techgrumps 3.38 - The bettification of everything

WEBVTT
00:00:01.099 --> 00:00:05.559
Welcome, everyone, to Tech Grumps 3 .38. Tonight
00:00:05.559 --> 00:00:08.279
is Sunday, the 8th of March. My name is Ryan
00:00:08.279 --> 00:00:11.199
Alexander. I am an Internet Engineering Manager
00:00:11.199 --> 00:00:14.699
and General Person about London. And joining
00:00:14.699 --> 00:00:19.120
us today for Tech Grumps is David. Hi there.
00:00:19.199 --> 00:00:22.039
David Eastman, again, from London. Man about
00:00:22.039 --> 00:00:24.620
the town as well, of course. Eastman on Blue
00:00:24.620 --> 00:00:28.300
Sky, and I write for thenewstack .io. Sadly,
00:00:28.480 --> 00:00:33.020
mainly about AI. Also joining us is Ian. Hi,
00:00:33.159 --> 00:00:35.780
I'm Ian Forrester, and I am cubic garden everywhere,
00:00:36.060 --> 00:00:39.759
including the Fediverse and Blue Sky, although
00:00:39.759 --> 00:00:44.079
I'm also migrating to Eurosky, so look out for
00:00:44.079 --> 00:00:46.960
that. And I am in Manchester, not in London.
00:00:47.539 --> 00:00:51.590
And last but certainly never least, Wendy. I'm
00:00:51.590 --> 00:00:53.829
Wendy Grossman. I'm from New York, but based
00:00:53.829 --> 00:00:56.549
in London and everything, including my social
00:00:56.549 --> 00:01:00.070
media stuff and the NetWars blog and all linked
00:01:00.070 --> 00:01:07.430
from my website at pelicancrossing .net. Okay,
00:01:07.430 --> 00:01:11.069
as has become a Tech Grumps tradition, we are
00:01:11.069 --> 00:01:14.049
going to kick it off with our AI Roundup, because
00:01:14.049 --> 00:01:17.810
who's talking about anything else? We have six
00:01:17.810 --> 00:01:21.269
stories and points raised for the AI Roundup.
00:01:21.290 --> 00:01:23.569
I'm going to just go through them real quick,
00:01:23.650 --> 00:01:25.489
and then we will have some discussion around
00:01:25.489 --> 00:01:29.569
them. We have submissions from Ryan, David, and
00:01:29.569 --> 00:01:34.810
Wendy. So we have a link from Simon Willison.
00:01:35.200 --> 00:01:37.359
perennial friend of the podcast, although he's
00:01:37.359 --> 00:01:39.540
never on, which is about comparing and showing
00:01:39.540 --> 00:01:42.900
diffs of the OpenAI mission statement over time.
00:01:43.040 --> 00:01:45.019
So you can see how they've been changing their
00:01:45.019 --> 00:01:47.840
mission statement. From David, we have a link
00:01:47.840 --> 00:01:50.739
to the new web is marked down because agents
00:01:50.739 --> 00:01:54.799
don't want ads. From myself, we have an article
00:01:54.799 --> 00:02:00.489
from Tom Loosemore about... The end, how AI agents
00:02:00.489 --> 00:02:04.090
can end the policy of using friction as a slowdown
00:02:04.090 --> 00:02:06.150
to provide access to government services, which
00:02:06.150 --> 00:02:09.870
I found fascinating. From David, we have an app
00:02:09.870 --> 00:02:12.250
that predicts if smart glasses are nearby. Again,
00:02:12.349 --> 00:02:15.849
from David, we have LLMs can de -anonymize pseudonymous
00:02:15.849 --> 00:02:18.189
online accounts, similar to Discord offering.
00:02:18.699 --> 00:02:21.860
to judge age stochastically. And lastly, from
00:02:21.860 --> 00:02:25.080
Wendy, she has found a paper that is showing
00:02:25.080 --> 00:02:28.520
that LLMs can link profiles and postings at speed
00:02:28.520 --> 00:02:31.780
and at scale, pretty much killing online anonymity.
00:02:31.879 --> 00:02:37.979
Any of these catch our attention first? I definitely
00:02:37.979 --> 00:02:42.500
would like to have a little chat about the markdown
00:02:42.500 --> 00:02:46.780
stuff, which David... Yeah, yeah. Well, so...
00:02:48.039 --> 00:02:52.919
I quite like Markdown, actually, now. Yeah, I
00:02:52.919 --> 00:02:56.060
think Markdown is good in itself, quite separately
00:02:56.060 --> 00:03:02.840
from its current fancy by AI. When agents are
00:03:02.840 --> 00:03:04.939
talking, they have this thing called a context
00:03:04.939 --> 00:03:07.659
window, which is kind of how much they can process,
00:03:07.699 --> 00:03:09.620
how big their memory is. Maybe that's the closest
00:03:09.620 --> 00:03:12.599
thing. The problem is a lot of websites come
00:03:12.599 --> 00:03:15.020
back with a lot of verbiage, and basically they
00:03:15.020 --> 00:03:19.039
crunch down unnecessarily. including, you know,
00:03:19.120 --> 00:03:22.919
with navigation information that agents can't
00:03:22.919 --> 00:03:26.300
even see. And it sort of uses that window up
00:03:26.300 --> 00:03:28.539
and they become less good at browsing the web,
00:03:28.620 --> 00:03:30.520
assuming you want your agent to browse the web.
00:03:30.979 --> 00:03:34.960
So what Cloudflare started to do is on certain,
00:03:35.060 --> 00:03:37.840
some of their CDNs and edges, they're starting
00:03:37.840 --> 00:03:40.979
to automatically give agents or things that have
00:03:40.979 --> 00:03:43.900
announced themselves as agents a markdown. And
00:03:43.900 --> 00:03:45.819
markdown is much more efficient in that respect.
00:03:46.600 --> 00:03:49.360
The interesting thing here is that if you're
00:03:49.360 --> 00:03:52.719
starting to say that agents like efficient web
00:03:52.719 --> 00:03:54.659
pages without ads, a lot of humans are going
00:03:54.659 --> 00:03:56.419
to say, well, fuck it. We actually also like
00:03:56.419 --> 00:03:59.159
the idea of web pages without ads. Absolutely.
00:03:59.539 --> 00:04:02.960
And then you realize what's going on underneath.
00:04:03.219 --> 00:04:06.099
This is, and I've said this in three tech grunts
00:04:06.099 --> 00:04:08.419
in a row now, this is the AI companies trying
00:04:08.419 --> 00:04:10.539
to undermine the web. They're not necessarily
00:04:10.539 --> 00:04:12.840
doing it for any other reason that they can,
00:04:13.039 --> 00:04:15.560
and they're successfully doing it. Because now
00:04:15.560 --> 00:04:17.680
we have... You know, agents going, oh, I don't
00:04:17.680 --> 00:04:19.040
want to read this rubbish on the web. Give me
00:04:19.040 --> 00:04:21.879
something different. And this is the way I am
00:04:21.879 --> 00:04:24.279
a little bit terrified by how we've come full
00:04:24.279 --> 00:04:27.540
circle by defining as the rubbish web as the
00:04:27.540 --> 00:04:29.720
web and that by removing the rubbish, we are
00:04:29.720 --> 00:04:33.300
undermining it. I was just going to say, if it
00:04:33.300 --> 00:04:36.740
makes it more like the web was in 1998, say,
00:04:36.839 --> 00:04:39.279
when you could actually read a web page would
00:04:39.279 --> 00:04:41.980
load really quickly because there wasn't a lot
00:04:41.980 --> 00:04:44.759
of JavaScript and crap on it. Why isn't that
00:04:44.759 --> 00:04:49.060
a benefit? Oh, I think a lot of people find that
00:04:49.060 --> 00:04:51.560
it is an improvement. The only problem is that
00:04:51.560 --> 00:04:53.040
when they look at the statistics, the people
00:04:53.040 --> 00:04:55.370
driving improvement. aren't human it's going
00:04:55.370 --> 00:04:58.230
to be ai agents now initially we might go oh
00:04:58.230 --> 00:04:59.829
this is actually pretty good we get a better
00:04:59.829 --> 00:05:02.629
web we get less rubbish on it and the ads and
00:05:02.629 --> 00:05:04.850
the ads disappear and then the revenue disappears
00:05:04.850 --> 00:05:07.110
and which again we're not necessarily initially
00:05:07.110 --> 00:05:08.949
going to care about but eventually we're going
00:05:08.949 --> 00:05:11.449
to go oh wait a minute this whole web model the
00:05:11.449 --> 00:05:13.470
web model has nothing to do with advertising
00:05:13.470 --> 00:05:16.730
but it kind of does now so it's going to be interesting
00:05:16.730 --> 00:05:18.910
and there's no question they're doing on purpose
00:05:18.910 --> 00:05:23.500
so when i saw this i okay so Hear me out. I don't
00:05:23.500 --> 00:05:26.699
want to hear any of you. Oh, my God, Ian. Could
00:05:26.699 --> 00:05:28.740
you just say the thing you want to say instead
00:05:28.740 --> 00:05:31.740
of, like, pouching it? We love you. Please proceed.
00:05:32.399 --> 00:05:36.860
Okay. Thank you. So one of the beautiful things
00:05:36.860 --> 00:05:42.399
about HTML and XHTML is that you can use something
00:05:42.399 --> 00:05:45.519
to, because it's structured in a nice way. you
00:05:45.519 --> 00:05:48.939
can literally rip out whole sections so you could
00:05:48.939 --> 00:05:52.319
literally go hey i'm only interested in everything
00:05:52.319 --> 00:05:55.379
in the body and i'm not interested in anything
00:05:55.379 --> 00:06:00.519
that's javascript and so this way of doing things
00:06:00.519 --> 00:06:04.139
of using markdown is interesting but i i think
00:06:04.139 --> 00:06:08.180
the idea of delivering another file is problematic
00:06:08.180 --> 00:06:12.480
and i also worry that, yeah, like what we've
00:06:12.480 --> 00:06:14.740
already talked about, it strips away some of
00:06:14.740 --> 00:06:18.779
the beautiful things about the web. And I know
00:06:18.779 --> 00:06:21.259
why they're doing this, but I also keep on thinking
00:06:21.259 --> 00:06:24.639
about what was that Google thing where it was
00:06:24.639 --> 00:06:28.500
a shortened version of a web page, and I can't
00:06:28.500 --> 00:06:31.000
remember the name of it, but it strikes me as
00:06:31.000 --> 00:06:35.500
that. And that didn't go very well. Amp. That's
00:06:35.500 --> 00:06:39.740
it. Yes, amp. I feel like we're going to get
00:06:39.740 --> 00:06:41.839
this problem where we're going to have the web
00:06:41.839 --> 00:06:44.579
page and the AMP page and then a markdown file,
00:06:44.879 --> 00:06:47.420
and I feel like this is not the way that it should
00:06:47.420 --> 00:06:53.699
be done. But that's... I just... Go on. For humans,
00:06:53.819 --> 00:06:56.300
they were allowed to increase the size of page
00:06:56.300 --> 00:07:00.920
loads by only 1 ,000%, just so they could flash
00:07:00.920 --> 00:07:03.259
some pixels at us, track where we are going,
00:07:03.439 --> 00:07:05.379
and every square inch of the screen that isn't
00:07:05.379 --> 00:07:06.939
the one square inch that contains the information
00:07:06.939 --> 00:07:10.329
we're looking for. is buzzing and yelling at
00:07:10.329 --> 00:07:13.050
us to frigging buy this or buy that. But the
00:07:13.050 --> 00:07:14.829
moment we start dealing with AIs, they're like,
00:07:14.930 --> 00:07:18.129
oh, that's too hard. We need to make it nice
00:07:18.129 --> 00:07:20.149
for the AIs. And I'm like, what about for us?
00:07:20.649 --> 00:07:24.649
How about just giving us the text for once in
00:07:24.649 --> 00:07:28.310
the last 20 years? I think we're just collateral
00:07:28.310 --> 00:07:34.389
damage, you know? Yeah. Wow. So it's baby and
00:07:34.389 --> 00:07:36.709
bathwater. So yes, it's nice to clear up and
00:07:36.709 --> 00:07:39.189
make sure we don't actually lose the web in an
00:07:39.189 --> 00:07:45.689
incentive. Interestingly, you talk about some
00:07:45.689 --> 00:07:47.350
of the nice formatting stuff, but the thing is
00:07:47.350 --> 00:07:49.449
that it was always the content. And it's the
00:07:49.449 --> 00:07:52.829
extra content that is the problem. Like the actual
00:07:52.829 --> 00:07:55.290
stuff that we need focusing on knowledge is so
00:07:55.290 --> 00:07:58.470
small. I'm going to push this a little bit into
00:07:58.470 --> 00:08:00.449
the thing that I brought in about the end of
00:08:00.449 --> 00:08:05.040
friction as policy. The Tom's Loose More article,
00:08:05.240 --> 00:08:09.139
which basically said anyone who's working like
00:08:09.139 --> 00:08:11.560
there's this thing because this seems related,
00:08:11.740 --> 00:08:15.560
which is that one of the things that governments
00:08:15.560 --> 00:08:21.319
do to control access to public services is they
00:08:21.319 --> 00:08:25.000
make it harder for them to access, is they make
00:08:25.000 --> 00:08:30.420
it more complicated, more difficult. You think
00:08:30.420 --> 00:08:38.720
that's a deliberate strategy? It absolutely is
00:08:38.720 --> 00:08:41.120
a deliberate strategy. It is 100 % a deliberate
00:08:41.120 --> 00:08:43.860
strategy. It is designed to, like, and this is
00:08:43.860 --> 00:08:46.840
actually, this has been really well documented
00:08:46.840 --> 00:08:48.399
in a number of places where they're like, we
00:08:48.399 --> 00:08:50.039
just want to make it more difficult because we
00:08:50.039 --> 00:08:53.299
know that for every five minutes we add to this
00:08:53.299 --> 00:08:56.549
process, There will be some people who will drop
00:08:56.549 --> 00:09:01.769
off. And from certain perspectives, the idea
00:09:01.769 --> 00:09:04.149
is, well, that's just us getting rid of the people
00:09:04.149 --> 00:09:05.590
who actually really didn't need it, because if
00:09:05.590 --> 00:09:07.370
they really didn't, if they really needed it,
00:09:07.370 --> 00:09:09.830
they'd keep doing it forever. Because if they
00:09:09.830 --> 00:09:11.330
need it, then they can jump through any number
00:09:11.330 --> 00:09:13.269
of hoops to get to it, because that's clearly
00:09:13.269 --> 00:09:17.529
showing that. Which is not true, which is it's
00:09:17.529 --> 00:09:23.629
a very. It's a. default to scrounger if unless
00:09:23.629 --> 00:09:27.789
proved otherwise mindset and the thing that tom
00:09:27.789 --> 00:09:29.669
loose more rights is going guess what ais can
00:09:29.669 --> 00:09:33.429
do really well now like we like we've already
00:09:33.429 --> 00:09:36.129
seen a version of this like i have on my phone
00:09:36.129 --> 00:09:40.009
i have a button that went on my phone and when
00:09:40.009 --> 00:09:41.789
i get a phone call which is almost never because
00:09:41.789 --> 00:09:49.240
who calls but like when somebody calls My mom
00:09:49.240 --> 00:09:51.600
calls. I'm not going to justify that with a response,
00:09:51.759 --> 00:09:55.600
Ian. When I've numbered that I don't recognize
00:09:55.600 --> 00:09:57.960
calls, there's a little button I can press. And
00:09:57.960 --> 00:10:00.419
my phone will be like, hi, this is the Google
00:10:00.419 --> 00:10:02.899
voice assistant. The person you are trying to
00:10:02.899 --> 00:10:04.259
call would like to know what it is that you're
00:10:04.259 --> 00:10:06.019
calling about. And it'll print out a transcript.
00:10:07.120 --> 00:10:09.379
And all of a sudden, I get to keep doing whatever
00:10:09.379 --> 00:10:12.100
the hell I want. And if they can't be bothered.
00:10:12.620 --> 00:10:15.700
To say to the agent, so I can see the text, what
00:10:15.700 --> 00:10:17.659
it is they're calling about, I don't have to
00:10:17.659 --> 00:10:21.039
answer. And a good 70 % to 80 % of those phone
00:10:21.039 --> 00:10:23.899
calls don't even bother. Isn't it quicker to
00:10:23.899 --> 00:10:26.789
just answer the phone call? Certainly not. I
00:10:26.789 --> 00:10:28.409
don't want to talk to some random person. No,
00:10:28.529 --> 00:10:30.450
because that tells them someone will pick up
00:10:30.450 --> 00:10:32.350
the phone and they will tell all of their friends
00:10:32.350 --> 00:10:33.990
that someone, that this phone number is picked
00:10:33.990 --> 00:10:36.190
up. If they get an agent that doesn't reply,
00:10:36.450 --> 00:10:38.330
they know that no matter what they're doing,
00:10:38.409 --> 00:10:41.210
it's going to have 20 minutes of time or 20 seconds
00:10:41.210 --> 00:10:43.889
or whatever of delay. By the way, normally nothing
00:10:43.889 --> 00:10:46.429
happens anyway. It goes boom, I go disconnect.
00:10:46.610 --> 00:10:48.529
I don't think I want you for a friend, Ryan.
00:10:49.190 --> 00:10:51.470
If I recognize your number, I'll pick up. If
00:10:51.470 --> 00:10:52.730
you're my friend, then I have your number in
00:10:52.730 --> 00:10:54.210
my phone and I'll go, oh, it's Wendy and I'll
00:10:54.210 --> 00:11:02.269
pick up. Yeah, I mean, Ryan, I do like that feature.
00:11:02.730 --> 00:11:05.389
I find it very useful. That's another version
00:11:05.389 --> 00:11:08.029
of this. And what's going on is that you will
00:11:08.029 --> 00:11:09.870
have, and we've already got this, we've got agents
00:11:09.870 --> 00:11:12.690
calling agents. And then eventually what's going
00:11:12.690 --> 00:11:16.940
to happen is... Just so we have open claw for
00:11:16.940 --> 00:11:19.720
the tech bros, someone's going to be smart enough.
00:11:19.759 --> 00:11:21.960
They're probably doing it right now. And they're
00:11:21.960 --> 00:11:25.100
being like, okay, well, we know that filing for
00:11:25.100 --> 00:11:27.360
this kind of government assistance is incredibly
00:11:27.360 --> 00:11:29.500
difficult in the UK, but we know what the rules
00:11:29.500 --> 00:11:30.899
are. And if you can pass the rules, you can get
00:11:30.899 --> 00:11:32.919
it. And even though it takes four days to go
00:11:32.919 --> 00:11:35.480
through, you can have us, we can set up a thing
00:11:35.480 --> 00:11:37.559
that'll give it, you know, just take the information
00:11:37.559 --> 00:11:39.840
that we actually know you need. And then you
00:11:39.840 --> 00:11:41.500
have the agent do it. You get on with your life
00:11:41.500 --> 00:11:43.039
and the agent comes back in four days and goes,
00:11:43.100 --> 00:11:49.500
yeah, OK, good. This could kill journalism. I'm
00:11:49.500 --> 00:11:52.659
curious to hear. Journalists routinely have to
00:11:52.659 --> 00:11:54.740
call people who, A, don't want to hear from them
00:11:54.740 --> 00:11:59.399
and B, don't know them. Oh, we're still on the
00:11:59.399 --> 00:12:00.799
phone answering thing. I've moved on. Sorry.
00:12:00.919 --> 00:12:04.340
Sorry. Yeah, I was just. Yeah. All right. But
00:12:04.340 --> 00:12:07.110
if it's a journalist and they're calling. Like,
00:12:07.110 --> 00:12:09.450
OK, so this this is a separate topic when I'm
00:12:09.450 --> 00:12:11.289
happy to talk about, which is the who is entitled
00:12:11.289 --> 00:12:13.789
to who is entitled to interrupt my life and demand
00:12:13.789 --> 00:12:15.350
without my knowing who they are, that I pick
00:12:15.350 --> 00:12:16.870
up a phone and talk to them at any given point
00:12:16.870 --> 00:12:19.570
in time. Yeah. All right. I take your point.
00:12:19.629 --> 00:12:21.230
Let's move back to the government thing. It's
00:12:21.230 --> 00:12:23.490
obviously more important. But well, I mean, they're
00:12:23.490 --> 00:12:25.149
both the things. But let's also like we have
00:12:25.149 --> 00:12:31.389
a couple other ones. The speaking of this. We've
00:12:31.389 --> 00:12:34.289
got this de -anonymized pseudonymous online accounts,
00:12:34.470 --> 00:12:37.649
but also. Wendy's paper about online anonymity.
00:12:37.710 --> 00:12:39.629
Can we merge those two together, Dave and Wendy?
00:12:40.009 --> 00:12:44.350
Well, maybe a bit. The paper is interesting because
00:12:44.350 --> 00:12:47.490
it's always been possible for a really, you know,
00:12:47.490 --> 00:12:54.309
an insistent human to follow, to link somebody
00:12:54.309 --> 00:12:56.710
across LinkedIn, you know, say their LinkedIn
00:12:56.710 --> 00:13:00.429
profile to postings they've made on Reddit. Sometimes,
00:13:00.470 --> 00:13:04.120
you know, sometimes You can tell because there's
00:13:04.120 --> 00:13:06.100
a particular style or a particular obsession.
00:13:06.440 --> 00:13:10.480
But, you know, the intelligence authorities have
00:13:10.480 --> 00:13:13.899
gotten very good at that. But what the paper
00:13:13.899 --> 00:13:17.960
says is that large language models vastly speed
00:13:17.960 --> 00:13:21.019
up this process and vastly enlarge the scale
00:13:21.019 --> 00:13:23.120
at which it can operate. And so what they're
00:13:23.120 --> 00:13:25.919
actually talking about is we need a new threat
00:13:25.919 --> 00:13:30.100
model for privacy online. And they don't really
00:13:30.100 --> 00:13:34.240
have ideas about what that should be. But a lot
00:13:34.240 --> 00:13:36.259
of people and I think the pseudonymous thing
00:13:36.259 --> 00:13:38.120
comes into this. A lot of people think that,
00:13:38.120 --> 00:13:39.899
you know, they post under one name on Reddit
00:13:39.899 --> 00:13:42.940
and they, you know, that their employer is not
00:13:42.940 --> 00:13:44.840
going to match that to their LinkedIn profile.
00:13:45.240 --> 00:13:48.159
But this would enable much, much more detailed
00:13:48.159 --> 00:13:53.059
and much more democratized version of that. Yeah,
00:13:53.100 --> 00:13:55.799
I just see the paper that I was pointing to.
00:13:57.230 --> 00:13:58.990
Kind of the same thing that Discord are doing.
00:13:59.090 --> 00:14:01.070
So you might remember, and we've talked about
00:14:01.070 --> 00:14:03.629
it before, Discord are offering to say, we can
00:14:03.629 --> 00:14:07.009
judge the age of people just by the stuff they
00:14:07.009 --> 00:14:09.169
say, which just means stochastically hitting
00:14:09.169 --> 00:14:12.330
their posts or whatever and go, that guy clearly
00:14:12.330 --> 00:14:14.509
a teenager because he's talking about trains,
00:14:14.629 --> 00:14:18.730
you know, or not. And same thing, exactly as
00:14:18.730 --> 00:14:20.269
when he was saying, people writing in different
00:14:20.269 --> 00:14:22.669
accounts still write in a similar way. They're
00:14:22.669 --> 00:14:25.049
still often, you know, sometimes they even...
00:14:25.230 --> 00:14:28.450
and abuse the same people. So a patient reading
00:14:28.450 --> 00:14:31.549
a lot of information. The only reason I use the
00:14:31.549 --> 00:14:33.370
word stochastically is just to make it understood.
00:14:33.750 --> 00:14:36.409
It's not one formula. You just have to hit a
00:14:36.409 --> 00:14:38.330
lot of things and go, oh, yeah, that looks similar.
00:14:38.529 --> 00:14:41.929
So as a judge, we had to do it. We had a case.
00:14:41.929 --> 00:14:45.169
I helped run a journalism forum on CompuServe
00:14:45.169 --> 00:14:48.289
years ago. And we had a guy who was a legitimate
00:14:48.289 --> 00:14:51.350
Fleet Street photographer. And he was he was
00:14:51.350 --> 00:14:52.929
I don't know what was wrong with him or what
00:14:52.929 --> 00:14:57.600
drug he was on. And he was threatening people
00:14:57.600 --> 00:15:00.000
like he was going to fill their shoes with concrete
00:15:00.000 --> 00:15:01.860
and drop them off a bridge. And we decided to
00:15:01.860 --> 00:15:05.200
ban him. And he kept coming back under new pseudonyms
00:15:05.200 --> 00:15:07.059
and stuff. And we always would recognize him
00:15:07.059 --> 00:15:09.259
really quickly because he would end up telling
00:15:09.259 --> 00:15:12.080
the same stories and making the same threats.
00:15:12.259 --> 00:15:14.440
And it was just really obvious after a while.
00:15:14.539 --> 00:15:19.039
It dropped from taking a few postings to two
00:15:19.039 --> 00:15:23.590
seconds of, okay, this is this guy again. You
00:15:23.590 --> 00:15:27.190
know, so it's always been possible, but you wouldn't
00:15:27.190 --> 00:15:28.870
need the expert. You wouldn't need the human
00:15:28.870 --> 00:15:30.529
in the loop at this point. Well, this is exactly
00:15:30.529 --> 00:15:33.129
it. Previously, we associated computing with
00:15:33.129 --> 00:15:36.149
having to do some cold hard facts and not capable
00:15:36.149 --> 00:15:38.549
of judgment. LLMs are now very good at certain
00:15:38.549 --> 00:15:41.190
types of judgment call based on patterns, exactly
00:15:41.190 --> 00:15:42.990
the same as human. They're also good at making
00:15:42.990 --> 00:15:45.450
wrong judgment calls, of course. Just like humans.
00:15:45.509 --> 00:15:48.370
Yeah, that's what I was going to say. Yeah, the
00:15:48.370 --> 00:15:51.139
problem is a scale. obviously if humans deal
00:15:51.139 --> 00:15:53.460
with five things and they get every tenth wrong
00:15:53.460 --> 00:15:55.019
that's not a problem but if they're dealing with
00:15:55.019 --> 00:15:56.620
a thousand things and they're going to get i
00:15:56.620 --> 00:15:58.259
think it depends on what the tenth thing they
00:15:58.259 --> 00:16:01.019
get wrong is oh of course if it's you it's really
00:16:01.019 --> 00:16:03.539
annoying i also think that this is actually coming
00:16:03.539 --> 00:16:05.480
back to wendy's point about journalism because
00:16:05.480 --> 00:16:08.519
journalism isn't about making a statistical guess
00:16:08.519 --> 00:16:11.899
the people who follow the discipline actually
00:16:11.899 --> 00:16:14.279
link these things that's i think one of the one
00:16:14.279 --> 00:16:17.029
of the things that makes Real journalism is people
00:16:17.029 --> 00:16:19.309
caring about the facts and doing proper linking,
00:16:19.450 --> 00:16:23.029
not just going, well, I have a 72 % statistical
00:16:23.029 --> 00:16:25.570
chance that this is the same person. Right. And
00:16:25.570 --> 00:16:29.769
there's also that line, isn't there, about if
00:16:29.769 --> 00:16:33.629
you're not writing things that somebody, if you're
00:16:33.629 --> 00:16:35.370
not publishing things that somebody doesn't want
00:16:35.370 --> 00:16:40.169
published, you're just doing PR. Yes. Yeah. I
00:16:40.169 --> 00:16:41.350
can't remember who said that originally, but
00:16:41.350 --> 00:16:43.370
that's a well -known phrase. It might have been
00:16:43.370 --> 00:16:48.000
Orwell. I don't know. Somebody. Let's get an
00:16:48.000 --> 00:16:50.019
AI agent to go and look that up while we do something
00:16:50.019 --> 00:16:54.139
else. I'll just give a little spin on this. So
00:16:54.139 --> 00:16:58.669
I'm worried about this because I can see... I
00:16:58.669 --> 00:17:01.649
could definitely see it being very useful for
00:17:01.649 --> 00:17:04.289
certain things, like the case that Wendy mentioned.
00:17:05.190 --> 00:17:07.609
Someone's bugging you or someone's interfering
00:17:07.609 --> 00:17:09.609
with something and they keep on coming back under
00:17:09.609 --> 00:17:12.890
new accounts and then it just goes, oh, yeah,
00:17:12.970 --> 00:17:18.190
this is the same person. My problem always is
00:17:18.190 --> 00:17:22.430
that a lot of this stuff when it comes to AI
00:17:22.430 --> 00:17:25.369
is that we suddenly become more reliant on it.
00:17:25.769 --> 00:17:29.490
And then when you are, on the other side of it
00:17:29.490 --> 00:17:32.930
you know then it becomes a real problem and we
00:17:32.930 --> 00:17:36.450
end up relying on it too much and we start to
00:17:36.450 --> 00:17:42.230
kind of go well it's 19 2 that is definitely
00:17:42.230 --> 00:17:45.289
you know you're definitely the same person that
00:17:45.289 --> 00:17:49.009
we kicked out 10 weeks ago so we're just going
00:17:49.009 --> 00:17:51.190
to automatically just remove you we're not even
00:17:51.190 --> 00:17:53.829
going to look at it and that's what that's my
00:17:53.829 --> 00:17:56.819
ultimate worry there's no kind of like what's
00:17:56.819 --> 00:17:58.940
not transparency there's no transparency but
00:17:58.940 --> 00:18:01.480
there's no kind of like way of kind of going
00:18:01.480 --> 00:18:05.799
hey tell me why you got there how you think i
00:18:05.799 --> 00:18:07.720
think there's i think there's more subtlety to
00:18:07.720 --> 00:18:09.200
this like the situation that wendy describes
00:18:09.200 --> 00:18:12.180
it's kind of like i would i would say like okay
00:18:12.180 --> 00:18:16.380
let's say a person comes on who is not the troll
00:18:16.380 --> 00:18:19.059
who's come on 10 times before and starts doing
00:18:19.059 --> 00:18:21.420
all of the same things that the troll was doing
00:18:21.420 --> 00:18:24.440
and i'm like well you should still kick their
00:18:24.440 --> 00:18:28.440
ass out Yeah, OK. The problem is the behavior,
00:18:28.519 --> 00:18:32.299
not necessarily the person. The person who repeated
00:18:32.299 --> 00:18:35.039
the behavior gave us an opportunity to refine
00:18:35.039 --> 00:18:37.660
our own algorithms to go, oh, that's really shitty
00:18:37.660 --> 00:18:40.140
behavior. We don't want that around here. So
00:18:40.140 --> 00:18:42.299
it's a slightly different situation like that.
00:18:42.539 --> 00:18:46.319
And it also really depends on what kind of judgment
00:18:46.319 --> 00:18:48.019
you're making. Like there was an article about
00:18:48.019 --> 00:18:52.099
how a certain well -known transphobic fantasy
00:18:52.099 --> 00:18:56.180
author started writing under a pseudonym. I wonder
00:18:56.180 --> 00:18:59.420
who that is. Many, many years ago. And originally,
00:18:59.440 --> 00:19:01.180
this person did not acknowledge this was a pseudonym.
00:19:01.299 --> 00:19:02.859
They were just probably because everyone was
00:19:02.859 --> 00:19:05.180
getting a lot of hate on them for perfectly legitimate
00:19:05.180 --> 00:19:11.539
reasons. And then some people did this lexicographical
00:19:11.539 --> 00:19:14.180
mapping and went, yeah, this is the same person.
00:19:14.539 --> 00:19:18.779
Yes, a bit of analysis. Doris Lessing years ago
00:19:18.779 --> 00:19:21.759
did an experiment and submitted a novel under
00:19:21.759 --> 00:19:24.859
a pseudonym to see if... She was trying to sort
00:19:24.859 --> 00:19:27.460
of see what the publishing industry was up to.
00:19:28.319 --> 00:19:32.839
And it got rejected all sorts of places. But
00:19:32.839 --> 00:19:35.259
there was one publisher, I think it was one she'd
00:19:35.259 --> 00:19:37.720
actually worked with, that said it sounded a
00:19:37.720 --> 00:19:40.119
little like early. It sounded a little like Doris
00:19:40.119 --> 00:19:42.940
Lessig. And she did get the novel published eventually,
00:19:43.039 --> 00:19:44.759
but I think she ended up having to put her name
00:19:44.759 --> 00:19:47.319
on it. But she was unusual in that she sort of
00:19:47.319 --> 00:19:49.119
darted into doing some science fiction and back
00:19:49.119 --> 00:19:52.079
again, which very few. literary authors could
00:19:52.079 --> 00:19:57.460
manage, but she did. I liked those novels, the
00:19:57.460 --> 00:20:01.539
Shikasta novels. Was it Shikasta? Yes, but the
00:20:01.539 --> 00:20:05.359
name is The Canyon. We've got one more article
00:20:05.359 --> 00:20:07.099
that we haven't really touched on, which is The
00:20:07.099 --> 00:20:09.319
Smart Glasses Nearby. David, do you want to tell
00:20:09.319 --> 00:20:11.779
us a little about that? Yeah, I'm not going to
00:20:11.779 --> 00:20:16.440
go, because I think it's kind of inside the words.
00:20:16.559 --> 00:20:20.019
The idea here is that a couple of hobbyists have
00:20:20.019 --> 00:20:24.430
been looking at the Bluetooth signatures of smart
00:20:24.430 --> 00:20:27.869
glasses. I guess it's Meta's Ray -Ban, they're
00:20:27.869 --> 00:20:29.730
the popular ones now, obviously it used to be
00:20:29.730 --> 00:20:33.470
a Google one. And they've managed to pirate this
00:20:33.470 --> 00:20:35.690
thing to a Bluetooth signature. And so people
00:20:35.690 --> 00:20:38.970
with this app can go, there's someone with a,
00:20:38.970 --> 00:20:43.250
you know, there's a glass hole nearby. Did you
00:20:43.250 --> 00:20:46.730
notice that the same time as that article came
00:20:46.730 --> 00:20:49.269
out, maybe a day or two before, at least I saw
00:20:49.269 --> 00:20:52.430
it a day or two before, a couple of Swedish newspapers
00:20:52.430 --> 00:20:56.509
did an investigation and talked to some of the
00:20:56.509 --> 00:20:59.309
poorly paid data labelers in Kenya and found
00:20:59.309 --> 00:21:03.769
out that these meta glasses are sending intimate
00:21:03.769 --> 00:21:09.309
video images to people in Kenya to use to train
00:21:09.309 --> 00:21:13.880
AI models. And, you know, there's no, you know,
00:21:13.880 --> 00:21:17.599
there's never turn off. Kenya doesn't have compatible
00:21:17.599 --> 00:21:21.500
data protection laws. You know, people don't
00:21:21.500 --> 00:21:24.180
necessarily realize that there's a pair of glasses
00:21:24.180 --> 00:21:26.579
sitting on their nightstand that is watching
00:21:26.579 --> 00:21:28.819
them come out of the bathroom and reading their
00:21:28.819 --> 00:21:30.940
bank card, which they happen to leave nearby.
00:21:31.079 --> 00:21:34.480
I mean, it's like this is absolute privacy horror.
00:21:35.309 --> 00:21:38.609
and uh you know and then then so i'm not sure
00:21:38.609 --> 00:21:41.029
whether this app was was in response to that
00:21:41.029 --> 00:21:44.170
or not but no i think it was just a i just i
00:21:44.170 --> 00:21:46.289
want to know when there's a glass on there but
00:21:46.289 --> 00:21:49.589
which frankly is the thing that amazes me is
00:21:49.589 --> 00:21:51.509
that they can only do that because they suck
00:21:51.509 --> 00:22:00.240
at making their spyware My brain goes, why would
00:22:00.240 --> 00:22:02.359
you make a pair of covert recording glasses give
00:22:02.359 --> 00:22:03.819
off a fucking Bluetooth signal? What's wrong
00:22:03.819 --> 00:22:08.700
with you? Oh, Meta. Yeah, okay. What really bugs
00:22:08.700 --> 00:22:10.859
me is different. What really bugs me is that
00:22:10.859 --> 00:22:13.180
there must have been hundreds or maybe even thousands
00:22:13.180 --> 00:22:15.119
of people at Meta who worked on this and they
00:22:15.119 --> 00:22:18.900
did not care or think it even worthy of anything.
00:22:19.720 --> 00:22:24.119
I would push back on that. Some of them probably
00:22:24.119 --> 00:22:26.140
cared, but they also possibly wanted to keep
00:22:26.140 --> 00:22:28.960
their job. And they also and they also recognize
00:22:28.960 --> 00:22:31.079
like it's very easy to set up a system where
00:22:31.079 --> 00:22:33.460
like this is I'm going to expand this a bit.
00:22:33.500 --> 00:22:35.799
This is the AI chat. Here's the thing. Right
00:22:35.799 --> 00:22:39.359
now in technology, everyone I like have in the
00:22:39.359 --> 00:22:44.140
last four months, no one is even willing to say
00:22:44.140 --> 00:22:46.700
maybe this is all a really bad idea anymore because
00:22:46.700 --> 00:22:51.130
they want to stay employed. And right now, everything
00:22:51.130 --> 00:22:53.849
has moved, and it's basically – and unfortunately,
00:22:53.950 --> 00:22:56.549
nobody knows what the future is, but what everyone
00:22:56.549 --> 00:22:58.970
is doing is going. The only thing that we can
00:22:58.970 --> 00:23:01.210
do is just show, yep, we're exploring it. We're
00:23:01.210 --> 00:23:02.730
excited about it. This is going to be great.
00:23:02.769 --> 00:23:04.849
Learning as fast as we can. Moving forward. Yay.
00:23:05.390 --> 00:23:09.309
I, for one, welcome our new robot ant overlords.
00:23:12.170 --> 00:23:17.079
Yeah, I don't know what to say about this. You
00:23:17.079 --> 00:23:20.880
are right, Ryan. You are right. I think to be
00:23:20.880 --> 00:23:25.059
employed in one of these companies and to be
00:23:25.059 --> 00:23:29.039
the voice of dissent in this is very difficult.
00:23:29.660 --> 00:23:33.279
I think it is to not be employed anymore. Yeah.
00:23:34.480 --> 00:23:39.039
It seems insane that it is that, but yeah. But
00:23:39.039 --> 00:23:41.500
these Kenyan workers who blew the whistle on
00:23:41.500 --> 00:23:45.400
it, I mean, they can lose their jobs too. Oh,
00:23:45.400 --> 00:23:47.799
yeah, absolutely. And they did it anyway. Yeah,
00:23:47.819 --> 00:23:52.099
yeah. Yes. I mean, you're right. I mean, I think
00:23:52.099 --> 00:23:53.700
about all these whistleblowers and all those
00:23:53.700 --> 00:23:58.640
people that stand up, and I think... I mean,
00:23:58.640 --> 00:24:01.539
I think about this myself. Like, would I have
00:24:01.539 --> 00:24:04.099
called the whistle on some of the things that
00:24:04.099 --> 00:24:09.420
I saw on the BBC? If it was bad enough, I would
00:24:09.420 --> 00:24:12.690
have. But then... It is that kind of like, well,
00:24:13.049 --> 00:24:16.670
especially right now when positions are really
00:24:16.670 --> 00:24:20.910
hard and, you know, there's a real, yeah, it's
00:24:20.910 --> 00:24:24.430
really difficult. There's an interesting extra
00:24:24.430 --> 00:24:26.329
factor when you start to talk about people who
00:24:26.329 --> 00:24:29.750
work in the countries that we usually go to for
00:24:29.750 --> 00:24:33.470
very, very cheap tech labor. And that's that.
00:24:33.549 --> 00:24:36.710
And I've done work with outsourced teams to a
00:24:36.710 --> 00:24:39.410
bunch of those areas. And here's the thing. They're
00:24:39.410 --> 00:24:42.900
not stupid. They know that people are going to
00:24:42.900 --> 00:24:45.839
them because they're inexpensive, and they also
00:24:45.839 --> 00:24:50.740
know that no one cares about them. In a weird
00:24:50.740 --> 00:24:53.380
way, yeah, someone who's working at that tech
00:24:53.380 --> 00:24:56.000
center in Kenya isn't going to be hounded for
00:24:56.000 --> 00:24:57.440
the rest of their days and kicked out of the
00:24:57.440 --> 00:25:02.339
industry because nobody cares. Well, actually,
00:25:02.460 --> 00:25:04.240
it's exactly this problem, that they are quite
00:25:04.240 --> 00:25:06.869
smart, and there's a wave from which way. they
00:25:06.869 --> 00:25:09.170
can be exploited until they go, yeah, we're actually
00:25:09.170 --> 00:25:11.309
quite good now. You can't exploit us quite so
00:25:11.309 --> 00:25:13.289
easily. And then the wave moves on to the next
00:25:13.289 --> 00:25:15.630
country. Well, yeah, I mean, if you look at...
00:25:15.630 --> 00:25:20.029
A really interesting perspective on this is in
00:25:20.029 --> 00:25:22.809
Kim Stanley Robinson's The Ministry for the Future
00:25:22.809 --> 00:25:26.490
book, which talks about certain areas of the
00:25:26.490 --> 00:25:28.589
world just going, oh, yeah, we're not going to
00:25:28.589 --> 00:25:30.670
be your cheap labor anymore because we actually
00:25:30.670 --> 00:25:35.650
need to not die now. Which has an interesting...
00:25:36.190 --> 00:25:37.529
example that but yeah you're absolutely right
00:25:37.529 --> 00:25:41.450
um okay anybody else for any of the ai roundup
00:25:41.450 --> 00:25:43.130
or have we talked ourselves into a depression
00:25:43.130 --> 00:25:45.289
once again i just thought we talked ourselves
00:25:45.289 --> 00:25:49.730
into a coma add one technical we um did a bit
00:25:49.730 --> 00:25:52.410
of a uk government baseball thing we talked about
00:25:52.410 --> 00:25:54.849
tom loose more about saying who he is he saw
00:25:54.849 --> 00:25:57.490
he sort of wrote the government digital strategy
00:25:57.490 --> 00:26:00.130
in the uk and he was the i think the deputy director
00:26:00.130 --> 00:26:02.569
of gds for about four or five years so i think
00:26:02.569 --> 00:26:05.670
that's what He's got a really kind of interesting,
00:26:05.750 --> 00:26:07.950
informed opinion on the government using friction
00:26:07.950 --> 00:26:12.369
as a way to... And many other things, yes. Worked
00:26:12.369 --> 00:26:20.329
alongside him. Yes, just for Americans. Okay,
00:26:20.450 --> 00:26:22.970
that's it for the roundup. And next up is Ian
00:26:22.970 --> 00:26:24.869
with an article about the prediction market.
00:26:24.950 --> 00:26:28.609
Take it away, Ian. Oh, yeah. So I've been following
00:26:28.609 --> 00:26:31.569
this for a while. I'm not on a prediction market.
00:26:34.290 --> 00:26:36.869
So for people who don't know what prediction
00:26:36.869 --> 00:26:39.849
markets are, it's for me, and I'm sure everyone
00:26:39.849 --> 00:26:43.529
would probably agree, it's really, it's very
00:26:43.529 --> 00:26:47.109
subtle, but it's basically gambling. You're gambling
00:26:47.109 --> 00:26:51.210
on things that are, I would say, ridiculous.
00:26:51.789 --> 00:26:56.630
So I know there was, anything you can think of,
00:26:56.730 --> 00:27:01.369
you can predict it, and they will go away and
00:27:01.369 --> 00:27:05.950
they will give you some odds. on it and um you
00:27:05.950 --> 00:27:09.549
put some money in there and then if it passes
00:27:09.549 --> 00:27:12.910
then you get a certain amount of money but other
00:27:12.910 --> 00:27:16.150
people can also bet on it at the same time um
00:27:16.150 --> 00:27:21.519
it's basically unlicensed gambling And I know
00:27:21.519 --> 00:27:23.799
it's really big in America. It's starting to
00:27:23.799 --> 00:27:26.380
get really big here. I know it's illegal in other
00:27:26.380 --> 00:27:28.539
parts of the world, but because it's on the internet,
00:27:28.759 --> 00:27:34.440
then it's very easy to get to. And I worry about...
00:27:34.440 --> 00:27:39.720
I deeply worry about how embedded this has become.
00:27:39.980 --> 00:27:42.980
So, for example, if you were not aware of the
00:27:42.980 --> 00:27:49.809
news of America taking out our... bombing iran
00:27:49.809 --> 00:27:53.890
um then that was on the prediction market and
00:27:53.890 --> 00:27:57.750
obviously there's also a slew of this kind of
00:27:57.750 --> 00:28:00.349
um was it when someone's kind of got information
00:28:00.349 --> 00:28:03.970
in the background i forgot the word is now um
00:28:03.970 --> 00:28:07.109
but there's a whole bunch of that kind of insider
00:28:07.109 --> 00:28:11.190
information insider information so yeah you can
00:28:11.190 --> 00:28:13.809
imagine i'm not saying this is this is what happened
00:28:14.670 --> 00:28:18.529
But someone who knew what was going to happen,
00:28:18.690 --> 00:28:21.349
could have put some stuff on the prediction market,
00:28:21.509 --> 00:28:25.069
made a killing, because they knew that it was
00:28:25.069 --> 00:28:27.910
going to happen at a certain time, in a certain
00:28:27.910 --> 00:28:31.529
place, and people bet with it or against it,
00:28:31.650 --> 00:28:33.930
and then they could walk away with loads of money.
00:28:34.069 --> 00:28:35.990
The other thing, just talking about the White
00:28:35.990 --> 00:28:39.069
House, sorry, is that there's talk about Trump
00:28:39.069 --> 00:28:43.359
setting up their own... set up his own prediction
00:28:43.359 --> 00:28:46.599
market as well or kind of getting into this field
00:28:46.599 --> 00:28:51.559
so I just I'm deeply worried about this but not
00:28:51.559 --> 00:28:55.319
only because I'm very conscious of people and
00:28:55.319 --> 00:28:57.839
addictions and I don't want to go there too much
00:28:57.839 --> 00:29:01.319
but I'm worried about this stuff but also I just
00:29:01.319 --> 00:29:04.400
feel like the things that are on there are so
00:29:04.400 --> 00:29:08.480
ridiculous And it just encourages this kind of
00:29:08.480 --> 00:29:11.140
mentality of like, well, is it on the prediction
00:29:11.140 --> 00:29:14.299
market? What's the possibilities? And you're
00:29:14.299 --> 00:29:15.859
up against people who have inside information
00:29:15.859 --> 00:29:19.160
and you're just losing money. And money is...
00:29:19.160 --> 00:29:21.480
I'll leave it at that. Has anyone else got any
00:29:21.480 --> 00:29:22.960
thoughts about this? Because it really pisses
00:29:22.960 --> 00:29:25.240
me off. I just want to add one point. I think
00:29:25.240 --> 00:29:28.099
it's an important technical point. They are setting
00:29:28.099 --> 00:29:30.259
up... This is the thing that keeps it, quote,
00:29:30.299 --> 00:29:32.880
legal. They are setting up markets where people
00:29:32.880 --> 00:29:35.029
are effectively betting against each other. It
00:29:35.029 --> 00:29:37.329
isn't one central person. If it's one central
00:29:37.329 --> 00:29:40.170
person giving odds, that isn't legal and that
00:29:40.170 --> 00:29:41.730
will be stopped. But because they're setting
00:29:41.730 --> 00:29:44.529
up a market where people are effectively accepting
00:29:44.529 --> 00:29:47.390
and giving odds to each other, that is pseudo
00:29:47.390 --> 00:29:49.109
-legal. I mean, until it isn't. Do you see what
00:29:49.109 --> 00:29:52.049
I mean? So it's kind of eBay for gamblers. Yes.
00:29:52.289 --> 00:29:54.450
Yeah. It's an important difference because most
00:29:54.450 --> 00:29:56.349
countries would ban it if it was just like...
00:29:56.349 --> 00:30:00.329
Okay, yeah. You are right. You are right. Sorry,
00:30:00.430 --> 00:30:03.890
that's an important point. But I think... That
00:30:03.890 --> 00:30:07.609
normalizes it. And actually, with what Wendy
00:30:07.609 --> 00:30:10.170
just said, we're talking about it like the eBay
00:30:10.170 --> 00:30:14.029
of gambling. And they literally want it to be
00:30:14.029 --> 00:30:16.470
across everything. And you're starting to see
00:30:16.470 --> 00:30:19.930
it. They're funding different events. I've not
00:30:19.930 --> 00:30:21.650
watched much football, but I gather there's a
00:30:21.650 --> 00:30:25.009
bunch of stuff in the football advertising about
00:30:25.009 --> 00:30:28.630
this stuff. I bet the Olympics were huge for
00:30:28.630 --> 00:30:33.220
them. Oh, yeah. Yeah. And it just... I guess
00:30:33.220 --> 00:30:35.819
what I'm worried about, what I'm concerned about,
00:30:35.880 --> 00:30:39.599
sorry, is that the normalization of gambling,
00:30:40.039 --> 00:30:43.019
but people don't think about it that way. That's
00:30:43.019 --> 00:30:45.940
why they call it prediction markets. And I just,
00:30:46.119 --> 00:30:52.000
I don't know. It just worries me. I wonder, like,
00:30:53.059 --> 00:30:55.819
it strikes me as a very fractional thing. Like,
00:30:55.819 --> 00:31:01.039
I think there's, I think, so in the ongoing meta
00:31:01.039 --> 00:31:03.410
battle, of finding the people who will be addicted
00:31:03.410 --> 00:31:07.130
if connecting the people who are addictive gamblers
00:31:07.130 --> 00:31:12.670
to methodologies of gambling um is a is a subset
00:31:12.670 --> 00:31:15.730
of people who are you know kind of really evil
00:31:15.730 --> 00:31:19.309
but we're never going to stop them and and so
00:31:19.309 --> 00:31:21.170
you have this certain percentage of folks who
00:31:21.170 --> 00:31:22.890
are just going to be like i am so into gambling
00:31:22.890 --> 00:31:24.869
that i will always find a way to gamble that
00:31:24.869 --> 00:31:26.589
is an addictive mindset and that's where your
00:31:26.589 --> 00:31:28.930
concern comes from the question is like how many
00:31:28.930 --> 00:31:32.380
more of that Because with any with any distinction,
00:31:32.599 --> 00:31:35.759
it's not there's a line. It's yeah, it's a gradient
00:31:35.759 --> 00:31:38.799
off into that. And it's like I so what I think
00:31:38.799 --> 00:31:42.240
is that like when for me, I'm not interested
00:31:42.240 --> 00:31:44.440
in gambling. So when it's like, hey, is there
00:31:44.440 --> 00:31:46.160
a new place I could go throw money down a well?
00:31:46.319 --> 00:31:53.759
That's wonderful to know. I bet on every. more
00:31:53.759 --> 00:31:55.559
or less every world cup match that's over four
00:31:55.559 --> 00:31:57.799
years i admit and the thing that puts me off
00:31:57.799 --> 00:32:00.900
is i only do it on paper in a betting shop and
00:32:00.900 --> 00:32:02.680
no one wants to spend time in a betting shop
00:32:02.680 --> 00:32:04.960
so i just spend the minimum amount of time and
00:32:04.960 --> 00:32:08.579
that's a good way so and that's i think that's
00:32:08.579 --> 00:32:11.180
kind of my point is that there there's there's
00:32:11.180 --> 00:32:14.920
a tremendous amount of nuances i mean but we
00:32:14.920 --> 00:32:16.460
have these two things here which is like are
00:32:16.460 --> 00:32:18.440
we making a moral argument against getting people
00:32:18.440 --> 00:32:19.720
addicted to gambling and all that other stuff
00:32:19.720 --> 00:32:22.940
but there's also like and that's not great But,
00:32:22.940 --> 00:32:24.660
you know, just like any other addiction, the
00:32:24.660 --> 00:32:26.980
addictive people are going to find their addiction.
00:32:27.200 --> 00:32:29.480
But then there's this entirely secondary market,
00:32:29.539 --> 00:32:31.039
which is like, well, when we're talking about,
00:32:31.039 --> 00:32:33.839
you know, like being addicted to drink or drink,
00:32:33.859 --> 00:32:36.240
be addicted to some other drugs, there's that.
00:32:36.359 --> 00:32:38.180
But then there's like we've got this whole new
00:32:38.180 --> 00:32:42.759
market and the market itself is like, and I lost
00:32:42.759 --> 00:32:45.539
I lost a thousand pounds because we just didn't
00:32:45.539 --> 00:32:47.480
hit the over under on how many children were
00:32:47.480 --> 00:32:51.089
killed in the bombing. Yes, exactly. There's
00:32:51.089 --> 00:32:53.490
a moral hazard to this that's different from
00:32:53.490 --> 00:32:57.829
drinking and drugs. And I think that that's a
00:32:57.829 --> 00:33:04.609
not to be ignored. Fuck no. But it is it is deserved
00:33:04.609 --> 00:33:08.930
by putting it in the same context as the moral
00:33:08.930 --> 00:33:10.730
hazard of some of these established gambling
00:33:10.730 --> 00:33:12.710
things. And I think that it's possible to look
00:33:12.710 --> 00:33:14.680
at. these two together. And I think, yeah, like,
00:33:14.779 --> 00:33:16.319
there are some people, there's going to be some
00:33:16.319 --> 00:33:18.099
folks for whom, like, the extra moral hazard
00:33:18.099 --> 00:33:21.539
is kind of the buzz for them. That's terrifying.
00:33:22.440 --> 00:33:25.500
But... Yeah, love a spell of napalm in the morning.
00:33:25.619 --> 00:33:27.039
Yeah, there's a bit of a drug in the people,
00:33:27.140 --> 00:33:31.339
actually, onto that purpose. Sorry, I think you're
00:33:31.339 --> 00:33:35.420
right. And I think this is my... It's a normalization
00:33:35.420 --> 00:33:39.920
of this stuff, which is kind of, like, deeply
00:33:39.920 --> 00:33:43.690
worrying. So even I'm hearing... i've heard in
00:33:43.690 --> 00:33:45.529
some podcasts not the podcast i'm listening to
00:33:45.529 --> 00:33:48.849
but i've heard from others where they are referring
00:33:48.849 --> 00:33:54.769
to the the bet on the prediction market you know
00:33:54.769 --> 00:33:57.369
in relation to a news story to the news stories
00:33:57.369 --> 00:33:59.250
that they're that they're talking about and i'm
00:33:59.250 --> 00:34:04.069
like okay this is this is a level that we shouldn't
00:34:04.069 --> 00:34:07.329
be going down i i worry that that normalization
00:34:07.329 --> 00:34:09.989
of this where people are just like oh what is
00:34:09.989 --> 00:34:12.809
that oh you know maybe i could also throw some
00:34:12.809 --> 00:34:16.650
money down the well as well, you know. It just,
00:34:16.869 --> 00:34:20.730
I don't know. I worry a lot. I feel there's a
00:34:20.730 --> 00:34:24.309
big difference between throwing some money at
00:34:24.309 --> 00:34:26.030
the World Cup because it increases your excitement
00:34:26.030 --> 00:34:28.289
of watching the match. I don't think that's anywhere
00:34:28.289 --> 00:34:31.030
near on the same level. What is that about? That's
00:34:31.030 --> 00:34:33.469
the part I really don't understand. Like, if
00:34:33.469 --> 00:34:36.210
the sport isn't exciting enough by itself, why
00:34:36.210 --> 00:34:39.090
are you watching it? Well, actually, you can
00:34:39.090 --> 00:34:41.110
also bet the matches you're not watching because
00:34:41.110 --> 00:34:43.230
there's often... two or three matches on at the
00:34:43.230 --> 00:34:47.110
same time in the World Cup. Yeah, okay, but Ryan
00:34:47.110 --> 00:34:50.610
just said, increasing the excitement of watching
00:34:50.610 --> 00:34:52.969
the sport. And this is the part I don't get.
00:34:53.090 --> 00:34:57.150
Why? Well, I like being right and then getting
00:34:57.150 --> 00:34:58.969
paid for being right. I think that's not fun.
00:34:59.170 --> 00:35:01.510
You don't like getting strung out on heroin either.
00:35:02.159 --> 00:35:05.079
But some people do. I mean, to be fair, there
00:35:05.079 --> 00:35:08.760
is absolutely no amount I could gamble on a World
00:35:08.760 --> 00:35:11.079
Cup match that would make me want to watch a
00:35:11.079 --> 00:35:14.420
World Cup match. But this is a different problem.
00:35:14.719 --> 00:35:18.159
I'm quite happy to watch without betting. I mean,
00:35:18.159 --> 00:35:23.280
trying to understand the mindset of what makes
00:35:23.280 --> 00:35:25.199
that is, you know, we live in a diverse world.
00:35:25.239 --> 00:35:26.260
Different people like different things. Some
00:35:26.260 --> 00:35:28.440
people like going out and watching. Some people
00:35:28.440 --> 00:35:30.679
like going out and watching birds. And I don't
00:35:30.679 --> 00:35:32.619
think there's anything wrong with that. It might
00:35:32.619 --> 00:35:33.920
not be the thing that excites me. I have some
00:35:33.920 --> 00:35:36.320
people who I actually really like a lot who do
00:35:36.320 --> 00:35:38.639
really hardcore bird watching. And I'm like,
00:35:38.679 --> 00:35:40.659
I am happy you have found something that you
00:35:40.659 --> 00:35:42.940
can engage with. I don't think less of you for
00:35:42.940 --> 00:35:45.239
it. I'm just not interested in doing it. It would
00:35:45.239 --> 00:35:47.699
have to be a brightly colored parrot for me to
00:35:47.699 --> 00:35:55.630
see it. But OK, so that example, right? So. Imagine,
00:35:55.769 --> 00:35:58.210
and this is the thing you could do on a polymarket,
00:35:58.329 --> 00:36:00.190
right, on a polymarket, is you could literally
00:36:00.190 --> 00:36:05.869
go, I'm going to put a bet on that not only am
00:36:05.869 --> 00:36:07.510
I going to see this bird, I'm going to take a
00:36:07.510 --> 00:36:10.230
picture of it, and it's going to be in this area.
00:36:10.510 --> 00:36:12.969
And that is something that someone else could
00:36:12.969 --> 00:36:16.530
bet on. And it's that kind of, like, betting
00:36:16.530 --> 00:36:20.610
on things that are... I'm not going to talk about
00:36:20.610 --> 00:36:22.329
the football because I have very particular views
00:36:22.329 --> 00:36:25.889
about football. But I think this idea of betting
00:36:25.889 --> 00:36:29.230
on everything is the bet -ification, I'm going
00:36:29.230 --> 00:36:31.889
to use that word for now, bet -ification of everything
00:36:31.889 --> 00:36:37.130
really pisses me off big time. A lot of life
00:36:37.130 --> 00:36:40.360
is based on betting. It doesn't have to be. Why
00:36:40.360 --> 00:36:43.599
does it have to be? David made a really important
00:36:43.599 --> 00:36:46.579
point, which is that you only get to bet on these
00:36:46.579 --> 00:36:49.239
things in the way the market is set up. You only
00:36:49.239 --> 00:36:51.420
get to bet on these things when you find someone
00:36:51.420 --> 00:36:54.760
else stupid enough to bet the other way. That's
00:36:54.760 --> 00:36:57.599
what makes a market. Yeah, but it's all that
00:36:57.599 --> 00:37:01.460
insurance. It's hedging if you're a professional,
00:37:01.500 --> 00:37:05.000
but for us, it's just market making. Actually,
00:37:05.219 --> 00:37:07.000
that's a really interesting question. Why is
00:37:07.000 --> 00:37:10.820
it exciting when you're gambling on a football
00:37:10.820 --> 00:37:15.679
match, but not when you're, quote, ensuring your
00:37:15.679 --> 00:37:19.000
health or your life? Why don't you find your
00:37:19.000 --> 00:37:24.559
pension exciting? Needs the same skill. Of course,
00:37:24.579 --> 00:37:26.420
you're doing something interesting. You're referring
00:37:26.420 --> 00:37:29.760
to what is gambling. You're not wrong that insurance
00:37:29.760 --> 00:37:32.679
represents an odd kind of gambling. What an insurance
00:37:32.679 --> 00:37:34.659
really is is just an installment payment against
00:37:34.659 --> 00:37:37.639
the lottery, the negative lottery ticket. Yes,
00:37:37.679 --> 00:37:41.599
yeah, I agree. I'm actually talking about everyone's
00:37:41.599 --> 00:37:44.590
pension funds, which is indeed. gambling on the
00:37:44.590 --> 00:37:47.590
market and i will not it is it is gambling that's
00:37:47.590 --> 00:37:49.829
it's just done by people with formal names who
00:37:49.829 --> 00:37:52.130
you pay so you go oh that's okay my pension fund's
00:37:52.130 --> 00:37:54.429
not being gambled with it's being carefully allotted
00:37:54.429 --> 00:37:57.690
to van garden blackstone and so forth which of
00:37:57.690 --> 00:38:00.489
course we think is fine by the way they'll do
00:38:00.489 --> 00:38:02.289
pretty well out of the wall but that's not related
00:38:02.289 --> 00:38:09.130
well i bet you didn't think how invested we'd
00:38:09.130 --> 00:38:12.250
be in that discussion did you Sorry, couldn't
00:38:12.250 --> 00:38:15.570
resist. Next up is Wendy talking about the disappearing
00:38:15.570 --> 00:38:18.329
web. Wendy, what have you got for us? Well, there's
00:38:18.329 --> 00:38:20.949
two stories that came up around the same time.
00:38:21.030 --> 00:38:23.809
One is in the Register based on a Pew survey
00:38:23.809 --> 00:38:27.789
that shows that 25 % of web pages vanish in a
00:38:27.789 --> 00:38:31.550
decade. So, you know, sort of the disappearing
00:38:31.550 --> 00:38:33.889
web. And then there's a second story at Tech
00:38:33.889 --> 00:38:38.869
Dirt, which talks about the Wayback Machine at
00:38:38.869 --> 00:38:42.070
the Internet Archive is being blocked by some
00:38:42.070 --> 00:38:46.690
of the main news sites. And so their articles
00:38:46.690 --> 00:38:50.239
won't necessarily be preserved in. the Wayback
00:38:50.239 --> 00:38:54.699
Machine. And as we know, lots of commercial publications
00:38:54.699 --> 00:38:57.400
do actually keep redesigning their websites and
00:38:57.400 --> 00:39:00.900
screwing up all their old URLs. So these are
00:39:00.900 --> 00:39:05.280
both cases where things would count on. And there's
00:39:05.280 --> 00:39:08.360
a third thing, which is not a story, but simply
00:39:08.360 --> 00:39:11.219
my own comment, which is that I worry an awful
00:39:11.219 --> 00:39:13.500
lot about the Internet Archive because we all
00:39:13.500 --> 00:39:18.360
treat it as the repository of the history of
00:39:18.360 --> 00:39:20.960
the web. And it's a single point of failure,
00:39:21.079 --> 00:39:23.519
you know, at least organizationally. I know they
00:39:23.519 --> 00:39:25.559
have servers around the world and they have backups
00:39:25.559 --> 00:39:27.940
and all the rest of it, but it's still, it's
00:39:27.940 --> 00:39:31.780
a single point of failure. So it's interesting
00:39:31.780 --> 00:39:33.840
what you should say, because I can't see the
00:39:33.840 --> 00:39:37.059
Wayback Machine because my ISP blocks it. And
00:39:37.059 --> 00:39:38.920
I don't blame them for blocking it. They block
00:39:38.920 --> 00:39:41.880
it? Why would they block it? Because it can see
00:39:41.880 --> 00:39:45.389
anything on the web. So it's an indirection.
00:39:45.610 --> 00:39:48.389
If they want, they can have a whole load of porn
00:39:48.389 --> 00:39:51.849
pages. Wait a minute. Name and shame your ISP.
00:39:51.949 --> 00:39:56.389
Yeah, really. That's outrageous. I think it's
00:39:56.389 --> 00:39:59.349
three, actually, who do that. So to get around
00:39:59.349 --> 00:40:01.889
that, I would probably have to... No, wait a
00:40:01.889 --> 00:40:07.150
minute. I have an account with three. But I didn't
00:40:07.150 --> 00:40:09.730
try... But it's on my phone, and I've never tried...
00:40:10.010 --> 00:40:12.690
you know certainly at home but i think the only
00:40:12.690 --> 00:40:14.769
i think you can get around it if you go to them
00:40:14.769 --> 00:40:17.170
and say i am an adult or something which is automatic
00:40:17.170 --> 00:40:20.329
yeah credit card i don't even know yeah mobile
00:40:20.329 --> 00:40:23.869
mobile all mobile phones are are blocked in the
00:40:23.869 --> 00:40:27.789
uk uh or they're you're blocked as if you're
00:40:27.789 --> 00:40:29.710
a child in the uk when you get it you always
00:40:29.710 --> 00:40:33.530
have to go in and uh and tell them you're an
00:40:33.530 --> 00:40:36.929
adult you only have to do it once And they throw
00:40:36.929 --> 00:40:38.730
away that machine for that very reason, this
00:40:38.730 --> 00:40:40.369
indirection to things that they would otherwise
00:40:40.369 --> 00:40:44.829
block. Oh, well, that's okay. It's silly, but
00:40:44.829 --> 00:40:50.170
yeah, just point that out. Okay, so going, yeah.
00:40:50.849 --> 00:40:53.489
Thank you. I've got a free business account,
00:40:53.630 --> 00:40:55.789
but I'm not using it for my broadband, that's
00:40:55.789 --> 00:40:58.949
for sure. So I do think, Wendy, you're absolutely
00:40:58.949 --> 00:41:02.090
right. And I'm always worried about... these
00:41:02.090 --> 00:41:05.909
things that we take for granted such as wikipedia
00:41:05.909 --> 00:41:10.449
or wikimedia um the you know the archive archive
00:41:10.449 --> 00:41:15.150
you know it's just it's really hard and these
00:41:15.150 --> 00:41:19.030
things should be supported like properly supported
00:41:19.030 --> 00:41:24.530
not relying on funds or you know kind of throwing
00:41:24.530 --> 00:41:26.949
the hat around trying to get money from people
00:41:26.949 --> 00:41:30.150
um because yeah when you really need it you really
00:41:30.150 --> 00:41:33.840
really need it I also am aware there's lots of
00:41:33.840 --> 00:41:38.400
other ones that are archiving, like archive .ph
00:41:38.400 --> 00:41:40.219
and stuff like that, but I don't know if they
00:41:40.219 --> 00:41:44.659
actually do long -term archive. We know nothing
00:41:44.659 --> 00:41:46.920
about them. We know nothing about them. We don't
00:41:46.920 --> 00:41:50.900
know who owns them. Yeah, that's true. We don't
00:41:50.900 --> 00:41:55.980
know who funds them. You get the point. And a
00:41:55.980 --> 00:41:59.980
lot of what that's used for is to get by paywalls.
00:42:00.059 --> 00:42:04.530
Paywalls, yeah. And, you know, the day is surely
00:42:04.530 --> 00:42:07.030
coming when somebody tries to sue it out of existence.
00:42:07.590 --> 00:42:11.309
Yeah. I knew someone in the Netherlands who was
00:42:11.309 --> 00:42:15.210
running something like that. And actually, they
00:42:15.210 --> 00:42:19.030
came to the BBC and asked if the BBC would be
00:42:19.030 --> 00:42:22.789
OK with it. Obviously, that conversation didn't
00:42:22.789 --> 00:42:26.750
go very far, you know. They are trying, but you're
00:42:26.750 --> 00:42:28.090
right. You're absolutely right. We don't know
00:42:28.090 --> 00:42:31.030
who's behind it. We don't know their ethos. We
00:42:31.030 --> 00:42:33.610
don't know what their... We don't know anything
00:42:33.610 --> 00:42:37.570
about them, so you're absolutely right. I really
00:42:37.570 --> 00:42:43.550
wish we had some actual proper funding. Maybe
00:42:43.550 --> 00:42:47.230
this should be something that the public service
00:42:47.230 --> 00:42:51.309
internet should fund. Maybe the BBC? Well, the
00:42:51.309 --> 00:42:56.090
National Archives do keep quite a bit of stuff.
00:42:56.429 --> 00:43:00.150
I don't know that they try to back up the British
00:43:00.150 --> 00:43:03.130
web, but they, for example, I remember hearing
00:43:03.130 --> 00:43:08.170
about a project to archive games and things.
00:43:08.789 --> 00:43:12.269
Yeah. I worked with them on this before. Yeah.
00:43:12.329 --> 00:43:14.550
I think both my old games are on the Internet
00:43:14.550 --> 00:43:18.449
Archive. Yes, they do. No, I said the National
00:43:18.449 --> 00:43:20.510
Archives. I'm talking about the National Archives
00:43:20.510 --> 00:43:23.809
here in Kew. Yeah, well, that's basically, yes.
00:43:25.329 --> 00:43:29.530
But you get into a problem where... You get into
00:43:29.530 --> 00:43:32.530
a problem where, obviously, most of the commercial
00:43:32.530 --> 00:43:35.909
entities do not want their stuff archived, and
00:43:35.909 --> 00:43:38.929
then you get into this big row, and... Yeah,
00:43:39.130 --> 00:43:42.190
and I can see the National Archive and other
00:43:42.190 --> 00:43:44.699
archives just going, you know what? We're just
00:43:44.699 --> 00:43:46.519
going to leave it alone because it's just not
00:43:46.519 --> 00:43:50.079
worth the hassle. Yes, I mean, there's rules
00:43:50.079 --> 00:43:52.659
about certain things can be archived at a certain
00:43:52.659 --> 00:43:55.340
time, but obviously... But that's why we have
00:43:55.340 --> 00:43:59.699
copyright libraries. Yeah, exactly. We do have
00:43:59.699 --> 00:44:02.440
systems in place. We just haven't really updated
00:44:02.440 --> 00:44:07.699
them maybe as much as we need to. Maybe all libraries...
00:44:07.699 --> 00:44:14.349
Sorry, carry on. Maybe this is something that
00:44:14.349 --> 00:44:18.389
has been talked about quite a lot, is that libraries
00:44:18.389 --> 00:44:22.989
can be a really good place for actual distributed
00:44:22.989 --> 00:44:27.090
archiving. So they have data centres of their
00:44:27.090 --> 00:44:30.590
own and it's a really good place to do that.
00:44:30.889 --> 00:44:33.030
And that's something that could be an actual
00:44:33.030 --> 00:44:35.829
proper project if it was supported by the government.
00:44:37.579 --> 00:44:39.619
Yeah, but it has to be supported not just by
00:44:39.619 --> 00:44:42.280
this government, but by governments down the
00:44:42.280 --> 00:44:46.460
echoing corridors of time. Yeah, that's not happening.
00:44:48.340 --> 00:44:54.000
I wonder if the price and speed of our capacity
00:44:54.000 --> 00:44:58.519
to store data more durably may provide something
00:44:58.519 --> 00:45:01.340
of a solution to this. I saw an article go past,
00:45:01.460 --> 00:45:04.280
which seems only vaguely related. poked in my
00:45:04.280 --> 00:45:06.639
brain. So yeah, I'll talk about it, which is
00:45:06.639 --> 00:45:11.079
like this new format of etching data onto glass
00:45:11.079 --> 00:45:14.800
and crystal sheets that they've found to, you
00:45:14.800 --> 00:45:18.280
know, it's a read once format, but once you've
00:45:18.280 --> 00:45:21.639
done it, it can hold an incredible amount of
00:45:21.639 --> 00:45:24.480
data. You mean a write once format. Yes, you're
00:45:24.480 --> 00:45:27.000
correct. I do mean a write once process, but
00:45:27.000 --> 00:45:29.760
once it's written, it's incredibly durable. Like
00:45:29.760 --> 00:45:31.500
they don't think it's going to degrade for 10
00:45:31.500 --> 00:45:35.239
,000 years. And if it turns out that the storage
00:45:35.239 --> 00:45:37.480
medium, which is something like glass, which
00:45:37.480 --> 00:45:41.519
is not very expensive, and then you can hold
00:45:41.519 --> 00:45:45.219
on to it, then what we've got is at least some
00:45:45.219 --> 00:45:49.840
capacity to do archives. So the stuff can be
00:45:49.840 --> 00:45:53.559
stored, and it shifts the question to, one, how
00:45:53.559 --> 00:45:56.119
do you find what you're looking for, and how
00:45:56.119 --> 00:46:02.719
fast can you get to it? And in lots of ways,
00:46:02.820 --> 00:46:05.739
I think that that may be like when we talk about
00:46:05.739 --> 00:46:08.559
the Internet Archive, you go to it, it's a page
00:46:08.559 --> 00:46:10.360
you can go to. It's actually pretty fast to do
00:46:10.360 --> 00:46:12.219
what you want to do for what you want to do.
00:46:12.440 --> 00:46:14.619
Like you go, I want to go to this website and
00:46:14.619 --> 00:46:17.579
then I want to see what it was like on the 14th
00:46:17.579 --> 00:46:22.820
of June in 2017 or something. That process is
00:46:22.820 --> 00:46:25.019
very fast. And so there's this question of like,
00:46:25.079 --> 00:46:27.889
how accessible is accessible? like if it turns
00:46:27.889 --> 00:46:29.590
out that somebody has this recorded somewhere
00:46:29.590 --> 00:46:32.090
like this this this relates to the distributed
00:46:32.090 --> 00:46:35.010
libraries just in the same way that we're like
00:46:35.010 --> 00:46:38.170
oh well hosmer branch doesn't have the current
00:46:38.170 --> 00:46:40.570
copy of dungeon crawler carl but we can order
00:46:40.570 --> 00:46:42.230
it from this other branch and it'll be here in
00:46:42.230 --> 00:46:44.550
three days like we just don't think about that
00:46:44.550 --> 00:46:47.469
in terms of internet time but that could be something
00:46:47.469 --> 00:46:51.710
that we're doing like there's there's ways it's
00:46:51.710 --> 00:46:53.349
just a question of how accessible is it and how
00:46:53.349 --> 00:46:57.469
easy is it that's the books are we um I was wondering
00:46:57.469 --> 00:46:59.750
about this. Are we looking at it the wrong way?
00:47:01.010 --> 00:47:05.650
I hear what you're saying, Ryan, and I do love
00:47:05.650 --> 00:47:07.750
the ability just to be able to go, I'm on a site,
00:47:07.889 --> 00:47:12.750
I want to go back. But I'm also, as someone who's
00:47:12.750 --> 00:47:20.789
using Amazon Glacier, I like the fact that it's
00:47:20.789 --> 00:47:24.289
archived somewhere. And maybe what... the problem
00:47:24.289 --> 00:47:28.170
is we're relying on an entity to archive it maybe
00:47:28.170 --> 00:47:30.610
what what we should do is have something where
00:47:30.610 --> 00:47:35.949
if you are creating this kind of thing um or
00:47:35.949 --> 00:47:38.769
creating something you know that is maybe like
00:47:38.769 --> 00:47:42.050
a new site it should be archived and it it means
00:47:42.050 --> 00:47:44.909
that that should also go into the national archive
00:47:44.909 --> 00:47:47.170
so that you're not relying on the national archive
00:47:47.170 --> 00:47:50.929
going and doing the the grabbing but you are
00:47:50.929 --> 00:47:54.809
as under uk law or the european law going you
00:47:54.809 --> 00:47:58.429
have to do this you know that would save them
00:47:58.429 --> 00:48:00.530
quite a bit of time and obviously that could
00:48:00.530 --> 00:48:04.969
also work for archive .org for example you know
00:48:05.739 --> 00:48:08.239
I'm getting dangerously close to you coming up
00:48:08.239 --> 00:48:10.880
with a Fediverse version of the Internet Archivian.
00:48:12.079 --> 00:48:17.380
I could talk about that if you want me to. Look,
00:48:18.719 --> 00:48:21.420
in the late 90s, the security engineer Ross Anderson
00:48:21.420 --> 00:48:24.139
proposed the Eternity Service. The paper is still
00:48:24.139 --> 00:48:29.639
on his website. you know it was an idea that
00:48:29.639 --> 00:48:32.539
was later i think the the basic ideas in that
00:48:32.539 --> 00:48:36.039
paper were later became free net and became oh
00:48:36.039 --> 00:48:39.820
yeah you know various other efforts to have distributed
00:48:39.820 --> 00:48:42.800
basically the idea was to distribute information
00:48:42.800 --> 00:48:45.900
so that it couldn't be censored and there's been
00:48:45.900 --> 00:48:52.119
various attempts at that i i just because i i
00:48:52.119 --> 00:48:54.719
i mean ultimately i just i think that relying
00:48:54.719 --> 00:49:00.929
on one person or one entity to do it, it's probably,
00:49:01.010 --> 00:49:05.250
we see the limits of that already. It's fragile,
00:49:05.469 --> 00:49:11.489
but at the same time, it's also like, this is
00:49:11.489 --> 00:49:13.530
always the problem with anything that's depending
00:49:13.530 --> 00:49:15.510
on people to do it via a volunteering mechanism.
00:49:15.869 --> 00:49:19.690
They have to want to do it. If you're going to
00:49:19.690 --> 00:49:21.070
ask people to do something they're not interested
00:49:21.070 --> 00:49:23.650
in doing that, you're going to have to move it
00:49:23.650 --> 00:49:25.449
to a paid model because you have to pay them.
00:49:26.699 --> 00:49:28.940
You know, one of the things I found interesting
00:49:28.940 --> 00:49:33.579
recently, not recently, a while ago, was when
00:49:33.579 --> 00:49:40.940
they re -found the GeoCities archive, and they
00:49:40.940 --> 00:49:46.019
distributed that on BitTorrent. And I'm not saying
00:49:46.019 --> 00:49:49.699
BitTorrent is a storage mechanism, but... It's
00:49:49.699 --> 00:49:52.039
a distribution mechanism. It makes it possible
00:49:52.039 --> 00:49:55.550
to distribute the storage, though. Yeah, I think
00:49:55.550 --> 00:49:57.570
there's something, I just think that there's
00:49:57.570 --> 00:50:00.469
stuff like that that is pretty useful, but that
00:50:00.469 --> 00:50:03.090
doesn't, but this goes back to what Ryan was
00:50:03.090 --> 00:50:06.210
saying about the, we're so used to, I go to archive
00:50:06.210 --> 00:50:08.449
.org and I just go back and I can go to the wave
00:50:08.449 --> 00:50:10.510
app machine and I can just go back to, and I
00:50:10.510 --> 00:50:12.170
think that maybe there is, there should be some
00:50:12.170 --> 00:50:16.150
friction because the headache of trying to maintain
00:50:16.150 --> 00:50:20.349
that kind of very easy mechanism is probably
00:50:20.349 --> 00:50:23.480
more painful than it's on. one of these disks
00:50:23.480 --> 00:50:26.900
somewhere, and you have to wait, or you have
00:50:26.900 --> 00:50:28.679
to download it. It'd be a lot easier if they
00:50:28.679 --> 00:50:29.900
just, like, one of the reasons why they could
00:50:29.900 --> 00:50:31.699
do that with the old GeoCities pages is because
00:50:31.699 --> 00:50:33.699
that was back before the day they made every
00:50:33.699 --> 00:50:35.760
web page a thousand times bigger because they
00:50:35.760 --> 00:50:37.960
just didn't know how to. Like, that's what enabled
00:50:37.960 --> 00:50:40.340
it. If only we were moving back into an age where
00:50:40.340 --> 00:50:42.099
we made simplified versions of all the pages
00:50:42.099 --> 00:50:43.800
in something like, say, Markdown, then we could
00:50:43.800 --> 00:50:51.070
do all that again. Or RSS. Yeah. I'm not going
00:50:51.070 --> 00:50:55.829
to start about XHTML. Markdown and RSS are not
00:50:55.829 --> 00:50:58.389
in competition. Markdown is a format. RSS is
00:50:58.389 --> 00:51:01.889
a syndication protocol. I know, but I'm not going
00:51:01.889 --> 00:51:04.110
to start. I'm not going to start. Just leave
00:51:04.110 --> 00:51:10.590
it alone. All right. Thank you for that, Ian
00:51:10.590 --> 00:51:13.889
and Wendy. And once again, our last point of
00:51:13.889 --> 00:51:16.889
the show before any other business is going to
00:51:16.889 --> 00:51:19.409
be Wendy with an article on surveillance pricing.
00:51:19.670 --> 00:51:22.570
Wendy, what's going on with that? Well, surveillance
00:51:22.570 --> 00:51:25.610
pricing is an interesting new term. You've heard
00:51:25.610 --> 00:51:28.210
of surveillance capitalism, which is the exploitation
00:51:28.210 --> 00:51:32.389
of advertising data to target us all. And surveillance
00:51:32.389 --> 00:51:35.989
pricing is the use of data to create personalized
00:51:35.989 --> 00:51:39.420
prices that you know, where, where some, some
00:51:39.420 --> 00:51:42.280
brilliant data crunching figures out exactly
00:51:42.280 --> 00:51:45.340
how much we are, would be able to able and willing
00:51:45.340 --> 00:51:48.179
to pay for something, you know, like you're trying
00:51:48.179 --> 00:51:51.519
to get a, it started with the airlines really,
00:51:51.639 --> 00:51:53.519
you know, they would love to be able to say,
00:51:53.579 --> 00:51:56.079
oh, you're trying to get to your, to, to, to
00:51:56.079 --> 00:51:59.199
your, see your dying mother before she, before
00:51:59.199 --> 00:52:02.739
she dies. Let me charge you $3 ,000 for this
00:52:02.739 --> 00:52:04.400
ticket. You know, you, you know, they would love
00:52:04.400 --> 00:52:09.300
to do that. And so, I think it's a couple of
00:52:09.300 --> 00:52:13.559
U .S. lawmakers are proposing a ban on surveillance
00:52:13.559 --> 00:52:17.719
pricing because apparently they've been in grocery
00:52:17.719 --> 00:52:20.500
chains. That's the specific it's specifically
00:52:20.500 --> 00:52:25.199
for the retail grocery sector, because now that
00:52:25.199 --> 00:52:27.780
you have electronic shelf tags that you can update
00:52:27.780 --> 00:52:30.780
in real time, you can use them to to change prices
00:52:30.780 --> 00:52:37.510
as people walk past. So this is a whole new bit
00:52:37.510 --> 00:52:45.250
of awfulness that has sprung up. Yeah, I mean,
00:52:45.250 --> 00:52:48.389
at the core of that is, oh, is it raining? Umbrellas
00:52:48.389 --> 00:52:53.090
are now 200 % more. Right. Or if you want to
00:52:53.090 --> 00:52:55.550
go extra evil, it's like, it's a bottle of water,
00:52:55.710 --> 00:52:59.409
but you're in the desert, buddy. Yeah, and it's
00:52:59.409 --> 00:53:01.469
not just that. It's also all the stuff they know
00:53:01.469 --> 00:53:06.230
about us, which is, of course, a lot. And traditionally,
00:53:06.610 --> 00:53:10.909
personalized pricing has not, I think, not been
00:53:10.909 --> 00:53:13.030
permitted in the United States, particularly
00:53:13.030 --> 00:53:16.050
among the airlines. They're supposed to publish
00:53:16.050 --> 00:53:18.829
a tariff and they are supposed to be able to,
00:53:18.869 --> 00:53:21.150
anybody is supposed to be able to read the tariff
00:53:21.150 --> 00:53:22.590
and say, this is what my ticket should cost.
00:53:22.710 --> 00:53:25.630
In practice, can any of us read airline tariffs
00:53:25.630 --> 00:53:27.050
and understand what our tickets should cost?
00:53:27.170 --> 00:53:30.579
No, we can't. which is what Edward Hasbrook writes
00:53:30.579 --> 00:53:33.440
about quite a lot at his Papers, Please blog.
00:53:34.480 --> 00:53:36.360
That's why everyone spends more than half a ton,
00:53:36.519 --> 00:53:40.920
yeah. But, you know, imagine this being a daily
00:53:40.920 --> 00:53:44.900
reality in grocery stores. Oh, you're gluten,
00:53:45.039 --> 00:53:48.300
you can't eat gluten. Well, let me just raise
00:53:48.300 --> 00:53:50.400
the price on this gluten -free product for you.
00:53:50.719 --> 00:53:52.840
You know, I mean, it's a horrible thought. Ah,
00:53:52.840 --> 00:53:54.539
if you're gluten intolerant, you probably deserve
00:53:54.539 --> 00:54:00.570
it. Whoa! I do like the fact that you can walk
00:54:00.570 --> 00:54:04.369
past a label and see it go up because it knows
00:54:04.369 --> 00:54:08.369
who you are. I did read that there's a study
00:54:08.369 --> 00:54:11.889
of these electronic shelf tags, and so far they
00:54:11.889 --> 00:54:14.610
have only found them used to discount prices,
00:54:14.789 --> 00:54:18.190
not to raise them. But you know that the capability
00:54:18.190 --> 00:54:22.639
is there. Actually, I would argue that... So
00:54:22.639 --> 00:54:25.079
reducing and increasing is the same thing. It's
00:54:25.079 --> 00:54:27.739
a mechanism to get you to buy stuff. The mechanism
00:54:27.739 --> 00:54:31.039
of the shelf tags isn't the one that, like, my
00:54:31.039 --> 00:54:32.820
brain's already shooting a thousand holes in
00:54:32.820 --> 00:54:34.380
that because who's to say someone else isn't
00:54:34.380 --> 00:54:35.860
looking at the tag at the same price? Well, exactly.
00:54:36.480 --> 00:54:39.280
You have to give it to them at that price. The
00:54:39.280 --> 00:54:41.320
place where this really starts to happen sneakily
00:54:41.320 --> 00:54:42.619
and the place that you have to be afraid of is
00:54:42.619 --> 00:54:45.329
when you go to the checkout counter. Yeah, I
00:54:45.329 --> 00:54:46.969
was going to say the same. I mean, I've seen
00:54:46.969 --> 00:54:50.550
that where, yeah, you literally, it's not actually
00:54:50.550 --> 00:54:52.409
got a price. Go to the checkout and it's suddenly,
00:54:52.610 --> 00:54:55.650
oh, it's this much. And you're like, and who's
00:54:55.650 --> 00:54:58.750
going to, you know, it takes a very strong willed
00:54:58.750 --> 00:55:02.989
person to go, hey, actually, I don't want this.
00:55:03.309 --> 00:55:05.949
I don't want it anymore. You know, people usually
00:55:05.949 --> 00:55:08.250
just go, oh, well, whatever. I've done that.
00:55:08.389 --> 00:55:11.400
I don't do it too often, but like Tesco is. Tesco
00:55:11.400 --> 00:55:16.260
is horrible about this. They have set up a complex
00:55:16.260 --> 00:55:20.820
system of pricing based on the size of the shop.
00:55:21.500 --> 00:55:24.579
So if you had a very, very big Tesco, you get
00:55:24.579 --> 00:55:26.480
the best prices. But as you go, because I live
00:55:26.480 --> 00:55:28.079
somewhere that's more urban and more close to
00:55:28.079 --> 00:55:31.639
that, my Tesco, and this is hilarious. So my
00:55:31.639 --> 00:55:39.400
Tesco used to be a Tesco Metro. Metro, yeah.
00:55:39.659 --> 00:55:43.099
And as far as I can tell, what happened was they
00:55:43.099 --> 00:55:45.159
had this extra section. It was not very big,
00:55:45.219 --> 00:55:46.900
and they did it by size. Like, if you have this
00:55:46.900 --> 00:55:48.900
much square feet, you're going to be this, this,
00:55:48.940 --> 00:55:50.840
or this. And every time the Tesco gets smaller,
00:55:51.019 --> 00:55:53.159
so there's, like, Mega Tesco, which is the biggest
00:55:53.159 --> 00:55:54.659
and the cheapest, then there's Tesco, which is
00:55:54.659 --> 00:55:56.760
smaller, Tesco Superstore, then you get Tesco
00:55:56.760 --> 00:56:02.260
Extra, then you get Tesco Metro, and then, like,
00:56:02.420 --> 00:56:04.760
because I have two Tescos I can walk to, one
00:56:04.760 --> 00:56:05.820
of which is bigger than the other one, and I
00:56:05.820 --> 00:56:08.690
can, like, I know. Because I'll be like, I don't
00:56:08.690 --> 00:56:10.489
feel like paying 20 pence extra for this thing.
00:56:10.530 --> 00:56:12.929
I'm going to take a longer walk. And then there's
00:56:12.929 --> 00:56:14.489
a Tesco Express down the road, and that gets
00:56:14.489 --> 00:56:17.550
even more expensive down there. But what they
00:56:17.550 --> 00:56:20.989
did for this Tesco is they went, oh, you know
00:56:20.989 --> 00:56:24.250
what? If we could just make this store a little
00:56:24.250 --> 00:56:26.989
bit smaller, we could charge extra for everything
00:56:26.989 --> 00:56:28.769
in here. And what they did is they put a goddamn
00:56:28.769 --> 00:56:32.250
Griggs in it. And that reduced the floor plan
00:56:32.250 --> 00:56:37.670
of the Tesco enough to justify that. And so this
00:56:37.670 --> 00:56:41.190
is a part of that, but I will go in it. But what
00:56:41.190 --> 00:56:43.969
happens because of this system is that sometimes
00:56:43.969 --> 00:56:48.550
they forget that some of their shops are supposed
00:56:48.550 --> 00:56:50.210
to be more expensive than others. And occasionally
00:56:50.210 --> 00:56:52.369
you will see things advertised for multiple prices
00:56:52.369 --> 00:56:55.190
because someone in one part of the organization
00:56:55.190 --> 00:56:58.030
will be like, let's make all the standees for
00:56:58.030 --> 00:57:01.989
the new thing, and they'll put the Tesco. You
00:57:01.989 --> 00:57:04.309
know, the Tesco mega Tesco prices on it, and
00:57:04.309 --> 00:57:06.170
then they'll send the standees to hold stuff
00:57:06.170 --> 00:57:08.650
out to the Tesco Express. And they'll be like,
00:57:08.650 --> 00:57:10.489
that's funny, because on the standee, it says
00:57:10.489 --> 00:57:12.150
this can of Pringles supposed to be one pound
00:57:12.150 --> 00:57:14.909
50. But when I take it to the counter, it's going
00:57:14.909 --> 00:57:17.349
to be two pounds. And when that happens, I'll
00:57:17.349 --> 00:57:19.289
be like, I'll just I'll go grab the standee.
00:57:19.289 --> 00:57:23.130
I'll be like, it's right here, guys. Great gesture.
00:57:23.469 --> 00:57:28.710
Gesturing for podcast is a great plan. I think
00:57:28.710 --> 00:57:30.349
I've had that problem in Asda where I've seen
00:57:30.349 --> 00:57:33.010
the price, but actually it was the member price.
00:57:33.289 --> 00:57:35.010
Of course, when I go out to check it out, I go,
00:57:35.050 --> 00:57:37.630
oh, you're not that nice. Which is kind of annoying.
00:57:37.989 --> 00:57:41.250
Even the member prices go up at those places.
00:57:41.369 --> 00:57:44.469
I can confirm this. I am within walking distance
00:57:44.469 --> 00:57:47.389
of four different Tescos of three different sizes,
00:57:47.530 --> 00:57:50.809
and it enrages me. It made me stop using them
00:57:50.809 --> 00:57:53.010
from those things. You're making me very happy
00:57:53.010 --> 00:57:55.690
that I do so much shopping at the shop around
00:57:55.690 --> 00:57:59.559
the corner. I know it charges more. I don't care.
00:57:59.679 --> 00:58:03.460
I like the people. And in a pandemic when no
00:58:03.460 --> 00:58:05.460
one can get toilet paper, I know they will find
00:58:05.460 --> 00:58:09.960
some for me. Yeah, that's what matters. Yeah,
00:58:10.039 --> 00:58:15.179
I mean, I do. I have the same problem as Ryan.
00:58:15.500 --> 00:58:18.639
So there's, I have a Sainsbury's right on the
00:58:18.639 --> 00:58:22.179
same block as my flat. And so I know it's going
00:58:22.179 --> 00:58:25.019
to be more money. And sometimes I'm like, you
00:58:25.019 --> 00:58:28.559
know what? I know that and I won't go. So there's
00:58:28.559 --> 00:58:30.820
an Aldi around the corner. I'll just go there
00:58:30.820 --> 00:58:33.780
for eggs because the price of six eggs in Sainsbury's
00:58:33.780 --> 00:58:37.639
local is the same price as six eggs in Aldi.
00:58:38.500 --> 00:58:40.500
But it just means I have to walk a bit further.
00:58:41.019 --> 00:58:44.880
I do think that I remember reading something
00:58:44.880 --> 00:58:51.800
about the e -paper displays. Yeah, I think like
00:58:51.800 --> 00:58:55.500
what David was saying. They're used for discounting.
00:58:56.119 --> 00:58:59.659
I don't think under EU law they can actually
00:58:59.659 --> 00:59:03.679
use it for personalisation. The thing that I'm
00:59:03.679 --> 00:59:06.059
interested in, and basically what you're saying,
00:59:06.139 --> 00:59:09.380
Wendy, is that I think when you go to other countries
00:59:09.380 --> 00:59:13.460
outside of the EU, they may not do it yet, but
00:59:13.460 --> 00:59:16.099
they're going to start doing it. And I do think
00:59:16.099 --> 00:59:18.960
that there will be... I do think that the problem
00:59:18.960 --> 00:59:21.079
with what Ryan's described, where you've got
00:59:21.079 --> 00:59:24.289
multiple people... walking past something, I
00:59:24.289 --> 00:59:25.809
think they will just get rid of all the pricing
00:59:25.809 --> 00:59:29.449
and it will be done at the checkout. I think
00:59:29.449 --> 00:59:31.949
that's where they're going to go. Well, you have
00:59:31.949 --> 00:59:34.090
to have an app to see what the price is. Yeah.
00:59:36.070 --> 00:59:39.909
There's a few places I've been where I had to
00:59:39.909 --> 00:59:43.489
download the menu and that's something also where
00:59:43.489 --> 00:59:46.650
you could change the... Not just download the
00:59:46.650 --> 00:59:49.769
menu, it's like the menu is dynamically... generated
00:59:49.769 --> 00:59:53.250
there is an easy way to do it there is an entire
00:59:53.250 --> 00:59:56.929
subset of bargain hunters who are specializing
00:59:56.929 --> 01:00:02.409
in using these handheld pay and scan apps that
01:00:02.409 --> 01:00:04.530
some places like sainsbury's got them some other
01:00:04.530 --> 01:00:06.670
places have them they usually have them in the
01:00:06.670 --> 01:00:08.590
bigger stores but the thing that they've discovered
01:00:08.590 --> 01:00:12.230
is that those things get updated much faster
01:00:12.230 --> 01:00:15.849
with new prices than the floor ones and if you
01:00:15.849 --> 01:00:17.769
are going through and scanning it with your app
01:00:18.280 --> 01:00:20.980
what you can do and so they'll go into the store
01:00:20.980 --> 01:00:23.760
and they'll scan it with the thing go oh that's
01:00:23.760 --> 01:00:26.099
interesting this thing is 50 pence cheaper on
01:00:26.099 --> 01:00:28.940
my scanner that's great and they'll take it and
01:00:28.940 --> 01:00:31.059
they can walk out because that's what the scanner
01:00:31.059 --> 01:00:33.619
says so it's yeah it's a pound on the scanner
01:00:33.619 --> 01:00:36.239
pound 50 on the shelf and they'll like i guess
01:00:36.239 --> 01:00:37.940
they haven't figured out how to make these scanners
01:00:37.940 --> 01:00:42.639
work for the different size stores yet All right,
01:00:42.800 --> 01:00:46.360
and last stop on Tech Grumps, we've got Any Other
01:00:46.360 --> 01:00:48.360
Business. And for Any Other Business, we've got
01:00:48.360 --> 01:00:50.460
two items from Ian. Ian, what you got for us?
01:00:51.070 --> 01:00:55.670
So I went to an event in London called Protocols
01:00:55.670 --> 01:00:58.690
for Publishers. They did a really nice summary
01:00:58.690 --> 01:01:01.889
of what the event was about. If you are interested
01:01:01.889 --> 01:01:07.670
in the Fediverse or Blue Sky or AT Proto, I do
01:01:07.670 --> 01:01:10.789
recommend having a little read. It was a really
01:01:10.789 --> 01:01:15.500
good event. And then the second one is... I wanted
01:01:15.500 --> 01:01:18.760
to give a shout out to someone that's mentioned
01:01:18.760 --> 01:01:25.940
on this podcast quite a lot, Andy, because they
01:01:25.940 --> 01:01:32.840
describe the hassle or the headache of Fediverse
01:01:32.840 --> 01:01:37.119
share this link, which we're used to on stuff
01:01:37.119 --> 01:01:42.150
like UX and... wherever else instagram and blah
01:01:42.150 --> 01:01:45.710
blah but there's a real unique challenge because
01:01:45.710 --> 01:01:49.550
it's distributed and it's it just it's a they
01:01:49.550 --> 01:01:52.050
describe it and i think it's well worth reading
01:01:52.050 --> 01:01:55.389
because it they went through the hard work and
01:01:55.389 --> 01:01:58.889
they did it the right way and it really really
01:01:58.889 --> 01:02:02.650
makes people It really helps people to understand
01:02:02.650 --> 01:02:06.389
why you can't just have these features and how
01:02:06.389 --> 01:02:09.590
the features currently work on platforms like
01:02:09.590 --> 01:02:14.789
X and Instagram when it comes to privacy. So
01:02:14.789 --> 01:02:18.750
I recommend it. Okay, and that's it for another
01:02:18.750 --> 01:02:20.969
round of Tech Grumps. The only thing left is
01:02:20.969 --> 01:02:22.929
to say goodbye. My name is Ryan. It's been great
01:02:22.929 --> 01:02:26.170
talking to you. Goodbye. Next up is David. Bye
01:02:26.170 --> 01:02:31.980
from me. Ian. It's bye from me. And Wendy. Bye
01:02:31.980 --> 01:02:36.820
from me. All right. Thanks from all of us. See
01:02:36.820 --> 01:02:42.619
you next month. Bye. Bye. And don't forget, if
01:02:42.619 --> 01:02:44.659
you'd like to join us or follow us on Tech Grumps,
01:02:44.719 --> 01:02:46.599
just look for the hashtag Tech Grumps on any
01:02:46.599 --> 01:02:48.199
of your favorite federated platforms.