.
00:13:12.677 --> 00:13:16.732
That's, that's the only way this would actually kind of, uh, change the, change the outcome in, in many sense.
00:13:17.197 --> 00:13:17.498
Yeah.
00:13:17.498 --> 00:13:21.398
And, and, you know, I think for a variety of reasons, I think, I think it's wrong.
00:13:21.437 --> 00:13:25.878
I think it's, there are other precedents on this that they sort of poo poo.
00:13:25.878 --> 00:13:28.388
And we're just like, the hat doesn't, doesn't really apply here.
00:13:28.725 --> 00:13:30.163
but it is something that happens.
00:13:30.163 --> 00:13:31.653
I think, I think it's a, bad look.
00:13:31.653 --> 00:13:37.196
I think it's a bad look for, you know, a supposedly free country in the U S to sort of take this viewpoint.
00:13:37.265 --> 00:13:44.778
And I think it will justify all sorts of other bad stuff from elsewhere as well, based on things like this, uh, it's, I think it's a bad look.
00:13:45.019 --> 00:13:46.739
I know some people really.
00:13:46.864 --> 00:14:09.956
Are concerned and there may be legitimate reasons to be concerned about the Chinese government and their connections here But I just feel that this law goes against, Basic american principles on on things and and i'm disappointed by by the ruling which you know again I've only read it once and really focused on the first amendment section, which is sort of the second half of the ruling And, um, I just, uh, it strikes me as very unconvincing.
00:14:10.291 --> 00:14:10.650
yeah.
00:14:10.780 --> 00:14:17.390
And that national security concern has yet to be proven as real or anything beyond really hypothetical at this stage.
00:14:17.725 --> 00:14:18.046
Yep.
00:14:18.086 --> 00:14:18.395
Yep.
00:14:18.426 --> 00:14:20.212
And, you know, it, it, that frustrates me.
00:14:20.212 --> 00:14:35.302
I mean, I have this come up in all sorts of, other cases around like Fourth Amendment stuff and encryption and just, you know, the willingness of the courts to accept the government just sort of giving this blanket statement that this is a national security concern without proof is, is just kind of frustrating.
00:14:35.648 --> 00:14:35.988
Yeah.
00:14:36.048 --> 00:14:36.368
Okay.
00:14:36.778 --> 00:14:48.758
it links, um, nicely to, to our next story, Mike, which you have, Figured out, um, it's, it's a TikTok story, but it's about a country that we don't often talk about in controlled speech, um, Romania.
00:14:49.038 --> 00:14:57.140
and I guess this is kind of really what the, in some senses, the threat that the, courts, and the government in the U S are trying to kind of mitigate against.
00:14:57.140 --> 00:14:57.390
Right.
00:14:57.666 --> 00:14:58.306
Yeah.
00:14:58.505 --> 00:14:58.855
Yeah.
00:14:58.855 --> 00:15:01.936
It's, it's kind of an interesting, interestingly related.
00:15:02.005 --> 00:15:24.128
Um, there's an election going on this coming Sunday, in Romania, but in the first round of the election, that's sort of a, you know, a two round election, the first round to figure out who are the top two candidates and then the, final election is you know, with just those two top two candidates, there was a surprise, you know, second place finish that knocked out the incumbent.
00:15:24.138 --> 00:15:28.927
The incumbent, I believe, came in third and so is not taking part in the election this weekend.
00:15:29.427 --> 00:15:38.918
there was a candidate, and I may pronounce his name wrong, but Kalin Georgescu, who was considered a sort of mostly unknown far right.
00:15:39.138 --> 00:15:44.508
populist candidate who sort of came out of nowhere with a big TikTok following.
00:15:44.927 --> 00:15:50.217
Um, and he has been sort of very pro Russia, pro Putin.
00:15:50.227 --> 00:15:56.368
And in fact, a lot of his successful TikToks are sort of reminiscent of, Putin.
00:15:57.097 --> 00:16:06.490
Style, uh, press appearances, riding a horse, doing judo, apparently running on a track without breaking a sweat.
00:16:06.791 --> 00:16:09.140
It's just sort of this like physical prowess.
00:16:09.591 --> 00:16:11.980
nothing, nothing to, you know, who doesn't like that?
00:16:11.980 --> 00:16:14.441
That sounds like a great, a great TikTok experience.
00:16:14.701 --> 00:16:15.140
yes.
00:16:15.181 --> 00:16:15.551
Yeah.
00:16:15.831 --> 00:16:18.754
And so he built up a really big following on, Tik TOK.
00:16:18.754 --> 00:16:23.313
The polls in the country had suggested that he was, not going to get that many votes.
00:16:23.677 --> 00:16:26.717
and a lot of people, it appears sort of chalked up his.
00:16:27.008 --> 00:16:30.878
Big Tik TOK following to foreign influence campaigns.
00:16:30.878 --> 00:16:34.898
They also noted that those videos were like highly stylized and produced.
00:16:35.197 --> 00:16:40.638
And there were questions of who was doing some of the producing and the stylizing behind it.
00:16:41.168 --> 00:16:47.004
There were questions about, whether or not the followers were real, were they foreign efforts?
00:16:47.004 --> 00:16:47.864
Were they bots?
00:16:48.227 --> 00:16:50.148
there were questions around.
00:16:50.523 --> 00:16:55.133
Influencers who were promoting the videos and were they paid or not?
00:16:55.712 --> 00:17:13.624
I, I, I found this really interesting point around, like, whether this candidate was, marked in the same way as other political candidates, basically suggesting that, they weren't kind of tagged as a political candidate running an election and therefore gotten larger reach than the other candidates, which is a really interesting kind of element to this, isn't there?
00:17:13.953 --> 00:17:14.463
Yeah.
00:17:14.663 --> 00:17:17.173
And that's unclear because tick tock denies that.
00:17:17.449 --> 00:17:30.144
but deal is because there are restrictions on political advertising within Romania, If all the candidates are treated as candidates, then in theory, there were limitations on kind of what sorts of promotions they could do.
00:17:30.144 --> 00:17:34.444
And if they were paying influencers to promote, is that a political ad?
00:17:34.500 --> 00:17:35.781
and how does that apply?
00:17:36.061 --> 00:17:51.817
And so there are all these questions and the suggestion from many, including some government officials is that TikTok didn't designate him as a, politician, as a candidate, and therefore didn't have these restrictions and that allowed his videos to go much further and build up a bigger following.
00:17:52.376 --> 00:17:53.686
Again, TikTok denies it.
00:17:53.686 --> 00:17:56.356
So it's a little unclear whether or not that is true.
00:17:56.527 --> 00:17:59.535
There is though, that, the things that TikTok.
00:17:59.815 --> 00:18:08.271
Did do it was at one point they did remove what they refer to as a covert network that was trying to boost his videos.
00:18:08.461 --> 00:18:10.434
And so, they did.
00:18:10.711 --> 00:18:20.241
Something, you know, they certainly, paid attention to something and found some inauthentic behavior behind him and that they claim that they stopped.
00:18:20.241 --> 00:18:25.932
And they also said that they found some similar types of inauthentic behavior for some of the other candidates.
00:18:26.311 --> 00:18:28.662
Um, so it wasn't that they did nothing.
00:18:29.008 --> 00:18:37.739
they were willing to step up and do something, but Romanian officials, feel pretty sketched out by this and they've asked the EU to investigate.
00:18:38.204 --> 00:18:53.174
and you know, it, it brings up all the questions that happen when, whenever there are unexpected results in elections and everyone's looking for explanations and, and these days, oftentimes they quickly jump to, well, it was this internet platform that caused the problem.
00:18:53.174 --> 00:18:53.394
Yeah.
00:18:53.525 --> 00:18:53.924
Yeah.
00:18:53.974 --> 00:19:02.825
I mean, there's a couple of interesting points here is in some ways this story is, one we've seen many times before, which is platform being used by political candidate.
00:19:03.138 --> 00:19:06.259
with the kind of looming threat of foreign interference in the background.
00:19:06.548 --> 00:19:11.848
And, you know, we've there's countless examples of that, Analytica being the kind of most, most famous one, I guess.
00:19:12.249 --> 00:19:16.420
And, It's nice to see that things don't change in the best part of a decade.
00:19:17.400 --> 00:19:22.926
But I was really interested in, you know, if that is the case, if there is a kind of lack of oversight here, Romania is pretty small.
00:19:23.017 --> 00:19:25.134
It has 20 million citizens.
00:19:25.564 --> 00:19:29.223
8 million of those apparently use TikTok or have a kind of TikTok account.
00:19:29.223 --> 00:19:31.186
So it's a significant number.
00:19:31.406 --> 00:19:39.194
You know, why was it that there wasn't really, I guess, more attention paid to Making sure that candidates were, given the same kind of platform.
00:19:39.865 --> 00:19:48.734
And I decided to kind of go into the DSA data, my find out how many, how many moderators Romania has, um, or how many are Romanian speaking.
00:19:49.144 --> 00:19:58.845
And, and so there are, according to the DSA report from January to June, 95 Romanian speaking moderators for a country of 20 million.
00:19:59.459 --> 00:20:01.308
And we don't know if that's good or bad.
00:20:01.308 --> 00:20:04.788
It doesn't necessarily take into account, I guess, people setting policy either.
00:20:04.788 --> 00:20:08.845
That's probably just the folks who are, looking at reports and appeals and those kinds of things.
00:20:09.224 --> 00:20:11.335
But it doesn't seem a lot compared to other countries.
00:20:11.481 --> 00:20:15.128
so the Netherlands, which has roughly the same number of.
00:20:15.294 --> 00:20:22.693
citizens, we don't know exactly how many users, of TikTok, but they have 160 moderators, speaking Dutch.
00:20:22.693 --> 00:20:23.713
roughly double,
00:20:23.903 --> 00:20:30.940
So roughly double, um, Sweden has, roughly the same number of moderators, 99, but has half the population.
00:20:31.329 --> 00:20:35.210
And again, we don't, we don't know how many, Of those people use, use Tik TOK.
00:20:35.619 --> 00:20:41.289
So again, you know, it's interesting to me that kind of how resources are applied.
00:20:41.660 --> 00:20:55.650
And again, this is a tale as old as time, how platforms, in the kind of non core markets, set policy and operationalized policy is something that we, we see time and time again as, as a sticking point, do you think that's kind of relevant here?
00:20:55.650 --> 00:20:58.913
Does that feel like a, a pathway trod a little bit?
00:20:59.263 --> 00:20:59.544
Yeah.
00:20:59.544 --> 00:21:24.611
I mean, you know, it is one of these, big questions that comes up all the time, so many of these discussions, you know, and we try to, you know, one of the reasons you and I are always looking for stories outside of, the U S and, The big countries that everybody talks about is because there is an important story there about how these companies handle it and the, you know, less followed countries where, where less attention is, paid to them.
00:21:25.080 --> 00:21:28.000
And so I think, I think it is a, it's a huge story.
00:21:28.330 --> 00:21:32.540
And, you know, there is this element of this one where I sort of feel like, I think.
00:21:32.885 --> 00:21:41.685
A lot of people really believe that so much of his following, in this case, were bots, that they didn't think that the votes would follow, and yet the votes did.
00:21:41.695 --> 00:21:49.732
And so, there was this part of me when I first story where I was like, and, you know, the first thing I was reading was like, oh, you know, it's all fake followers and bots.
00:21:49.732 --> 00:21:52.383
I was like, yeah, but it wasn't fake voters, you know.
00:21:52.772 --> 00:21:55.022
So, so something is happening here.
00:21:55.490 --> 00:22:03.914
but to be honest too, like, I don't know that, you can say necessarily that like, Oh, it's 95 or whatever moderators too little.
00:22:04.397 --> 00:22:05.198
I don't know.
00:22:05.258 --> 00:22:15.758
And we, we don't know, like, was that not enough where the policy's not in place again, there's the whole thing where like TikTok has denied this and they did take down some inauthentic behavior.
00:22:16.065 --> 00:22:19.766
so it, we, we have, you know, Partial information, not full information.
00:22:19.955 --> 00:22:30.400
If the EU does an investigation, it would be really interesting to find out if more details come out of it, but it is, it is sort of an interesting story pay attention to and to see what comes of this.
00:22:30.795 --> 00:22:31.464
yeah, definitely.
00:22:31.464 --> 00:22:36.365
And I mean, I think the other part of this story, something that kind of is a thread that runs through U.
00:22:36.365 --> 00:22:36.555
S.
00:22:36.565 --> 00:22:40.285
TikTok ban as well, is just the, is a media story, really.
00:22:40.285 --> 00:22:48.911
It's about the fact that there is so many people kind of consuming TikTok and the shift away from traditional media to, new forms of media.
00:22:48.911 --> 00:22:52.121
The fact that politicians can bypass traditional forms of media.
00:22:52.122 --> 00:22:52.296
Yeah.
00:22:52.506 --> 00:22:56.930
as this candidate has done and still perform very well in an early round.
00:22:56.930 --> 00:22:58.509
So we'll keep a tabs.
00:22:58.690 --> 00:23:01.980
I've never said this before, Mike, but I'll be keeping tabs on the Romanian election this Sunday.
00:23:02.059 --> 00:23:02.559
Uh,
00:23:03.744 --> 00:23:04.315
There you go.
00:23:04.545 --> 00:23:06.714
This, this, this coming Sunday is my birthday.
00:23:06.714 --> 00:23:10.964
So I will celebrate it by paying attention to the Romanian election as well.
00:23:11.349 --> 00:23:12.470
happy birthday in advance.
00:23:12.470 --> 00:23:13.029
And, uh,
00:23:13.184 --> 00:23:13.325
you.
00:23:13.559 --> 00:23:14.759
What, what a way to celebrate
00:23:15.924 --> 00:23:16.625
Yes, yes.
00:23:16.989 --> 00:23:25.890
so if those kind of two stories paired together, I guess, represent in some ways the kind of ghost of, antitrust past Mike, you know, the way that the,
00:23:26.144 --> 00:23:26.835
a segue.
00:23:26.865 --> 00:23:27.954
What a segue.
00:23:28.349 --> 00:23:31.813
you know, I think the next story is the ghost of, of antitrust future.
00:23:32.240 --> 00:23:38.101
next one you picked out and it's about a man that I actually didn't know about, uh, up until, an hour or so ago.
00:23:38.631 --> 00:23:41.201
Um, but he's an incredibly frightening looking man.
00:23:41.201 --> 00:23:44.250
I'll have to say, and I'm not sure if I'm ever going to forget his face.
00:23:45.161 --> 00:23:49.270
Um, so tell us about Andrew Ferguson and, and what he's come out with this week.
00:23:49.441 --> 00:23:50.111
Oh my goodness.
00:23:50.121 --> 00:23:52.590
So Andrew Ferguson is an FTC commissioner.
00:23:52.810 --> 00:24:04.184
and as a Republican, the FTC has five commissioners, three are appointed by whichever party has the presidency and two are the, other party.
00:24:04.204 --> 00:24:06.880
So right now there are two Republican commissioners.
00:24:06.932 --> 00:24:11.133
FTC commissioners and three Democratic ones, that will flip in January.
00:24:11.782 --> 00:24:15.715
And, Andrew Ferguson is one of the Republican commissioners.
00:24:16.006 --> 00:24:25.079
He is vying for very clearly vying for the chair, taking over what Lena Khan's position is, is right now under a Trump administration.
00:24:25.675 --> 00:24:37.403
and there was a New York Post article Today or yesterday that sort of detailed, there are three leading candidates, the two current FTC commissioners, which is Ferguson and the other one's Melissa Holyoke.
00:24:37.759 --> 00:24:44.220
and then there's, there's a third person, who, worked for Senator Mike Lee that he's really pushing for to be chair.
00:24:44.539 --> 00:24:53.200
And the, the question that people around Trump are apparently asking is like, which one of these is going to be toughest on, they say, big tech.
00:24:53.441 --> 00:24:54.950
but they really mean Trump's enemies.
00:24:55.180 --> 00:25:03.029
Especially considering how much support he got from certain tech sectors this time around, it'll be, you know, the companies that they don't like.
00:25:03.569 --> 00:25:10.619
And so the FTC took a fairly typical FTC action this week against an e commerce platform called Goat.
00:25:11.224 --> 00:25:16.795
The details aren't even that interesting in the case, but basically they lied about shipping times.
00:25:16.835 --> 00:25:19.884
People were paying for premium shipping and not getting it in time.
00:25:20.424 --> 00:25:25.508
And then also they had a, buyer protection, thing where they're saying like, if anything goes wrong, we'll protect you.
00:25:25.508 --> 00:25:27.327
And then they weren't, they weren't living up to that.
00:25:28.028 --> 00:25:32.795
The FTC took action on them and basically saying, you know, they were making promises and they weren't living up to it.
00:25:32.795 --> 00:25:34.644
That's an unfair and deceptive practice.
00:25:35.015 --> 00:25:37.194
Very typical standard FTC.
00:25:37.829 --> 00:25:39.680
Nothing at all interesting about that.
00:25:40.202 --> 00:25:56.915
The other Republican commissioner, Holyoke, put out a concurring statement on this that basically just said, this also proves that, we can use the same authorities to go after big tech companies for unfair moderation decisions.
00:25:57.411 --> 00:25:57.750
Okay.
00:25:58.226 --> 00:25:59.266
Which is nonsense.
00:26:00.191 --> 00:26:00.550
Yeah.
00:26:00.962 --> 00:26:12.455
but it was like a one paragraph thing and it looks like to me, at least Andrew Ferguson then said, okay, I see that I got a one up it because we're in, we're in a fight here for the chair position.
00:26:12.611 --> 00:26:13.411
I raise you.
00:26:13.682 --> 00:26:16.199
I raised you, I raised you crazy.
00:26:16.262 --> 00:26:19.208
and put out this like four page concurring thing.
00:26:19.278 --> 00:26:26.214
Again, none of this has anything to do with the ruling on goat, which is the company that this is ostensibly about, saying like, yes.
00:26:26.234 --> 00:26:28.164
And I agree with Holyoke that we can.
00:26:28.412 --> 00:26:44.762
Use this power to go after unfair moderation, but also we, we can use our powers to go after advertisers who stopped advertising on X because that must be antitrust collusion to censor free speech.
00:26:44.883 --> 00:26:46.772
And we have to support free speech.
00:26:46.903 --> 00:26:54.103
And right now there's only one free speech platform Elon Musk, the brilliant, wonderful free speech supporter.
00:26:54.482 --> 00:27:03.347
And, you know, How dare, any other platform censor American speech and that must be illegal and how dare they not advertise.
00:27:03.718 --> 00:27:13.046
so he goes after, advertisers who stopped, he goes after Garm, which we've talked about in the past saying, you know, that was, clear evidence of, of collusion.
00:27:13.276 --> 00:27:25.395
He goes after NewsGuard, saying that, NewsGuard, who I've written about a few times that the Republicans have gone crazy about, NewsGuard, all they do is just say which news sources are trustworthy and which are not.
00:27:25.776 --> 00:27:30.913
And he sort of admits when he's talking about NewsGuard, like, yes, NewsGuard can have its own opinions.
00:27:31.282 --> 00:27:37.903
But if multiple companies are basing decisions on those opinions, that is antitrust collusion.
00:27:38.303 --> 00:27:51.039
It, it is four pages of crazy indicating once again, like with Brendan Carr, we talked about a few weeks ago, like with Brendan Carr that, these.
00:27:51.242 --> 00:28:01.057
Bureaucrats really intend to use the powers of government to attack speech online, and they're framing it all within the language of free speech.
00:28:01.387 --> 00:28:05.351
The whole thing over and over again, he talks about, how important free speech is.
00:28:05.371 --> 00:28:08.740
And he does the Elon Musk is the only believer in free speech.
00:28:08.951 --> 00:28:13.964
And every platform has to use the same policies as, Elon does.
00:28:14.174 --> 00:28:15.625
And how dare they not do that?
00:28:15.867 --> 00:28:32.414
so just to clarify, so we have a situation where we're antitrust as we've talked about in the last couple of weeks of the podcast which is going to be a big theme within the Trump administration, and, you know, he's run out of the FTC has a guy in Brendan Carr who doesn't know much.
00:28:33.420 --> 00:28:35.200
Brendan Carr's the FCC, not the
00:28:35.464 --> 00:28:36.394
Sorry, sorry.
00:28:36.394 --> 00:28:43.488
Yeah, so, so the FTC commissioners also don't know very much according to this letter, at least this particular guy.
00:28:43.698 --> 00:28:48.290
And so we have a situation where like maybe nobody knows anything about
00:28:49.346 --> 00:28:55.326
it's, I mean, the question is, do they know or are they just like putting on a show for Trump?
00:28:55.355 --> 00:28:55.655
Right.
00:28:55.685 --> 00:28:58.415
And, and it's just this sort of like populist thing.
00:28:58.806 --> 00:29:02.205
I don't know Andrew Ferguson that well, Brendan Carr, I know a little bit.
00:29:02.205 --> 00:29:04.115
And so I know he knows that he's.
00:29:04.570 --> 00:29:08.201
Lying, like Brendan Carr is smart enough to know what the law is.
00:29:08.607 --> 00:29:12.345
I don't know enough about Ferguson to know whether or not he knows this is crazy.
00:29:12.845 --> 00:29:32.381
You know, one of the lines that really got me in this letter was like, he claims at one point that the proof of collusion among big tech companies to censor content in an illegal manner is that simultaneously, he specifically says simultaneously, all of the big tech platforms blocked.
00:29:32.615 --> 00:29:39.006
All discussion and reporting on the Hunter Biden laptop in 2020.
00:29:39.340 --> 00:29:40.790
Which just isn't the case.
00:29:41.175 --> 00:29:42.996
None of that happened, right?
00:29:43.006 --> 00:29:48.500
The only thing that happened was two companies took some action, Twitter and Facebook.
00:29:49.046 --> 00:29:51.705
The action that Twitter took was it blocked the link.
00:29:51.786 --> 00:29:53.996
It did not block any other reporting on it.
00:29:54.006 --> 00:29:55.445
There was other reporting on it.
00:29:55.586 --> 00:29:57.526
It did not block any discussion of it.
00:29:57.526 --> 00:29:59.256
There was lots of other discussion on it.
00:29:59.256 --> 00:30:01.705
In fact, like it was like a trending topic.
00:30:02.185 --> 00:30:04.175
The only thing they did was they blocked the link.
00:30:04.316 --> 00:30:08.355
To a single New York post story for 24 hours.
00:30:08.395 --> 00:30:12.026
Then they reversed their policy and allowed that link to be shared.
00:30:12.615 --> 00:30:18.435
The only thing that Facebook did was it said, well, there's some questions about the story, so we're going to keep it out of the trending.
00:30:18.586 --> 00:30:20.685
It won't go into the trending topics.
00:30:21.115 --> 00:30:23.486
and they reversed that policy relatively quickly.
00:30:23.945 --> 00:30:28.195
That was the only thing it did, but he declares unequivocally that the entire.
00:30:28.421 --> 00:30:32.601
Tech industry simultaneously blocked all discussion of this.
00:30:33.174 --> 00:30:51.637
And I, you know, one of the things that gets me is I pointed out earlier this year that Elon Musk did all that and more when apparently Iranian hackers got access to the Trump campaigns dossier on JD Vance and they passed around and most, most media sources didn't bite on it.
00:30:51.647 --> 00:30:59.557
Finally, Ken Klippenstein, who has a substack posted it and Elon banned Ken.
00:30:59.788 --> 00:31:01.087
He blocked all.
00:31:01.512 --> 00:31:06.309
Links to, any part of Ken's sub stack, not just that one article.
00:31:06.700 --> 00:31:08.730
He, you know, pulled down all sorts of stuff.
00:31:08.730 --> 00:31:11.049
And to this day, I don't think you can share that.
00:31:11.059 --> 00:31:22.826
He did let Ken back on the platform after like two weeks, So everything that, they have accused Twitter of doing to the Hunter Biden laptop story, Elon has done and more and gone much further.
00:31:23.076 --> 00:31:24.195
And yet in this.
00:31:24.477 --> 00:31:39.397
comment from this FTC commissioner, he claims that Elon is the big free speech supporter and the actions taken on the, the Hunter laptop, which didn't happen, prove that they're illegal censoring, collusion, antitrust.
00:31:39.788 --> 00:31:41.798
Not everything about this.
00:31:41.887 --> 00:31:43.387
You've got your hands on your head, Mike.
00:31:43.387 --> 00:31:44.667
It's
00:31:45.107 --> 00:31:49.196
it's so wrong, but this is, this is unfortunately the world that we're living in.
00:31:49.196 --> 00:31:56.170
and it, gives a sense of how the incoming administration is going to attack content moderation.
00:31:56.259 --> 00:31:57.779
They're going to make these claims.
00:31:57.960 --> 00:32:04.869
They are going to try and use every legal lever they have, even as they are crazy and totally counterfactual to reality.
00:32:05.119 --> 00:32:05.369
Yeah.
00:32:05.680 --> 00:32:15.170
There's no I mean, the idea that FTC could prove that platforms coordinated on policy changes or anything like that would be so difficult to do, right?
00:32:15.170 --> 00:32:26.463
If this actually was, you know, is a route you want to go down, how do you go about saying that this company over here would Has done the same thing as this company over here a way that amounts to collusion.
00:32:26.747 --> 00:32:42.990
Yeah, well, the thing that they can do, and they probably will do, is that they can conduct investigations, and they can demand to see all sorts of internal files, and that is what is going to happen, almost certainly, and then, I would guess what would come out of it is, probably a, really misleading investigation.
00:32:43.335 --> 00:32:50.419
findings and they'll release things selectively that take things out of context and make, make claims that are just not accurate.
00:32:50.772 --> 00:32:52.449
and it's going to be a mess.
00:32:52.582 --> 00:33:04.634
and this is why we're seeing, tech companies, you know, trying to kiss the ring of Donald Trump and, and try and make nice because they know if they don't, they're going to face all of this kind of, authoritarian nonsense.
00:33:04.839 --> 00:33:06.099
yeah, no, indeed.
00:33:06.460 --> 00:33:13.047
And, and, you know, currying favor, it's actually, uh, again, a very good segue onto our next story, um, is, is fast becoming
00:33:13.051 --> 00:33:14.231
did it on purpose, Ben.
00:33:15.666 --> 00:33:16.794
The theme of this, episode.
00:33:16.844 --> 00:33:23.676
So, you know, we have a situation where, FTC commissioners are, are, cozying up to, the new administration.
00:33:23.777 --> 00:33:29.467
We also have a situation where Meta in a very coordinated way is doing the same thing.
00:33:29.676 --> 00:33:47.487
and so this week, a number of different outlets, including the financial times and the verge reported on, Nick Clegg, the, president of global affairs at Metta, talking about how, Metta essentially overstepped the line when it came to content moderation during COVID, the comments were made in a, reporter.
00:33:48.092 --> 00:33:58.902
Briefing, which is, uh, you know, does happen, but it's a very kind of coordinated, very kind of controlled environment, for a very senior person within Metta to kind of make these statements.
00:33:59.281 --> 00:34:10.454
And, yeah, the, coverage has been essentially that Metta is not apologizing, but admitting that it, it overstepped the line in terms of, COVID information controlled during that period.
00:34:10.849 --> 00:34:11.068
And.
00:34:11.449 --> 00:34:20.298
This comes on the back of, you'll remember Mike, that letter by Mark Zuckerberg, uh, to Jim Jordan, which we had a good laugh about, um,
00:34:21.173 --> 00:34:22.094
I thought it was a cry.
00:34:23.418 --> 00:34:47.704
we cried a bit, um, it, it, it felt like a letter that had been, written at gunpoint, um, I remember, I remember saying, and, uh, almost kind of made, made to write that, and it was clearly in lieu of this situation, right, that where Donald Trump becomes president again, and, you have a situation as was reported this week, where Zuckerberg is invited to Mar a Lago to talk about the future of tech policy, out of the U S.
00:34:47.713 --> 00:34:53.347
So in a number of different ways we have, you know, in here, we have meta kind of cozying up to the new administration.
00:34:53.347 --> 00:34:55.628
We have the FTC commissioners doing the same.
00:34:56.121 --> 00:34:58.780
Is there anybody that has any dignity left?
00:34:59.550 --> 00:35:02.391
You know, I think this stuff is so obvious, right?
00:35:02.391 --> 00:35:14.963
It's, you know, it's so obvious and in some respects, I'm frustrated about the way that it's reported, in this way, because apart from the line that says This is a briefing with Nick Clegg and some journalists who've been invited there.
00:35:15.978 --> 00:35:24.213
quite clearly, you know, designed to be a signal to the Trump administration of, we know what you're going to ask us to do and we're happy to do it.
00:35:24.603 --> 00:35:24.974
Yeah.
00:35:25.233 --> 00:35:25.543
Yeah.
00:35:25.594 --> 00:35:26.554
That's exactly what it was.
00:35:26.614 --> 00:35:28.403
this was totally a messaging thing.
00:35:28.733 --> 00:35:35.083
It was, you know, a coordinated attempt by the company to lay out this message that will be embraced by.
00:35:35.391 --> 00:35:40.418
the sort of MAGA faithful, to insist that it proves, I mean, this will be extended, right?
00:35:40.418 --> 00:35:54.521
They'll say it proves that, not only that MEDA was overly, willing to suppress speech, but it will just be reinforced with the claim that, the demands for that came from the government, which is the part that he didn't say that.
00:35:54.764 --> 00:35:57.181
But that is, claim that lots of people are making.
00:35:57.664 --> 00:36:08.840
and so, This is, it's a spineless capitulation to this argument and basically it's giving the Republicans ammo to claim that we were right all along.
00:36:09.061 --> 00:36:10.570
We were unfairly targeted.
00:36:10.721 --> 00:36:14.744
We were unfairly censored and, even meta admits it.
00:36:14.900 --> 00:36:16.541
and we'll see that over and over again.
00:36:16.561 --> 00:36:18.530
And people will point to this as if it's proof.
00:36:18.791 --> 00:36:21.021
I had somebody yell at me this week,
00:36:21.577 --> 00:36:22.277
only one.
00:36:22.557 --> 00:36:29.168
well, there were a few people, but someone was yelling at me about this, where I was talking about some of this and they were saying, well, you know.
00:36:29.518 --> 00:36:34.458
Zuckerberg admitted under oath that the US government pressured him to take down content.
00:36:34.487 --> 00:36:37.402
He didn't want to take which is not What actually happened,
00:36:37.507 --> 00:36:37.907
Right.
00:36:38.452 --> 00:36:43.882
but like the message gets out there and kind of knows what they're doing when they, when they say this.
00:36:44.282 --> 00:36:46.422
and you know, I understand why they're doing it.
00:36:46.443 --> 00:36:55.353
You know, they feel like they need to do it to avoid, to hopefully avoid costly stuff, but it is, shows a real lack of principles as far as I'm concerned.
00:36:55.867 --> 00:36:56.197
Yeah.
00:36:56.197 --> 00:37:14.523
And if you're a trust and safety professional working in meta who, probably spent countless hours trying to figure out what the policy should be during an evolving situation that no one has ever seen before that was COVID and no one will ever see, you know, for a long time, that's going to be really, really tough to take.
00:37:14.748 --> 00:37:16.057
it's demoralizing, right?
00:37:16.057 --> 00:37:23.938
I mean, you know, the reality is what these companies should be doing and they don't do is saying, like, these are really, really, really difficult decisions.
00:37:24.108 --> 00:37:25.728
And there was no way to get it right.
00:37:25.797 --> 00:37:27.038
There was simply no way.
00:37:27.038 --> 00:37:30.018
I mean, I talk about sort of impossibility theorem here.
00:37:30.208 --> 00:37:31.487
There is no way to get it right.
00:37:31.677 --> 00:37:32.273
And that is it.
00:37:32.483 --> 00:37:35.103
Extra true in a case where you have something that is brand new.
00:37:35.112 --> 00:37:38.393
Nobody understands, you know, nobody understood the details of COVID.
00:37:38.592 --> 00:37:39.913
We didn't know what was right.
00:37:39.913 --> 00:37:41.123
We didn't know what was wrong.
00:37:41.393 --> 00:37:44.452
And people made choices and lots of people made wrong choices.
00:37:44.672 --> 00:37:49.822
Some of them made wrong choices because they just didn't have enough information and there's more information came out, they adjusted.
00:37:50.043 --> 00:37:54.079
Some people made wrong choices because they, you had crazy ideas in their head.
00:37:54.298 --> 00:37:57.148
There were all sorts of wrong choices that were made along the way.
00:37:57.268 --> 00:38:01.889
And it wasn't because of like any, you know, in many cases.
00:38:02.443 --> 00:38:04.914
It wasn't because of bad actions or bad ideas.
00:38:05.248 --> 00:38:06.748
I think, you know, the companies.
00:38:07.047 --> 00:38:08.768
Try to put forth their best effort.
00:38:08.867 --> 00:38:10.547
That's what Meta should be saying.
00:38:10.797 --> 00:38:16.101
You know, yes, we may have made mistakes, but we made best efforts based on what kind of information we had.
00:38:16.271 --> 00:38:17.240
We took this seriously.
00:38:17.240 --> 00:38:18.681
We wanted to keep people safe.
00:38:18.951 --> 00:38:24.291
And because of the changing nature of the information environment, we had to make decisions on the fly.
00:38:24.641 --> 00:38:30.831
And instead he comes out and he gives this statement, which is basically like, Oh, you know, we took down too much content because there was too much pressure on us.
00:38:30.831 --> 00:38:33.380
Like, you know, come on, stand up, have a spine.
00:38:33.411 --> 00:38:33.960
I, it's.
00:38:34.891 --> 00:38:36.721
It's really, really frustrating to me.
00:38:36.940 --> 00:38:37.690
This was a chance.
00:38:37.690 --> 00:38:44.001
This was an opportunity for them to educate people on how trust and safety works and what the real purpose of trust and safety is.
00:38:44.311 --> 00:38:48.791
And instead he's feeding into the narrative that it's this awful censorship machine.
00:38:48.880 --> 00:38:51.231
And it's, it's really, really frustrating.
00:38:51.411 --> 00:38:51.701
Yeah.
00:38:51.701 --> 00:38:56.764
I mean, I've talked to a few folks recently about trust and safety is kind of marketing problem.
00:38:56.994 --> 00:39:12.880
The fact that, you know, it needs to kind of present itself continually as that difficult challenge, those impossible trade offs, and it needs to kind of completely an ongoing message that out, um, it doesn't help when Nick comes out and gives a, a soft briefing to those journalists that the opposite is true.
00:39:13.271 --> 00:39:28.713
You know, he does say, interestingly, just before we move on about how, AI has all this potential, but right now there are some really pissed off people on, on Metas platforms for the fact that, they make mistakes and remove innocuous or innocent content.
00:39:28.713 --> 00:39:30.074
And, again, you know, just.
00:39:30.541 --> 00:39:33.612
it's kind of completely crazy because it was literally two weeks ago.
00:39:33.612 --> 00:39:42.099
We were talking about how, how threads was an absolute mess of a, of a moderation process and you had all of these kind of, terms that should have been moderated being taken down.
00:39:42.099 --> 00:39:54.132
So I'd love for media to be a bit more critical and, and, uh, clear on, on what it is that senior people like Clegg are saying, and I think that's, that's really part of our job, um, to, to, to do that too.
00:39:54.498 --> 00:40:01.818
they, they, you know, put this in context, like put, put his statements in context and I didn't feel like, like the media was really doing that.
00:40:02.072 --> 00:40:03.152
no, no, indeed.
00:40:03.501 --> 00:40:06.219
Mike, I'm gonna, um, we can't finish on that low note.
00:40:06.268 --> 00:40:08.429
we've got a couple of lighter stories.
00:40:08.509 --> 00:40:16.949
Um, but I'm gonna, throw to you to pick, you had, end with ChatGPT since that's where we started episode.
00:40:17.338 --> 00:40:20.188
Um, tell us about this kind of fun lighthearted story.
00:40:20.603 --> 00:40:20.804
Yeah.
00:40:20.804 --> 00:40:30.358
This was kind of an interesting story where suddenly it started spreading wide that, chat GPT would break if you tried to get it to say anything The name David Mayer.
00:40:31.085 --> 00:40:39.956
I saw it first in a very funny post on blue sky from Andy Bayo, who's is really interesting guy runs, uh, has the site waxy.
00:40:39.956 --> 00:40:40.266
org.
00:40:40.266 --> 00:40:43.525
And he does lots of really interesting stuff, but he had heard that.
00:40:43.525 --> 00:40:57.885
And so he, he created this question for chat GPT, which was combined the first name of the artist who recorded Ziggy Stardust and the last name of the artist who recorded your body is a wonderland into a single name and chat GPT starts.
00:40:58.065 --> 00:41:01.726
And says the artist who recorded Ziggy Sardust is David Bowie.
00:41:01.916 --> 00:41:05.385
And the artist who recorded Your Body is a Wonderland is John Mayer.
00:41:05.585 --> 00:41:11.266
Combining their names and it says David, and then it breaks and it says, I am unable to produce a response.
00:41:11.976 --> 00:41:14.576
It refuses to produce the name David Mayer.
00:41:15.286 --> 00:41:25.925
And what people, yeah, what people then quickly discovered was that there was a short list of names that chat GPT will break if you try and get it to produce those names.
00:41:26.646 --> 00:41:30.036
Then the hunt was on to sort of figure out who they were and why
00:41:31.596 --> 00:41:31.976
odd.
00:41:32.175 --> 00:41:37.286
So, and did they find out what the cause was that, what, what, why doesn't ChatGPT like these names?
00:41:37.335 --> 00:41:48.489
so as far as anyone can tell, and as I don't think open AI has come out and said anything yet, and I will note that they have fixed the David Mayer one, so that it now does work.
00:41:48.489 --> 00:41:51.998
But the other names on the list do not work, still do not work.
00:41:52.059 --> 00:41:52.449
Okay.
00:41:53.039 --> 00:41:57.219
It appears that OpenAI just created a block list.
00:41:58.309 --> 00:41:59.969
It appears to be about six names.
00:41:59.969 --> 00:42:06.159
There might be more, um, but there's a block list that if you try and get it to produce that name, it will break.
00:42:07.378 --> 00:42:10.782
And it just not, you know, it doesn't break in a nice way.
00:42:10.802 --> 00:42:15.083
It just, you know, we'll go halfway through a question and then say, I am unable to produce a response.
00:42:16.242 --> 00:42:35.126
I really love the idea that there's some incredibly well played, you know, some of the smartest kind of machine learning engineers, you know, sat next to some of the world's, brightest trust and safety experts, some of whom we know probably, um, you know, sat together being like, what are the list of six names that we've gotta add to a block list?
00:42:35.391 --> 00:42:35.760
Yeah.
00:42:36.101 --> 00:42:45.351
And it's just, it's so obviously just a straight up lock list that says we will not produce these names and people have gone through and reasoned out some of them and they're all slightly different.
00:42:45.771 --> 00:42:58.538
There's a few that we don't quite know why there was one person, Guido Scorza, who is a data protection expert in Italy, and he had posted that he had used the GDPRs.
00:42:58.746 --> 00:43:05.260
right to erasure process against open AI and said, basically you have to forget everything about me.
00:43:05.681 --> 00:43:13.420
And the only way that they could figure out to do so was to put him on this block list Because they don't actually remember anything about him.
00:43:13.431 --> 00:43:19.143
Like there's no this is the thing that a lot of people don't Understand about these systems is it's not a big database.
00:43:19.143 --> 00:43:20.784
It's not going in and collecting stuff.
00:43:21.003 --> 00:43:22.673
It is training and just thinking about stuff.
00:43:22.673 --> 00:43:33.190
So the only way to say like, never produce any information on Guido Scorza is to put them on a, I mean, there are better ways, but this was the sort of fast and quick way.
00:43:33.510 --> 00:43:43.333
There are other people, you know, the, Brian Hood was a mayor in Australia who claimed that it produced defamatory content about him and he threatened to sue.
00:43:43.583 --> 00:43:43.943
And so.
00:43:44.318 --> 00:43:46.007
that one probably came up first.
00:43:46.018 --> 00:43:50.608
Somebody had actually told me about that one, like a year and a half ago when it, when it came up and I had meant to investigate it.
00:43:50.608 --> 00:43:51.797
And then I just never got to it.
00:43:52.168 --> 00:43:58.947
Jonathan Turley, who's a law professor and sort of a famously sort of Trumpian law professor.
00:43:59.228 --> 00:44:03.858
Went on this big rant of claiming also that chat GPT had defamed him.
00:44:03.867 --> 00:44:11.398
And so he's on the The weird one is Jonathan Zatrain, who I know, and asked like, what the hell, what
00:44:11.498 --> 00:44:12.047
yeah,
00:44:12.378 --> 00:44:13.327
to get off this list?
00:44:13.557 --> 00:44:15.097
And he has no idea.
00:44:15.467 --> 00:44:16.768
right, right.
00:44:17.010 --> 00:44:18.693
he doesn't know what's going on.
00:44:18.724 --> 00:44:25.264
I joked with him that like, Oh, you know, are you trying to come up with something to prove a point or something about all this?
00:44:25.264 --> 00:44:26.204
He's like, I did nothing.
00:44:28.103 --> 00:44:28.603
So,
00:44:28.813 --> 00:44:31.574
And he didn't know he was on ChachiPT's naughty step.
00:44:31.594 --> 00:44:33.713
he, he had found it, he had actually discovered it.
00:44:33.724 --> 00:44:42.411
So he, he added himself when everyone was pointing out to this list of names, he pointed out, like he had discovered a few months ago that chat GPT breaks on his name.
00:44:42.773 --> 00:44:44.304
and so, you know, it's just.
00:44:44.548 --> 00:44:46.688
Basically, it's like this really simple way.
00:44:46.739 --> 00:44:52.998
I understand probably why it happened because all these people, except for Jonathan, as far as we know, got really mad.
00:44:53.009 --> 00:44:54.898
We still don't know which David Mayer there is.
00:44:54.898 --> 00:44:55.699
There's speculation.
00:44:55.708 --> 00:45:00.068
There's a few different, there's also a David Farber, Faber, and nobody's sure which one that is.
00:45:00.409 --> 00:45:05.865
but the other ones, like people got mad and open AI probably was like, you know what, this guy is a pain.
00:45:06.016 --> 00:45:06.775
Fucking headache.
00:45:07.666 --> 00:45:10.206
And therefore just put them on the ban list.
00:45:10.206 --> 00:45:14.735
Like, let's not deal with this because we don't want to go through a lawsuit and have to do this.
00:45:14.735 --> 00:45:19.295
And there's no other way to like effectively, because they're, you know, they don't understand.
00:45:19.295 --> 00:45:21.759
It's not that chat GPT is defamatory.
00:45:21.759 --> 00:45:24.708
It's whatever the prompt is led to defamation.
00:45:25.039 --> 00:45:31.389
And it's not that we're collecting data on this guy, but you know, he thinks we are, and do we really want to fight it in court?
00:45:31.429 --> 00:45:32.539
I just put them on the ban
00:45:32.753 --> 00:45:33.664
Put him on the list.
00:45:33.793 --> 00:45:36.364
I'm pretty sure that's the kind of thing that happened.
00:45:36.744 --> 00:45:46.617
You know, you would hope that a company with as many resources, and as you noted, as many, smart engineers would come up with a more sophisticated way than this, but they haven't.
00:45:46.668 --> 00:45:53.547
And so, you know, I know how things like that happen and I'm sure it was just kind of like, ah, just make this headache go away in the easiest way possible.
00:45:53.827 --> 00:45:56.760
And that means every time you try and produce this name, we're going to break it.
00:45:57.030 --> 00:45:57.550
Yeah.
00:45:57.601 --> 00:45:58.170
Interesting.
00:45:58.673 --> 00:46:05.951
just as an aside, there's some, there's the OpenAI put out some research in the past week around kind of red teaming in new, new and novel ways.
00:46:06.286 --> 00:46:08.135
I wonder if they should be reading their own research.
00:46:08.971 --> 00:46:12.121
Yeah, yeah, absolutely, absolutely.
00:46:12.465 --> 00:46:23.539
so yeah, if you, if David May is listening, to the podcast, if he's, if he's a regular listener, um, make yourself known, um, we would love to know what you did to be put on the, uh, on the shit list of, uh,
00:46:23.753 --> 00:46:24.153
Yeah.
00:46:24.184 --> 00:46:25.324
And then you got taken off.
00:46:25.324 --> 00:46:28.023
So if you have a legal complaint, get it, get it going
00:46:28.393 --> 00:46:30.474
Yeah, we know some lawyers.
00:46:30.713 --> 00:46:31.903
Um, great.
00:46:31.903 --> 00:46:32.443
Thanks Mike.
00:46:32.614 --> 00:46:36.177
Um, that's not the, you know, that's not the kind of end of today's episode.
00:46:36.177 --> 00:46:37.641
It's, we've got a great bonus chat.
00:46:37.641 --> 00:46:40.110
I don't know if you want to kind of give us a preview of that.
00:46:40.481 --> 00:46:40.780
Yeah.
00:46:40.780 --> 00:46:41.708
So this was really great.
00:46:41.708 --> 00:46:43.438
I had this discussion just yesterday.
00:46:43.514 --> 00:46:45.088
it's a topic that we've talked about a lot.
00:46:45.298 --> 00:46:50.159
The social media age bands in Australia and how they're popping up elsewhere.
00:46:50.338 --> 00:46:58.469
and so this is a discussion I had with, two folks from the internet society, Natalie Campbell, who's the senior director of North American government and regulatory affairs.
00:46:58.509 --> 00:47:02.438
And John Perrino, who's a senior policy and advocacy expert.
00:47:02.806 --> 00:47:08.135
And they're concerned about these age restrictions and age verification.
00:47:08.416 --> 00:47:27.146
And, we talked about Australia and also Canada has a bill that is, moving forward that one of the incredible things in the chat was that, it would require places like Starbucks to verify the age of anyone who wants to use the wifi in Starbucks and, you know, all sorts of stuff.
00:47:27.146 --> 00:47:29.536
And so they, they're sort of concerned about.
00:47:29.847 --> 00:47:36.887
these laws, the proliferation of these laws, what it means for the open internet and what age verification requirements would mean.
00:47:37.117 --> 00:47:40.434
The really fascinating discussion and we'll go to that right now.
00:47:51.025 --> 00:47:52.762
Natalie, John, welcome to the podcast.
00:47:52.802 --> 00:47:53.911
Glad to have you here.
00:47:54.061 --> 00:48:01.422
wanted to start by talking about the law that just passed in Australia, which effectively bans those under 16 from social media.
00:48:01.827 --> 00:48:12.271
John, can you talk through the details of that law, including sort of which sites are impacted and how do you, Does the Australian government expect sites to know the age of their visitors?
00:48:14.041 --> 00:48:30.293
great question, Mike, and, you know, for, for quick background for listeners, this legislation moved through in about less than two weeks and there was a consultation period open less than 24 hours.
00:48:30.804 --> 00:48:35.324
Um, the Australian social media age verification stuff really, really flew through.
00:48:35.786 --> 00:48:48.009
And there's a lot that we still don't know, even though the legislation passed, especially on what social media platforms would be required to do in order to comply with this.
00:48:48.648 --> 00:48:55.719
So right now, essentially all we know is that government ID would not be required.
00:48:56.268 --> 00:49:02.148
That should include things like a passport or government issued driver's license, also a digital ID.
00:49:02.748 --> 00:49:09.820
And we know that the social media is for Australians under the age of 16.
00:49:10.550 --> 00:49:18.501
and then the final thing that was kind of added on late was a digital duty of care, which again is to be determined.
00:49:18.940 --> 00:49:23.451
So the bottom line on this is much of the bill is to be determined.
00:49:23.690 --> 00:49:32.393
They don't know what age verification would be, as they say, reasonable for the social media platforms to use.
00:49:32.873 --> 00:49:47.996
And as a lot of the comments pointed out, there was actually already an age verification study ongoing, and that's still going on, which tools, maybe work best, what the trade offs are in different age verification methods.
00:49:48.237 --> 00:49:50.146
The Australian government was already doing this.
00:49:50.527 --> 00:49:51.856
They don't have the results.
00:49:51.916 --> 00:49:53.396
They probably won't have.
00:49:53.681 --> 00:49:59.512
The results until about six months before the social media platforms have to comply with this.
00:50:00.295 --> 00:50:05.795
I know that we've seen a bunch of other countries also exploring similar ideas around age verification.
00:50:06.101 --> 00:50:17.561
on the podcast, we've talked about it in the UK and somewhat in the US to a lesser extent, there are a couple of different issues in the US, but also now Canada is exploring a similar issue.
00:50:17.661 --> 00:50:22.168
So, Natalie, can you talk through what the proposal in Canada is about?
00:50:22.876 --> 00:50:33.052
Sure, so, Senate Bill S 210, this is an act to restrict young persons online access to sexually explicit material, is very close to becoming a law.
00:50:33.532 --> 00:50:36.802
This bill has been flying under the radar for quite a long time.
00:50:36.822 --> 00:50:41.233
I don't think that people are taking it very seriously because most Senate bills.
00:50:41.628 --> 00:50:42.726
Don't make it to law.
00:50:43.190 --> 00:50:55.175
but this has found its way into the very late stages of our parliamentary process, and it's probably the most dangerous bill to the Internet, in Canada right now.
00:50:55.434 --> 00:51:00.559
and one of the main reasons for that is because of its age verification mandates.
00:51:00.733 --> 00:51:07.474
That would apply to virtually every intermediary on the internet and some that are not on the internet as well.
00:51:08.054 --> 00:51:17.083
So essentially what the bill tries to do is to prevent young people from access to sexually explicit material, which is defined extremely broadly.
00:51:17.409 --> 00:51:31.268
and makes it so that every intermediary that would play a part in facilitating sexually explicit material for profit online would have to verify a user's age or face very high fines.
00:51:31.757 --> 00:51:51.088
So, this, you know, is this, is this, Probably the most extreme age verification proposal we've seen so far, because it's not just targeting websites, it's targeting internet service providers, content delivery networks, search engines, email services, even a Starbucks location that's providing access to Wi Fi.
00:51:51.509 --> 00:51:58.128
Because what content is flowing through this pipes could be sexually explicit material.
00:51:58.545 --> 00:52:05.248
they now have a duty to verify users age and in doing so like it's, it's different.
00:52:05.579 --> 00:52:33.739
When we're talking about websites doing this, but when we're thinking about every single intermediary on the Internet having a duty to verify users age, that gets really problematic when we're thinking about an open Internet first, because most traffic is encrypted, and it's not possible for most infrastructure intermediaries to know what's flowing through the pipes, and even if they could, content doesn't flow through the Internet as a whole piece of content.
00:52:34.108 --> 00:52:35.039
It's packets.
00:52:35.509 --> 00:52:46.329
And so it becomes extremely difficult to, one, identify, what is sexually explicit material and, you know, is Canada's definition of this type of content.
00:52:46.679 --> 00:52:59.846
And then, second of all, having to, decide to try and identify that stuff means you can't use things like encryption, which is the foundation of security for every service and user on the Internet.
00:53:00.585 --> 00:53:13.255
So it's very concerning in that, the implications for security online, but also the fact that you're placing huge barriers to access because now is an Internet user, because a lot of intermediaries.
00:53:13.550 --> 00:53:16.724
Won't know what's sexually explicit material.
00:53:16.733 --> 00:53:40.297
They might just start doing age verification to Everyone and that means I now, you know as somebody who's based in canada would have to trust a whole lot of entities with Very personal information whether it's government id or you know biometrics i'm having to trust a lot of third parties Who I might not have?
00:53:40.297 --> 00:53:48.077
You Any direct relationship with with my personal information, which is a huge barrier to privacy and anonymity online.
00:53:48.621 --> 00:53:55.054
so this is, from an open Internet aspect, this is super problematic because you're creating huge barriers to access.
00:53:55.539 --> 00:53:58.528
for people who might not be able to get government issued I.
00:53:58.528 --> 00:53:58.778
D.
00:53:58.778 --> 00:54:07.048
or might not be able to use the Internet without the promise of anonymity, which is a huge hurdle for a lot of people in marginalized communities and young people as well.
00:54:07.599 --> 00:54:12.108
But also the fact that this has huge implications for security online.
00:54:12.626 --> 00:54:28.398
and you know, not Enabling intermediaries to use encryption that could be extremely devastating to people's security online and making people vulnerable to a whole range of bad stuff that I don't think was the intention of this bill.
00:54:28.768 --> 00:54:31.139
Yeah, I mean you, you mentioned just really quickly.
00:54:31.139 --> 00:54:32.349
I wanted to follow up on one point.
00:54:32.509 --> 00:54:45.846
You mentioned the fact that since it hits every intermediary, the potential that, like a Starbucks, if you wanted internet in a Starbucks, is the idea there that, Starbucks would have to not just check your ID, but like record it somehow?
00:54:45.846 --> 00:54:51.135
Is that part of the fear that if you want to get on the Wi Fi at Starbucks, you have to first prove how old you are?
00:54:51.659 --> 00:55:09.699
Well, this is the problem is that the way that it defines who this bill applies to is internet service provider, but the definition it uses for that is a person who provides access to the Internet Internet content hosting or electronic mail.
00:55:10.588 --> 00:55:24.949
The person who provides internet access could mean Starbucks and could mean that, you Starbucks has to know what are you looking at when you're accessing the internet or just say, just show us your ID if you want to use the Wi Fi, right?
00:55:24.949 --> 00:55:30.096
So, problematic in both senses and just a reminder that please use VPNs.
00:55:30.456 --> 00:55:30.485
Okay.
00:55:32.101 --> 00:55:38.650
So, I mean, so we're seeing this all over, obviously, Australia and Canada, as you guys discussed, we've discussed the UK in the past.
00:55:38.670 --> 00:55:42.423
There are various bills in the U S that also touch on this elsewhere around the world.
00:55:42.804 --> 00:55:44.653
What should we make of this trend?
00:55:44.653 --> 00:55:47.244
What, you know, what, why is this all, all of a sudden happening?
00:55:47.244 --> 00:55:49.673
We've had the internet obviously for decades now.
00:55:49.793 --> 00:55:56.947
Why is it suddenly that everyone is trying to pass these kind of age related bans or age verification.
00:55:57.134 --> 00:55:59.824
ideas, John, do you want to start there?
00:56:00.248 --> 00:56:09.239
Yeah, I mean, to that point, it certainly feels like everything is happening all at once everywhere in every corner of the world.
00:56:09.878 --> 00:56:18.568
I mean, we're seeing this, as you mentioned, you know, the UK has been working on this almost a decade, but their age verification guidance comes out in January.
00:56:19.389 --> 00:56:21.088
The European Union is working on this.
00:56:21.583 --> 00:56:24.563
There are almost two dozen U.
00:56:24.563 --> 00:56:24.704
S.
00:56:24.704 --> 00:56:30.693
states that have passed age verification for adult websites, more with social media.
00:56:31.329 --> 00:56:32.597
it really is happening everywhere.
00:56:33.068 --> 00:56:33.557
Why?
00:56:33.648 --> 00:56:35.887
You know, it's a whole bunch of factors.
00:56:35.987 --> 00:56:46.331
Um, one kind of interesting story with the Australia legislation is it seems to have stemmed depending on who you ask, from Jonathan Haidt's book.
00:56:46.489 --> 00:56:52.623
and, you know, there's, has been a lot of discussion, because of recent literature.
00:56:53.338 --> 00:56:58.329
The Anxious Generation is Jonathan Height's book, as I'm sure many of your listeners are familiar with.
00:56:58.759 --> 00:57:00.778
There's just been more discussion on this.
00:57:00.818 --> 00:57:02.478
There's been a lot of discussion in the U.
00:57:02.478 --> 00:57:02.918
S.
00:57:02.938 --> 00:57:08.239
You know, the Kids Online Safety Act, but there are, you know, true age verification bills being introduced.
00:57:08.838 --> 00:57:11.079
There's even a case going to the Supreme Court.
00:57:11.411 --> 00:57:13.402
And that's the Free Speech Coalition v.
00:57:13.443 --> 00:57:14.222
Paxton case.
00:57:14.242 --> 00:57:17.967
And that's something that the Internet Society Weight in on with an amicus.
00:57:17.967 --> 00:57:24.710
We joined with the Center for Democracy and Technology, New America's Open Technology Institute, and some academics on that case.
00:57:25.034 --> 00:57:33.414
And, you know, the thing that we think doesn't get talked about enough is that age verification laws are not just about young people.
00:57:33.902 --> 00:57:34.625
to
00:57:34.824 --> 00:57:38.650
verification laws are about everyone, if not done right.
00:57:39.318 --> 00:57:56.085
You know, this puts everyone's privacy security at risk and like Natalie already said, you know, this can be discriminatory this the marginalized communities and there's really interesting research on this marginalized communities can benefit the most from social media can benefit the most from being online.
00:57:56.735 --> 00:57:58.706
It can be more difficult for them to get online.
00:57:58.905 --> 00:58:01.106
There can be more social factors that make it.
00:58:01.181 --> 00:58:08.670
Um, you know, generally their worlds, it's more difficult.
00:58:08.864 --> 00:58:10.543
To make those types of connections.
00:58:10.893 --> 00:58:16.873
so what we're really focused on is that everyone can get access to the Internet that is not.
00:58:17.434 --> 00:58:23.443
Gating off access to the Internet, to news, information, health, entertainment, right?
00:58:23.800 --> 00:58:32.201
so this is, this is really going to be a challenge and we see so many pieces of regulation that are being implemented and introduced right now.
00:58:32.543 --> 00:58:35.481
luckily at the Internet Society, we have an incredible community.
00:58:36.018 --> 00:58:38.277
for instance, our Australia chapter.
00:58:38.757 --> 00:58:39.637
Jumps to action.
00:58:39.637 --> 00:58:48.233
There was less than 24 hours to file, comments, they made the deadline, got the comments in, you know, and that's so important.
00:58:48.233 --> 00:58:57.561
And so many of our, our local, you know, regional chapters are engaging with governments in their home countries on this issue.
00:58:58.021 --> 00:59:01.891
So on that point, you know, there's really good debate on this.
00:59:02.326 --> 00:59:03.786
really great interaction.
00:59:04.076 --> 00:59:09.195
More and more technologists are getting involved in this, more standards development organizations.
00:59:09.445 --> 00:59:13.985
So the engineers who actually make the internet function, right, are getting involved on this.
00:59:14.266 --> 00:59:17.706
So that's encouraging, but it really is happening everywhere at once.
00:59:18.065 --> 00:59:27.938
And we really need to make sure that those who need the internet the most, can actually be safe online and are not having their privacy and security exposed through all this.
00:59:29.903 --> 00:59:47.173
you've both talked a little bit about sort of the implications of this and the reasons why there are concerns, but Natalie, I want to finish up with you and just say, you know, for policymakers, some of whom might be listening to this, hopefully, um, who are looking at these laws and thinking about them, what kinds of.
00:59:47.454 --> 00:59:53.833
Factors, what should they be considering before, you know, proposing or voting for these kinds of laws?
00:59:54.436 --> 01:00:01.228
So, first, like, it's really important to understand that the Internet Society works to make sure that the Internet is for everyone.
01:00:01.259 --> 01:00:08.489
And when we talk about an open Internet, we're talking about making sure we're lowering barriers to access to the Internet.
01:00:08.994 --> 01:00:11.074
We also want a healthy Internet.
01:00:11.094 --> 01:00:18.833
We are, you know, like John mentioned, we're a huge community, not just the Internet Society, but our chapters and members and organizational members around the world.
01:00:19.193 --> 01:00:23.994
We all want a healthy Internet and care about making sure there are safe spaces for people online.
01:00:24.664 --> 01:00:31.373
But we also want to make sure that, you know, things like encryption and the fundamentals of a secure Internet.
01:00:31.903 --> 01:00:41.273
Are not undermined and that people don't experience barriers to access, that could be complete hurdles to accessing the Internet in the first place.
01:00:41.813 --> 01:00:50.831
So, I mean, it's not to say that there might never be a solution for age verification that could not hinder things like security and open Internet.
01:00:51.400 --> 01:00:56.951
But policymakers really do have to be thinking through how could their proposals.
01:00:57.177 --> 01:01:00.927
Impact, people's access to the Internet and their safety online.
01:01:01.438 --> 01:01:07.128
And we have a tool that helps us analyze these proposals is called the Internet Impact Assessment Toolkit.
01:01:07.458 --> 01:01:11.403
We think of it like an environmental impact assessment for the Internet.
01:01:11.824 --> 01:01:21.240
And so what we do, is, um, Offer ourselves to work with governments who are thinking about, whatever issues are working out that may relate to the Internet.
01:01:21.251 --> 01:01:32.280
And we'll often use this framework that describes what the Internet needs to exist in the 1st place and to be more open, global connected, secure and trustworthy and we help them think through how might.
01:01:32.583 --> 01:01:36.893
Particular proposal impact these goals for a healthy Internet.
01:01:37.623 --> 01:01:45.278
And so always are available to work with governments to think through these aspects, but we have a toolkit that policymakers can use themselves.
01:01:45.858 --> 01:02:03.483
And we think that it's really important that, We just don't jump to law proposals that don't consider those impacts on the internet because such as a case in Canada, there can be very extreme consequences for people's access to the internet and their very safety and security online.
01:02:03.947 --> 01:02:04.447
All right.
01:02:04.478 --> 01:02:10.771
Well, Natalie and John, thank you for coming on the podcast and thank you for all the good work that the Internet Society does.
01:02:11.170 --> 01:02:25.181
And, uh, I hope if anyone's listening to this and you are, working on a bill like this or thinking about these kinds of laws that you, uh, listen closely and, take a look at what the Internet Society is doing on this and, and what, resources they have available.
01:02:25.731 --> 01:02:26.768
Thanks again for joining us.
01:02:27.103 --> 01:02:27.742
Thanks Mike.
01:02:27.891 --> 01:02:28.387
Thanks, Mike.
01:02:31.677 --> 01:02:33.818
Thanks for listening to Ctrl-Alt-Speech.
01:02:34.007 --> 01:02:37.197
Subscribe now to get our weekly episodes as soon as they're released.
01:02:37.998 --> 01:02:44.768
If your company or organization is interested in sponsoring the podcast, contact us by visiting ctrlaltspeech.Com.
01:02:44.958 --> 01:02:47.157
That's C T R L Alt Speech.
01:02:47.518 --> 01:02:47.847
com.
01:02:48.748 --> 01:03:01.818
This podcast is produced with financial support from the Future of Online Trust and Safety Fund, a fiscally sponsored multi donor fund at Global Impact that supports charitable activities to build a more robust, capable, and inclusive trust and safety ecosystem.
