Techgrumps 3.44 - My Nuclear Bunker is too woke

WEBVTT
00:00:01.259 --> 00:00:05.740
Welcome to Tech Grumps 3 .44. Today is Sunday,
00:00:06.059 --> 00:00:09.699
the 6th of September, and we are here to speak
00:00:09.699 --> 00:00:12.199
with you about things and tech and grumpiness.
00:00:13.220 --> 00:00:19.679
Joining us tonight, we have Ian Forrester. Hello,
00:00:19.739 --> 00:00:22.960
I am Ian, and I am Cubic Garden Everywhere, and
00:00:22.960 --> 00:00:27.399
I am calling from Manchester, UK, not Alabama.
00:00:28.649 --> 00:00:33.929
We have Wendy. Hi, I'm Wendy Grossman. My website
00:00:33.929 --> 00:00:36.789
is pelicancrossing .net, and I'm based in London.
00:00:37.509 --> 00:00:41.429
We have David. Hello, I'm David Eastman, also
00:00:41.429 --> 00:00:44.810
based in London. Eastman on blue sky and other
00:00:44.810 --> 00:00:47.710
things. I won't be talking too much tonight,
00:00:47.829 --> 00:00:50.130
as I've been on holiday for two weeks, so pass
00:00:50.130 --> 00:00:53.030
it on. I don't believe you for a second. Joining
00:00:53.030 --> 00:00:57.649
us tonight is Kim. Hello, I'm Kim. I live in
00:00:57.649 --> 00:00:59.689
Manchester, also United Kingdom and Alabama.
00:01:00.689 --> 00:01:02.590
I run a little studio called Geeks for Social
00:01:02.590 --> 00:01:05.409
Change, which you can find at gfsc .community.
00:01:05.409 --> 00:01:07.390
And thanks for having me. Oh, and I use she and
00:01:07.390 --> 00:01:09.510
they pronouns too. Hello. Thank you very much.
00:01:09.730 --> 00:01:12.969
And I, of course, am Ryan Alexander, R .N. Alexander,
00:01:13.129 --> 00:01:20.090
just about everywhere. Find me if you like. All
00:01:20.090 --> 00:01:22.590
right. And as usual at Tech Grumps, we like to
00:01:22.590 --> 00:01:25.310
cover all things AI all at once because, oh,
00:01:25.329 --> 00:01:27.870
God, how else would we do it? We have a collection
00:01:27.870 --> 00:01:30.170
of fine stories for you this evening, some even
00:01:30.170 --> 00:01:33.230
brought by our wonderful guest, Kim. So I'm just
00:01:33.230 --> 00:01:35.069
going to read them off and then we'll get into
00:01:35.069 --> 00:01:37.269
talking about them. The stories we have tonight
00:01:37.269 --> 00:01:40.230
is a story about AI may make software more secure
00:01:40.230 --> 00:01:42.730
about government hacking on TechCrunch. We have
00:01:42.730 --> 00:01:46.549
a story about NVIDIA buying Hugging Face on Ars
00:01:46.549 --> 00:01:50.250
Technica. We have a reference to... Trump, who's
00:01:50.250 --> 00:01:53.890
opposing people who oppose data centers, calling
00:01:53.890 --> 00:01:57.989
them backwards and poor. We have the Dyson $500
00:01:57.989 --> 00:02:02.230
smart toothbrush with AI, live camera, and water
00:02:02.230 --> 00:02:06.019
floss. We have a story about AI almost sending
00:02:06.019 --> 00:02:09.479
a hazmat train barreling towards Katas. And we
00:02:09.479 --> 00:02:12.099
have a story about New York school systems to
00:02:12.099 --> 00:02:15.340
ban generative AI for elementary and middle school
00:02:15.340 --> 00:02:18.580
students. Yay, Mamdani. And last but certainly
00:02:18.580 --> 00:02:21.460
not least, a story on The Guardian about a fake
00:02:21.460 --> 00:02:25.080
think tank from Israel that has been attempting
00:02:25.080 --> 00:02:29.080
to... Bias AI with massive amounts of propaganda.
00:02:29.520 --> 00:02:32.259
Where do we feel like starting with all of this
00:02:32.259 --> 00:02:37.659
tonight? I would definitely like to talk a little
00:02:37.659 --> 00:02:41.460
bit about this. These people who are backwards
00:02:41.460 --> 00:02:43.740
because they want to sell it for their rights.
00:02:43.800 --> 00:02:45.979
He didn't say they are backwards. He said they
00:02:45.979 --> 00:02:51.219
want to be backwards and poor. The point is they
00:02:51.219 --> 00:02:54.039
oppose data centers and therefore they want to
00:02:54.039 --> 00:02:57.789
be backwards and poor. I mean, you are right.
00:02:58.430 --> 00:03:02.129
I just have this feeling that it was very implied
00:03:02.129 --> 00:03:05.330
that you are all backwards and you want to be
00:03:05.330 --> 00:03:13.169
backwards. That's my feeling. Go on. No, we're
00:03:13.169 --> 00:03:19.889
all NPCs to billionaires. You know that. That's
00:03:19.889 --> 00:03:22.930
a good way of putting it. Yeah, I just, I felt
00:03:22.930 --> 00:03:27.810
when I heard this. I just, I mean, obviously
00:03:27.810 --> 00:03:31.789
I'm not American, but I also am very aware that
00:03:31.789 --> 00:03:34.770
these are the kind of things that spread, you
00:03:34.770 --> 00:03:38.469
know, this notion of, well, obviously, you know,
00:03:38.509 --> 00:03:40.930
why would you want to stand in the way of progress?
00:03:41.310 --> 00:03:44.270
You know, screw your rights, screw, you know,
00:03:44.270 --> 00:03:47.009
like actually having a life. Well, I'm going
00:03:47.009 --> 00:03:50.409
to bring this up again, that in the 1960s, the
00:03:50.409 --> 00:03:55.909
inevitability narrative was cars. And yet, a
00:03:55.909 --> 00:03:58.590
lot of the roads they planned to build in the
00:03:58.590 --> 00:04:01.090
end did not get built because people did oppose
00:04:01.090 --> 00:04:07.849
it. That's a very good analogy. We treat opposition
00:04:07.849 --> 00:04:13.210
as a binary. It is a spectrum. The funny question
00:04:13.210 --> 00:04:15.189
I have about AI data centers that I don't really
00:04:15.189 --> 00:04:19.620
know the answer to is... Because everything's,
00:04:19.620 --> 00:04:21.079
I feel like we don't really have a tech sector
00:04:21.079 --> 00:04:23.199
anymore. We have an AI sector. So are they AI
00:04:23.199 --> 00:04:25.699
data centers or are they just data centers because
00:04:25.699 --> 00:04:28.779
our general computing needs are massively spiking
00:04:28.779 --> 00:04:30.939
through the roof and AI is kind of the excuse
00:04:30.939 --> 00:04:33.560
to build them? Like, I mean, I think we all know
00:04:33.560 --> 00:04:36.259
how big the computation is of AI, but I'm just
00:04:36.259 --> 00:04:39.560
really curious, like to what degree it's more
00:04:39.560 --> 00:04:42.860
just like spiraling need for this. As I understand
00:04:42.860 --> 00:04:45.319
it, the data centers they're building now for
00:04:45.319 --> 00:04:48.779
AI are on a vastly bigger scale than anything
00:04:48.779 --> 00:04:51.399
we've had before. I mean, there was one that
00:04:51.399 --> 00:04:54.540
was canceled in Virginia that was described in
00:04:54.540 --> 00:04:58.319
the news pieces as 144 Walmart super centers
00:04:58.319 --> 00:05:03.160
stacked laterally. I mean, that's massive. That's,
00:05:03.160 --> 00:05:06.000
you know, that's not, that's not just a data
00:05:06.000 --> 00:05:09.990
center. It's, you know. So the scale difference
00:05:09.990 --> 00:05:12.689
is they used to say, you know, how many servers
00:05:12.689 --> 00:05:15.949
are in the data. Now they say, what is the power
00:05:15.949 --> 00:05:18.149
consumption? In fact, that is how you measure
00:05:18.149 --> 00:05:22.910
them. Yeah, that's true. What measurable percentage
00:05:22.910 --> 00:05:25.750
of the available power to the entire part of
00:05:25.750 --> 00:05:28.810
this country will we need? And if it is not sufficient,
00:05:28.990 --> 00:05:30.810
do you mind if we just plop down or restart a
00:05:30.810 --> 00:05:32.529
couple of nuclear reactors right next to you?
00:05:33.430 --> 00:05:36.029
Right. And they're really, as I understand it
00:05:36.029 --> 00:05:38.149
also, they're really unpleasant neighbors. They're
00:05:38.149 --> 00:05:41.250
noisy and they don't have very much staff. So
00:05:41.250 --> 00:05:43.850
there's this kind of forbidding giant building
00:05:43.850 --> 00:05:48.110
in your midst that is, you know, just has no
00:05:48.110 --> 00:05:50.550
interaction with the neighborhood whatsoever.
00:05:50.930 --> 00:05:52.410
I mean, that's what they're complaining about
00:05:52.410 --> 00:05:54.509
in Brick Lane in part. I mean, it's also that
00:05:54.509 --> 00:05:57.050
they wanted the old brewery to be converted for
00:05:57.050 --> 00:06:00.149
affordable housing, not for a data center to
00:06:00.149 --> 00:06:04.860
enable. uh you know london city traders to to
00:06:04.860 --> 00:06:09.639
trade nanoseconds faster i mean they've completely
00:06:09.639 --> 00:06:11.720
given up the narrow i've never seen a narrow
00:06:11.720 --> 00:06:14.279
it's odd because anytime previously to this kind
00:06:14.279 --> 00:06:17.740
of land grab scenario one of the four horsemen
00:06:17.740 --> 00:06:22.220
of the capitalist apocalypse is but it will create
00:06:22.220 --> 00:06:25.089
jobs and that horseman seems to be profoundly
00:06:25.089 --> 00:06:27.509
missing from all of these discussions which is
00:06:27.509 --> 00:06:30.110
weird by its absence because they always try
00:06:30.110 --> 00:06:31.870
to pull at least that part of the wool over as
00:06:31.870 --> 00:06:35.350
like it will drive they do say it it's just nobody
00:06:35.350 --> 00:06:37.209
believes them because they know that they aren't
00:06:37.209 --> 00:06:38.750
going to have very much staff once they've been
00:06:38.750 --> 00:06:41.329
built yeah this is like a long question here
00:06:41.329 --> 00:06:43.769
sorry i'd like so okay so they need to be huge
00:06:43.769 --> 00:06:45.750
why do they also need to be in residential areas
00:06:45.750 --> 00:06:52.610
do people know why because But, for example,
00:06:52.730 --> 00:06:54.509
the one that was canceled in Virginia was going
00:06:54.509 --> 00:06:57.250
to be right out the side of a Civil War battlefield
00:06:57.250 --> 00:07:00.930
that is of great significance to the people who
00:07:00.930 --> 00:07:04.670
live there. Other places where they're going
00:07:04.670 --> 00:07:07.750
to sit in the middle of particularly valued landscapes
00:07:07.750 --> 00:07:12.389
or, you know, or quite often there's a bunch
00:07:12.389 --> 00:07:14.670
being built in places where there's just not
00:07:14.670 --> 00:07:19.920
enough water. There's a factor of available infrastructure,
00:07:20.199 --> 00:07:22.180
not just in terms of can they get the high power
00:07:22.180 --> 00:07:23.939
stuff there without having to build a lot of
00:07:23.939 --> 00:07:26.040
it themselves. And this is important. Once again,
00:07:26.079 --> 00:07:29.560
it's a marginal thing where it's like they could
00:07:29.560 --> 00:07:31.459
build this in the middle of nowhere, but they
00:07:31.459 --> 00:07:33.279
don't want to have to pay for all of the infrastructure
00:07:33.279 --> 00:07:34.839
that it would take for them to get it in the
00:07:34.839 --> 00:07:37.100
middle of nowhere. Whereas they know, and this
00:07:37.100 --> 00:07:38.579
is a big part they don't say, is that it's so
00:07:38.579 --> 00:07:40.040
much cheaper for us to build this right next
00:07:40.040 --> 00:07:41.579
to you because you've already got the pipes and
00:07:41.579 --> 00:07:45.250
the power going in. Also, fuck them poor. It's
00:07:45.250 --> 00:07:47.370
mixed zoning. A lot of the time they purposely
00:07:47.370 --> 00:07:49.769
want mixed zoning precisely, as Ryan said, because
00:07:49.769 --> 00:07:52.970
the facilities are already there. Putting something
00:07:52.970 --> 00:07:55.050
out in the middle of nowhere is extremely expensive.
00:07:55.829 --> 00:07:58.930
And they don't want that expense. To bring it
00:07:58.930 --> 00:08:00.529
back to the article, presumably this will make
00:08:00.529 --> 00:08:04.610
the areas it's in poorer. It's literally the
00:08:04.610 --> 00:08:06.990
opposite of what Trump said. But the weird thing
00:08:06.990 --> 00:08:08.529
is, it's almost if they did the opposite, then
00:08:08.529 --> 00:08:10.470
they could sort of be making an argument for
00:08:10.470 --> 00:08:13.220
new. company towns of some description but they're
00:08:13.220 --> 00:08:17.860
obviously not doing that either we're not following
00:08:17.860 --> 00:08:20.319
any logic i mean the old thing is oh you'll get
00:08:20.319 --> 00:08:22.560
a new warm swimming pool well i mean afterwards
00:08:22.560 --> 00:08:24.779
if you've got 124 asters i'm never going to find
00:08:24.779 --> 00:08:27.100
your swimming pool so the ironic thing to my
00:08:27.100 --> 00:08:29.220
mind about the loss of the company town narrative
00:08:29.220 --> 00:08:33.759
is that it's not it we've lost it because they
00:08:33.759 --> 00:08:36.779
don't even want to try to lie that much they
00:08:36.779 --> 00:08:39.009
don't even want like they've managed to divorce
00:08:39.009 --> 00:08:42.370
the idea of responsibility of employment so far
00:08:42.370 --> 00:08:45.950
so that it's like, why would we take care of
00:08:45.950 --> 00:08:49.190
you? This is the gig economy. You show up and
00:08:49.190 --> 00:08:51.590
do exactly as much work as we want. Show up on
00:08:51.590 --> 00:08:53.250
your own dime, by the way. Show up and do exactly
00:08:53.250 --> 00:08:55.049
as much work as we want, and then you piss off.
00:08:55.470 --> 00:08:57.850
And if you die after that, that's great. If you
00:08:57.850 --> 00:09:00.450
die before then, we might need to charge you
00:09:00.450 --> 00:09:01.750
a fee because you were supposed to be getting
00:09:01.750 --> 00:09:04.509
work done for us. So where are we heading to?
00:09:04.590 --> 00:09:07.350
We're turning the server off and on again. Pretty
00:09:07.350 --> 00:09:11.039
much, yeah. I want to talk about the toothbrush.
00:09:11.639 --> 00:09:14.600
Oh, yeah. Yes. The toothbrush is great. I saw
00:09:14.600 --> 00:09:16.399
the story and I thought it was hilarious. And
00:09:16.399 --> 00:09:18.840
as it happened, I saw my dentist the next day
00:09:18.840 --> 00:09:21.100
and I asked him, I asked him about it. And he
00:09:21.100 --> 00:09:23.639
basically said, I tell my patients not to buy
00:09:23.639 --> 00:09:28.460
them because, you know, I mean, a $500 toothbrush
00:09:28.460 --> 00:09:31.100
to do things you can do for, you know, way less
00:09:31.100 --> 00:09:36.639
than that. I can just imagine the dentist, what,
00:09:36.720 --> 00:09:39.159
like. They're not saying it's like, oh, so you
00:09:39.159 --> 00:09:43.980
can spend $500 to never floss. You can never
00:09:43.980 --> 00:09:46.460
floss for free. Trust me. I'm a dentist. I know
00:09:46.460 --> 00:09:51.179
this. Mind you, I have a water flosser. Okay.
00:09:51.440 --> 00:09:54.440
I have a Sonicare toothbrush, and those are great.
00:09:55.179 --> 00:09:58.019
What is great about them? What is great about
00:09:58.019 --> 00:10:04.960
a toothbrush? Well, okay. So, I mean, do you?
00:10:05.859 --> 00:10:07.840
I'm going to need you to be more specific about
00:10:07.840 --> 00:10:08.960
what's great about a toothbrush, because part
00:10:08.960 --> 00:10:10.679
of me just wants to be like having clean teeth,
00:10:10.840 --> 00:10:15.179
but I think it's great. What makes you say the
00:10:15.179 --> 00:10:17.940
Sonic toothbrush is great? Oh, because it literally
00:10:17.940 --> 00:10:22.120
vibrates at 30 ,000 things a minute or a second
00:10:22.120 --> 00:10:23.779
or something, and it does make a difference in
00:10:23.779 --> 00:10:25.120
how clean it gets my teeth, particularly in getting
00:10:25.120 --> 00:10:26.779
the plaque off. I like that, but it doesn't need
00:10:26.779 --> 00:10:29.960
a camera to do it. Does it show up in less dental
00:10:29.960 --> 00:10:34.740
work? Yeah, I think it does. When I'm using it
00:10:34.740 --> 00:10:37.360
regularly, I come in, there's noticeably less
00:10:37.360 --> 00:10:42.799
scraping that occurs in my checkups because it
00:10:42.799 --> 00:10:45.240
does a good job of taking it off. It's very satisfying.
00:10:45.419 --> 00:10:49.159
You can feel it. You finally get over your stuff
00:10:49.159 --> 00:10:50.779
and you go to the dentist and then you get done.
00:10:50.899 --> 00:10:51.919
You get a dentist when you get done. You're like,
00:10:51.980 --> 00:10:55.019
hmm, the inside of my mouth has a luxury car
00:10:55.019 --> 00:10:58.500
feel for a short period of time. It's clean and
00:10:58.500 --> 00:11:03.919
slippery. Go ahead. Can I say... I only just
00:11:03.919 --> 00:11:07.580
recently bought an electric toothbrush. I've
00:11:07.580 --> 00:11:10.919
been visiting my dentist regularly. I've never,
00:11:11.000 --> 00:11:14.740
ever had anything done to my teeth, ever. And
00:11:14.740 --> 00:11:19.019
I've got to a kind of grand age. And that's just
00:11:19.019 --> 00:11:22.720
been brushing with a standard toothbrush. Yeah.
00:11:23.120 --> 00:11:26.139
Did you grow up in a place where they put fluoride
00:11:26.139 --> 00:11:29.659
in the water, perchance? Well, I guess Bristol
00:11:29.659 --> 00:11:33.080
may do. I don't know. Actually, no. No? Okay.
00:11:33.240 --> 00:11:36.299
I don't know. David? I don't know. I mean, they
00:11:36.299 --> 00:11:39.179
did do some fluoridation, but I think they've
00:11:39.179 --> 00:11:41.700
stopped that. Fluoridation really helps. Some
00:11:41.700 --> 00:11:45.379
of it's genetic. Some of it is how much sugar
00:11:45.379 --> 00:11:47.659
and crap you eat and how rigorous you are at
00:11:47.659 --> 00:11:50.600
brushing and flossing. There's a whole mix of
00:11:50.600 --> 00:11:53.860
things that go into whether you have good or
00:11:53.860 --> 00:11:57.279
bad teeth, you know. Yeah. It's real crap, too.
00:11:57.960 --> 00:12:00.600
We're not a dental podcast, so carry on. But
00:12:00.600 --> 00:12:03.419
I really don't think that, I think that $500
00:12:03.419 --> 00:12:06.659
or whatever it is for a supposedly smart toothbrush
00:12:06.659 --> 00:12:08.879
is just ridiculous. Honestly, at this point,
00:12:08.919 --> 00:12:12.679
every time I see AI associated with any device
00:12:12.679 --> 00:12:15.440
that has any recording of any type ability possible,
00:12:15.600 --> 00:12:17.679
my first thought is like, they want to train
00:12:17.679 --> 00:12:20.879
it on photos of my teeth. Well, they do have
00:12:20.879 --> 00:12:22.820
a camera in this toothbrush. I mean, that's what
00:12:22.820 --> 00:12:25.240
we want too. That is my immediate reaction to
00:12:25.240 --> 00:12:27.279
seeing anything that is being sold with a camera.
00:12:27.279 --> 00:12:30.879
Why would I want? Why would I want to stare at
00:12:30.879 --> 00:12:33.000
pictures of my mouth while I was brushing my
00:12:33.000 --> 00:12:35.899
teeth? The AI is looking at your mouth. No, no,
00:12:35.899 --> 00:12:38.299
no. They want to make it look cool to you so
00:12:38.299 --> 00:12:40.220
you don't think that what they really want is
00:12:40.220 --> 00:12:46.019
yet another source of undigested data. Because
00:12:46.019 --> 00:12:47.259
that's something that's going to be a problem
00:12:47.259 --> 00:12:50.419
now. It's like we're starting to get the AI equivalent
00:12:50.419 --> 00:12:52.220
of mad cow disease where they're feeding it its
00:12:52.220 --> 00:12:53.899
own data and then it starts to degrade and it
00:12:53.899 --> 00:12:56.440
goes bad. The term you're looking for is model
00:12:56.440 --> 00:13:00.320
collapse. Yeah, model collapse. yeah which is
00:13:00.320 --> 00:13:03.000
funny because it has the same initials the same
00:13:03.000 --> 00:13:06.080
acronym as mad cow is there something weird here
00:13:06.080 --> 00:13:08.399
does this seem quite off -brand for dyson to
00:13:08.399 --> 00:13:10.720
me because i feel like bryson dyson made their
00:13:10.720 --> 00:13:13.659
name by doing like quite smart engineering right
00:13:13.659 --> 00:13:15.220
like they were famous for a different kind of
00:13:15.220 --> 00:13:16.639
hoover design that was kind of shipped they look
00:13:16.639 --> 00:13:18.919
cool and they buy the air wraps and the hair
00:13:18.919 --> 00:13:21.220
dryers and they're all based on this kind of
00:13:21.220 --> 00:13:25.049
like idea of like smart modern engineering this
00:13:25.049 --> 00:13:27.350
just sounds like a load of iot crap from five
00:13:27.350 --> 00:13:29.350
years ago the load of stuff will glue together
00:13:29.350 --> 00:13:34.309
you know actually the main reason not to buy
00:13:34.309 --> 00:13:37.519
anything that dyson makes is that I am convinced
00:13:37.519 --> 00:13:39.679
that all the people who work at Dyson are deaf
00:13:39.679 --> 00:13:42.779
because every device I've ever come across of
00:13:42.779 --> 00:13:45.600
theirs is incredibly loud. I mean, those hand
00:13:45.600 --> 00:13:48.980
dryers in bathrooms ought to be shot, taken out
00:13:48.980 --> 00:13:51.259
and shot. Oh, I love those things. If that's
00:13:51.259 --> 00:13:54.120
true, then everyone who was fired from Dyson
00:13:54.120 --> 00:13:57.340
went directly to work for Ninja because those
00:13:57.340 --> 00:14:00.279
household appliances are like having a Harrier
00:14:00.279 --> 00:14:02.539
jet try to take off from your kitchen counter.
00:14:04.240 --> 00:14:06.940
I was going to say the loudest thing I've ever
00:14:06.940 --> 00:14:11.019
experienced in any kitchen was my now no longer
00:14:11.019 --> 00:14:14.299
working Ninja Blender system. The first time
00:14:14.299 --> 00:14:16.120
I turned it on, I actually thought there was
00:14:16.120 --> 00:14:19.500
something wrong. Wow. I was just going to add,
00:14:19.720 --> 00:14:24.159
I think there needs to, I'm sure there probably
00:14:24.159 --> 00:14:27.379
is some kind of little YouTube series on the
00:14:27.379 --> 00:14:31.779
downfall of Dyson, Kim. There's definitely, they
00:14:31.779 --> 00:14:34.690
started off well. I do know someone who used
00:14:34.690 --> 00:14:38.110
to work there and things changed. It feels like
00:14:38.110 --> 00:14:39.450
they had a, it feels like they had a catching
00:14:39.450 --> 00:14:41.049
lightning in a bottle moment. I've been chasing
00:14:41.049 --> 00:14:44.990
it ever since. Yeah. Do we have any thoughts
00:14:44.990 --> 00:14:48.149
about the hugging face purchase? I was about
00:14:48.149 --> 00:14:51.230
to ask the same. Yeah. I mean, it makes sense
00:14:51.230 --> 00:14:54.769
from an perspective. I mean, half the, like most
00:14:54.769 --> 00:14:58.750
of those model, like, okay. So I see the future
00:14:58.750 --> 00:15:01.330
of running these models. They're mostly going
00:15:01.330 --> 00:15:04.720
to be running on. you know, NVIDIA style hardware.
00:15:04.879 --> 00:15:06.740
I mean, the graphics cards are what they run
00:15:06.740 --> 00:15:10.340
these models on and therefore owning, having
00:15:10.340 --> 00:15:12.639
ownership of the biggest repository of them.
00:15:12.720 --> 00:15:14.620
A lot of these, which will be, you know, I mean,
00:15:14.620 --> 00:15:17.840
that makes a lot of sense to me. I don't know
00:15:17.840 --> 00:15:20.519
if it's a good or a bad thing. I think it might
00:15:20.519 --> 00:15:26.980
be, I could see NVIDIA trying to, they could,
00:15:27.019 --> 00:15:28.980
they could through, they could through encouraging
00:15:28.980 --> 00:15:33.120
the adaption of models. to running on nvidia
00:15:33.120 --> 00:15:34.539
hardware is already happening they could try
00:15:34.539 --> 00:15:37.200
to like increase lock into the hardware in case
00:15:37.200 --> 00:15:39.379
there's somebody with an nvidia killer coming
00:15:39.379 --> 00:15:41.899
up on the market hardware wise i mean no they
00:15:41.899 --> 00:15:44.200
don't do that their cuda system is open no i
00:15:44.200 --> 00:15:46.240
think it's just the like microsoft bought github
00:15:46.240 --> 00:15:48.480
because that's where the code is like those those
00:15:48.480 --> 00:15:51.159
bank robbers you know they're robbed backs because
00:15:51.159 --> 00:15:52.639
that's where the money is i don't think it's
00:15:52.639 --> 00:15:54.500
any more complicated than that yeah and we all
00:15:54.500 --> 00:15:56.259
know how well that's gone who's been trying to
00:15:56.259 --> 00:15:59.139
use github over the last couple of months Yeah.
00:15:59.299 --> 00:16:02.799
The weird stat I read for hugging faces, I don't
00:16:02.799 --> 00:16:04.799
know if this is real or not, but apparently the
00:16:04.799 --> 00:16:06.700
number that's the cost of the acquisition is
00:16:06.700 --> 00:16:10.840
the Unicode for the hugging face emoji. So there's
00:16:10.840 --> 00:16:13.100
a level to this too, which is just like, are
00:16:13.100 --> 00:16:16.000
they just taking the piss out of us? It's like
00:16:16.000 --> 00:16:18.539
money isn't real. Things are what we say they
00:16:18.539 --> 00:16:21.519
are. We haven't heard about reduction. We haven't
00:16:21.519 --> 00:16:24.340
heard about reduction. It's a gold mix. like
00:16:24.340 --> 00:16:26.240
the big seven are just basically passing around
00:16:26.240 --> 00:16:28.480
the 10, the same $10 million debt to each other.
00:16:28.559 --> 00:16:31.879
Doing all business, but it's just this. Jensen
00:16:31.879 --> 00:16:35.620
against your theory, Kim is the Jensen Huang
00:16:35.620 --> 00:16:39.919
is not Elon Musk and Elon Musk always picks a
00:16:39.919 --> 00:16:42.720
price that has four 20 in it because that's,
00:16:42.720 --> 00:16:44.899
that's apparently the time you go out and smoke
00:16:44.899 --> 00:16:47.740
and smoke dope behind the school bicycle shed
00:16:47.740 --> 00:16:51.059
or something. Yeah. But I don't think Jensen
00:16:51.059 --> 00:16:55.460
Huang has that particular kind of. goofiness.
00:16:55.460 --> 00:16:58.740
I'm not too sure. I think he's pretty goofy,
00:16:58.820 --> 00:17:03.600
actually. Yeah. I know he's on the money here.
00:17:03.840 --> 00:17:06.880
The Taiwanese treason is pretty goofy. He's a
00:17:06.880 --> 00:17:08.859
hero, but in the same way that Elon Musk is.
00:17:09.140 --> 00:17:12.799
Nowadays, it's hard to tell how much of an influence
00:17:12.799 --> 00:17:20.539
it is to get the crypto kitties in. The other
00:17:20.539 --> 00:17:26.980
thing... Did we talk at all about the piece that
00:17:26.980 --> 00:17:30.900
Matthew Green, who is a respected cryptography
00:17:30.900 --> 00:17:34.880
professor, apparently wrote this posting saying
00:17:34.880 --> 00:17:38.420
that AI was making software so secure that government
00:17:38.420 --> 00:17:40.880
wouldn't be able to hack into it? I mean, I was
00:17:40.880 --> 00:17:42.960
surprised because I would have expected Matthew
00:17:42.960 --> 00:17:44.980
Green to think that was a good thing, but apparently
00:17:44.980 --> 00:17:50.539
not. I'm torn. I am at this point desperate for
00:17:50.539 --> 00:17:52.500
anything that looks positive to come out of this
00:17:52.500 --> 00:17:55.579
situation. And the idea that being able to brute
00:17:55.579 --> 00:17:58.180
force via AI all of the stupid little hacks that
00:17:58.180 --> 00:17:59.759
have been letting all the other assholes get
00:17:59.759 --> 00:18:01.500
into all the systems, like I'll take a little
00:18:01.500 --> 00:18:11.200
bit of positivity. My sarcastic, cynical brain
00:18:11.200 --> 00:18:13.960
is going, that looks like it was written by someone
00:18:13.960 --> 00:18:16.220
who wanted to make AI sound good. But I can also
00:18:16.220 --> 00:18:18.259
see that. Actually making sense. And you said
00:18:18.259 --> 00:18:20.059
it's like this fellow who did this, he knows
00:18:20.059 --> 00:18:22.640
his stuff. He's not, you know. Oh, yeah. Matthew
00:18:22.640 --> 00:18:26.220
Green is a respected security guy. You know,
00:18:26.220 --> 00:18:31.140
no, he's not. I mean, that would be nice. I mean,
00:18:31.220 --> 00:18:34.640
I certainly would love to hear something being
00:18:34.640 --> 00:18:37.960
made slightly more secure. Yeah, that was why
00:18:37.960 --> 00:18:41.700
the thing surprised me was it sounded like. Anyway.
00:18:42.870 --> 00:18:45.130
Well, for the two things we haven't quite managed
00:18:45.130 --> 00:18:48.089
to get to yet, do we want to talk about banning
00:18:48.089 --> 00:18:51.329
generative AI for the school systems in New York?
00:18:51.369 --> 00:18:53.450
I would say, to be honest, this piece about,
00:18:53.650 --> 00:18:56.269
yeah, I didn't quite realize which way around
00:18:56.269 --> 00:18:57.990
this was. To me, it's kind of insane, the idea
00:18:57.990 --> 00:18:59.910
that we must have insecure software forever so
00:18:59.910 --> 00:19:01.750
that governments can hack people. I don't think
00:19:01.750 --> 00:19:03.990
that's good logic. But also, like, the piece
00:19:03.990 --> 00:19:06.170
I've pasted is about, there's a story of this
00:19:06.170 --> 00:19:08.549
massive Berlin data dump that's like terabytes
00:19:08.549 --> 00:19:11.559
of data. And a lot of it just seems to be like
00:19:11.559 --> 00:19:15.500
there was like a ransom demand that wasn't paid.
00:19:16.019 --> 00:19:19.180
And and a lot of if you read the article, it
00:19:19.180 --> 00:19:21.220
was stuff that was on like Windows 95, Windows
00:19:21.220 --> 00:19:24.019
98 computers. It doesn't seem like it's been
00:19:24.019 --> 00:19:25.839
hard to get. And there's a lot of stuff in there
00:19:25.839 --> 00:19:28.680
about like what the nuclear and chemical warfare
00:19:28.680 --> 00:19:34.849
plans are. So I don't know. I also, like, there
00:19:34.849 --> 00:19:37.329
was some massive hack that happened recently.
00:19:37.549 --> 00:19:39.730
It took down, I just saw, I wish I'd written
00:19:39.730 --> 00:19:43.569
it down. And the perennial reason for the hack
00:19:43.569 --> 00:19:46.230
was just a straight up social engineering thing.
00:19:46.369 --> 00:19:50.230
Like somebody found who the right person was,
00:19:50.430 --> 00:19:52.950
called their tech support, the person who had
00:19:52.950 --> 00:19:54.210
the right kind of permissions, called the tech
00:19:54.210 --> 00:19:57.049
support center, pretended to be them, got a password
00:19:57.049 --> 00:20:01.740
reset, logged in. Bob's your uncle. And there
00:20:01.740 --> 00:20:07.099
is no amount of AI improvement of the software
00:20:07.099 --> 00:20:10.319
that will get around the human element. Who was
00:20:10.319 --> 00:20:14.779
Bob? Where does that come from, do you know?
00:20:14.960 --> 00:20:17.279
I think it's an Australian term, originally.
00:20:17.880 --> 00:20:20.299
Really? Again, I've been reading He Who Fights
00:20:20.299 --> 00:20:21.920
With Monsters, which has an Australian protagonist,
00:20:22.119 --> 00:20:25.900
so it may be influencing me right now. I assumed
00:20:25.900 --> 00:20:29.059
it was British. Did we want to cover the school
00:20:29.059 --> 00:20:34.799
systems or the... The school system one is...
00:20:34.799 --> 00:20:39.859
There's so much. It feels like it's a whole...
00:20:39.859 --> 00:20:44.259
I think it's a really good testing of the waters.
00:20:44.440 --> 00:20:47.519
We'll keep an eye out for it. So my feeling on
00:20:47.519 --> 00:20:50.140
this is... I'm finding AI a really useful tool
00:20:50.140 --> 00:20:51.579
for coding, but it's only because I already know
00:20:51.579 --> 00:20:53.720
how to code, right? So like I've made a bunch
00:20:53.720 --> 00:20:55.539
of Rails apps and now I can make Rails apps and
00:20:55.539 --> 00:20:57.160
I know what a good Rails app looks like. And
00:20:57.160 --> 00:20:59.000
obviously it's not, you know, I think to be honest,
00:20:59.039 --> 00:21:01.400
it does a much better job than me. And I think
00:21:01.400 --> 00:21:03.400
like for me, why this really makes sense is because
00:21:03.400 --> 00:21:05.019
like you have to learn to do the thing before
00:21:05.019 --> 00:21:06.920
you try and automate the thing. And that's the
00:21:06.920 --> 00:21:09.220
really scary. That's like the model collapse.
00:21:09.259 --> 00:21:11.660
But for society, right? No context collapse for
00:21:11.660 --> 00:21:13.920
society, right? When like people are trying to
00:21:13.920 --> 00:21:15.680
do things they've never done the manual way.
00:21:15.779 --> 00:21:17.720
And then they're giving jobs where they have
00:21:17.720 --> 00:21:19.259
to automate a thing they've never done. Like
00:21:19.259 --> 00:21:21.240
it just doesn't add up to me. You don't put someone
00:21:21.240 --> 00:21:23.279
who doesn't know how to weld into a welding mech
00:21:23.279 --> 00:21:26.890
suit. Right. That's a recipe for disaster. yeah
00:21:26.890 --> 00:21:29.130
yeah probably won't leave their suit now now
00:21:29.130 --> 00:21:30.950
you can now you can not know what you're doing
00:21:30.950 --> 00:21:34.930
so much faster than before i'm definitely i'm
00:21:34.930 --> 00:21:36.410
i'm having that similar experience with coding
00:21:36.410 --> 00:21:39.269
and doing other things where having a large and
00:21:39.269 --> 00:21:42.150
deep breadth of knowledge having coded for decades
00:21:42.150 --> 00:21:44.230
means that i know what to ask for and i know
00:21:44.230 --> 00:21:46.029
how to watch and i know how to have the moments
00:21:46.029 --> 00:21:48.690
of sure that's right oh they're absolutely right
00:21:48.690 --> 00:21:50.430
to have asked me about that i will go correct
00:21:50.430 --> 00:21:53.049
that right now thank crap okay fine because an
00:21:53.049 --> 00:21:54.769
ai can tell you something's internally consistent
00:21:54.769 --> 00:21:57.500
but when it can't tell you The thing that only
00:21:57.500 --> 00:21:59.420
a human can do is tell you if it's valuable or
00:21:59.420 --> 00:22:02.579
useful. If you actually went the right direction.
00:22:02.839 --> 00:22:04.140
And it's interesting how we're building that
00:22:04.140 --> 00:22:06.900
skill set. We seem to be dodging around. Do we
00:22:06.900 --> 00:22:12.660
want to tackle the Israel think tank? You said
00:22:12.660 --> 00:22:17.799
it wasn't a real think tank. I'm up for it. They're
00:22:17.799 --> 00:22:22.960
faking a think tank to produce articles that
00:22:22.960 --> 00:22:26.490
are pro... Israel. And this is, this is, this
00:22:26.490 --> 00:22:28.029
is something we talked about. And for me, so
00:22:28.029 --> 00:22:29.609
how this dovetails into the discussions we've
00:22:29.609 --> 00:22:31.190
had earlier that I found interesting. It was
00:22:31.190 --> 00:22:33.690
like, I remember saying a year ago on tag run,
00:22:33.730 --> 00:22:36.269
something like if I was someone in Russia right
00:22:36.269 --> 00:22:39.890
now, I would be very keen to get a whole bunch
00:22:39.890 --> 00:22:41.869
of stuff into all the data that they're using
00:22:41.869 --> 00:22:44.049
to train these models to make it say the things
00:22:44.049 --> 00:22:45.809
that I want it to say right now. So that a year
00:22:45.809 --> 00:22:47.309
in the future, when these models comes out, they're
00:22:47.309 --> 00:22:49.789
like, what should I do? What package should I
00:22:49.789 --> 00:22:51.730
use? You should use Russian software package.
00:22:51.950 --> 00:22:55.240
Give me all of your money. you know and like
00:22:55.240 --> 00:22:56.680
well there's something else but you see where
00:22:56.680 --> 00:22:59.079
i'm going with that and i the funny thing for
00:22:59.079 --> 00:23:02.339
me was like i think in my mind i was wrong this
00:23:02.339 --> 00:23:04.920
has allowed me to you know reckless i'm thinking
00:23:04.920 --> 00:23:07.099
that there will be a day when we'll have trained
00:23:07.099 --> 00:23:09.920
the models and we will stop that is clearly not
00:23:09.920 --> 00:23:13.900
the case we will be trained until until the data
00:23:13.900 --> 00:23:16.700
centers are all molten piles of slag we will
00:23:16.700 --> 00:23:19.359
forever be training them and now literally like
00:23:19.359 --> 00:23:21.200
the thing for me that i i didn't realize and
00:23:21.200 --> 00:23:24.359
now i have a vision of this is that We are running
00:23:24.359 --> 00:23:29.079
out of original data. Oh, I wrote about that
00:23:29.079 --> 00:23:32.640
like two or three years ago. Yeah. There was
00:23:32.640 --> 00:23:35.079
a study that said that we were running out of
00:23:35.079 --> 00:23:37.539
high quality text because we don't produce it
00:23:37.539 --> 00:23:39.980
nearly as fast as... No. It can be consumed.
00:23:40.900 --> 00:23:42.460
There's been stories in the last couple of months
00:23:42.460 --> 00:23:46.779
of people bulk, of companies bulk buying... Shit
00:23:46.779 --> 00:23:50.339
old books. Antiquarian books and then destroying
00:23:50.339 --> 00:23:56.059
them. to every person who ever loved a book.
00:23:56.240 --> 00:23:58.440
Right. There's a great, great bookshop in New
00:23:58.440 --> 00:24:00.920
Zealand that realized that that's what was happening
00:24:00.920 --> 00:24:04.440
was that they were buying these niche titles
00:24:04.440 --> 00:24:07.480
and they actually canceled the orders for thousands
00:24:07.480 --> 00:24:11.259
of books because they worked it out. They're
00:24:11.259 --> 00:24:16.279
selling books to have them destroyed. Do you
00:24:16.279 --> 00:24:18.099
think that we are learning the craft of how to
00:24:18.970 --> 00:24:22.150
provide the right crap but it's also that now
00:24:22.150 --> 00:24:26.470
it's also a an ongoing game it's like sure we'll
00:24:26.470 --> 00:24:28.809
skew it this way for you know opus five is going
00:24:28.809 --> 00:24:32.690
to have a slight preference towards chocolate
00:24:32.690 --> 00:24:36.509
chip ice cream but the next time around the you
00:24:36.509 --> 00:24:38.930
know the rocky road folks are going to win by
00:24:38.930 --> 00:24:41.710
doing it he said bring it into a very carefully
00:24:41.710 --> 00:24:44.630
neutral territory the really interesting thing
00:24:44.630 --> 00:24:48.500
about that is that the people who use chatbots,
00:24:48.579 --> 00:24:52.559
generalized chatbots, as companions and lovers
00:24:52.559 --> 00:24:57.160
and so forth can actually get quite upset when
00:24:57.160 --> 00:25:00.740
they're updated and then reactions they're used
00:25:00.740 --> 00:25:04.380
to getting don't appear. Yeah, model updates
00:25:04.380 --> 00:25:06.660
on those. What's the one character or something?
00:25:06.779 --> 00:25:09.119
They've said they've slowed them down for exactly
00:25:09.119 --> 00:25:11.799
that reason. People got very upset. When the
00:25:11.799 --> 00:25:15.440
intelligence, I'm using air quotes, of the models
00:25:15.440 --> 00:25:17.099
increased and they stopped using the sort of
00:25:17.099 --> 00:25:19.079
same terms that they were endeared to. I mean,
00:25:19.099 --> 00:25:21.500
it's worse than having your desktop rearranged.
00:25:24.160 --> 00:25:28.000
That's my thing. I guess, like, for me, why this
00:25:28.000 --> 00:25:30.579
thing's so interesting is because, like, yeah.
00:25:31.099 --> 00:25:33.019
it's like if you're disingenuous and you and
00:25:33.019 --> 00:25:35.119
you don't care or like you're actively trying
00:25:35.119 --> 00:25:37.319
to cause harm like the israeli government is
00:25:37.319 --> 00:25:39.859
it's an incredible tool because like you know
00:25:39.859 --> 00:25:41.680
the whole way that like search engine optimization
00:25:41.680 --> 00:25:44.059
works on the net is basically around like authority
00:25:44.059 --> 00:25:46.039
and the amount of things you have pointing to
00:25:46.039 --> 00:25:48.559
you and i feel like for anyone who's worked in
00:25:48.559 --> 00:25:50.500
the third sector there is just this culture of
00:25:50.500 --> 00:25:53.680
like reports having their own seriousness as
00:25:53.680 --> 00:25:55.920
if they mean anything even though i think most
00:25:55.920 --> 00:25:58.359
people skim read the executive summary i found
00:25:58.359 --> 00:26:00.500
some clangers in like published reports that
00:26:00.500 --> 00:26:02.619
people have put out so i think it's interesting
00:26:02.619 --> 00:26:03.960
because a lot of the stuff that's happening now
00:26:03.960 --> 00:26:07.059
is sort of like taking advantage of these like
00:26:07.059 --> 00:26:09.759
cultural weaknesses we have to things like the
00:26:09.759 --> 00:26:12.680
idea of a report right and it's like and now
00:26:12.680 --> 00:26:15.240
you can produce them as fast as you can prompt
00:26:15.240 --> 00:26:17.720
your agent but it's like you know the thing that
00:26:17.720 --> 00:26:19.599
was the problem was the idea that these are things
00:26:19.599 --> 00:26:21.980
that should be like inherently given this air
00:26:21.980 --> 00:26:24.440
of seriousness anyway i don't know how to describe
00:26:24.440 --> 00:26:27.480
it we saw a very early version of that in the
00:26:27.480 --> 00:26:30.660
problem that the guardian was having with falsely
00:26:30.660 --> 00:26:33.359
generated article references in other people's
00:26:33.359 --> 00:26:35.140
work people were doing academic work and then
00:26:35.140 --> 00:26:40.039
the and then basically The AI would make up a
00:26:40.039 --> 00:26:43.759
quote to, like, the January 17th, you know, 2022
00:26:43.759 --> 00:26:46.980
article from The Guardian on why I'm right and
00:26:46.980 --> 00:26:50.579
you're wrong in depth analysis. And then The
00:26:50.579 --> 00:26:53.420
Guardian was getting, and many other newspapers,
00:26:53.559 --> 00:26:55.059
like, where's this article? I have to find this
00:26:55.059 --> 00:26:58.240
article. And this is a very interesting, like,
00:26:58.339 --> 00:27:01.740
if you're really technically minded, there's
00:27:01.740 --> 00:27:04.500
a, like, one of the most obnoxious things to
00:27:04.500 --> 00:27:07.720
do is to prove that something doesn't exist.
00:27:08.269 --> 00:27:10.849
in your corpus because the only way to do that
00:27:10.849 --> 00:27:14.170
with authority is to go through freaking everything
00:27:14.170 --> 00:27:18.630
and know that you couldn't find it and originally
00:27:18.630 --> 00:27:21.710
because no one had ever done this before until
00:27:21.710 --> 00:27:24.250
people at the guardian and these other places
00:27:24.250 --> 00:27:27.150
caught on that's like oh someone's making fake
00:27:27.150 --> 00:27:31.339
citations like until that point They were throwing
00:27:31.339 --> 00:27:33.380
a lot of energy at trying to find these articles
00:27:33.380 --> 00:27:35.619
because they presumed by that same methodology
00:27:35.619 --> 00:27:36.880
that you were just talking about, Kim, where
00:27:36.880 --> 00:27:40.299
it's like, oh, this is a thing that only exists
00:27:40.299 --> 00:27:42.519
when it's real, right? No one would make a fake
00:27:42.519 --> 00:27:46.299
article citation. AI would. AI did. didn't even
00:27:46.299 --> 00:27:48.579
specifically a human with bad intentions would
00:27:48.579 --> 00:27:51.359
prompt ai to do it right oh yeah now actually
00:27:51.359 --> 00:27:54.119
before it wasn't even a human prompting in many
00:27:54.119 --> 00:27:56.460
cases it was just ai trying really hard to do
00:27:56.460 --> 00:27:58.740
what you thought you wanted it to do and it goes
00:27:58.740 --> 00:27:59.940
oh there should be something that looks like
00:27:59.940 --> 00:28:01.039
a citation here i'll just make something that
00:28:01.039 --> 00:28:02.700
looks like a citation yeah yeah yeah yeah yeah
00:28:02.700 --> 00:28:06.180
yeah that's what you're saying yeah yeah and
00:28:06.180 --> 00:28:08.220
now the humans went oh that's a good idea why
00:28:08.220 --> 00:28:10.420
don't we do more of that oh did that really mess
00:28:10.420 --> 00:28:12.420
them up we should do more of that yep absolutely
00:28:14.379 --> 00:28:17.240
I'm still puzzled by the idea of a fake think
00:28:17.240 --> 00:28:23.920
tank. It's just the think tank. It's the presumed
00:28:23.920 --> 00:28:29.299
academic cred or sense of authority that is given
00:28:29.299 --> 00:28:31.500
when these articles come out of this quote -unquote
00:28:31.500 --> 00:28:33.819
think tank. Because how many of us, like we're
00:28:33.819 --> 00:28:38.380
building the habit now, but think tank, do we
00:28:38.380 --> 00:28:41.089
always go check a think tank? I mean, why would
00:28:41.089 --> 00:28:42.849
we say think tank if it wasn't supposed to mean
00:28:42.849 --> 00:28:44.369
that people who were really smart thought about
00:28:44.369 --> 00:28:46.230
it? Because that's what a think tank is. Except
00:28:46.230 --> 00:28:48.069
that this think tank is my goldfish bowl. I've
00:28:48.069 --> 00:28:50.690
just called it a think tank. But we don't think
00:28:50.690 --> 00:28:56.670
of that. And that's it for our wonderful roundtable
00:28:56.670 --> 00:29:00.170
discussion of all things AI. Now, we've got Wendy
00:29:00.170 --> 00:29:03.549
talking to us about The Nerd Reich by Gil Duran.
00:29:04.670 --> 00:29:08.069
Yeah, Gil Duran is a San Francisco journalist.
00:29:10.170 --> 00:29:13.809
His book, well, the subtitle of his book is Silicon
00:29:13.809 --> 00:29:17.789
Valley Fascism and the War on Democracy. And
00:29:17.789 --> 00:29:21.869
it joins a sort of subgenre of publishing about
00:29:21.869 --> 00:29:25.930
the rightward turn of Silicon Valley, which actually
00:29:25.930 --> 00:29:29.549
started more than 20 years ago with Paulina Borsuk's
00:29:29.549 --> 00:29:33.670
book, Cyber Selfish. And when she wrote it, it
00:29:33.670 --> 00:29:37.829
was widely ignored and practically bankrupted
00:29:37.829 --> 00:29:40.420
her, she keeps saying. And it's actually now
00:29:40.420 --> 00:29:42.420
going to be republished this year because people
00:29:42.420 --> 00:29:44.319
have rediscovered it and realized that she was
00:29:44.319 --> 00:29:50.039
right all along. And so there have been more
00:29:50.039 --> 00:29:52.900
recent entries into this particular little subgenre,
00:29:52.940 --> 00:29:55.380
Throwing Rocks at the Google Bus, I think, and
00:29:55.380 --> 00:29:58.680
The Survival of the Richest by Douglas Rushkoff,
00:29:58.839 --> 00:30:03.740
and Gilded Rage, which was mostly about Elon
00:30:03.740 --> 00:30:07.279
Musk from Jacob Silverman. But The Nerd Reich
00:30:07.279 --> 00:30:13.720
basically starts with a book called The Sovereign
00:30:13.720 --> 00:30:19.619
Individual by William Rees -Mogg and a co -author
00:30:19.619 --> 00:30:22.259
whose name I forget. But, you know, you sort
00:30:22.259 --> 00:30:23.819
of look at that and you go, Rees -Mogg, you mean
00:30:23.819 --> 00:30:26.059
like in Jacob? And, of course, it was his father.
00:30:26.799 --> 00:30:31.460
Right. Famous editor. But a lot of the stuff
00:30:31.460 --> 00:30:34.140
that we see from Silicon Valley is about this,
00:30:34.180 --> 00:30:37.799
is wanting to build little private states, for
00:30:37.799 --> 00:30:43.779
example. And so what Duran does is really follow
00:30:43.779 --> 00:30:46.119
the politics, which is kind of different from
00:30:46.119 --> 00:30:48.940
a lot of these books, which really focus on Silicon
00:30:48.940 --> 00:30:52.329
Valley. What he's really aiming is to show the
00:30:52.329 --> 00:30:54.990
real danger to democracy. And he's he seems to
00:30:54.990 --> 00:30:57.470
be most frightened of J .D. Vance because he's
00:30:57.470 --> 00:31:00.130
already in a position as vice president. And
00:31:00.130 --> 00:31:02.569
he describes him as a complete sort of completely
00:31:02.569 --> 00:31:06.089
manufactured creature of Peter Thiel. And Thiel,
00:31:06.210 --> 00:31:09.069
he sees as the guy behind a lot of this. And
00:31:09.069 --> 00:31:12.029
Thiel has acolytes all over the White House.
00:31:12.849 --> 00:31:18.559
So he picked them. Yeah. Yeah. Yeah. Yeah. So
00:31:18.559 --> 00:31:20.660
that's kind of money. Why not go out and pick
00:31:20.660 --> 00:31:22.859
up a couple of, you know, vice presidents along
00:31:22.859 --> 00:31:25.000
with the milk and the butter? Well, you only
00:31:25.000 --> 00:31:28.380
need one. You say that, but I bet he's probably
00:31:28.380 --> 00:31:33.099
got three or four. He probably has, yes, other
00:31:33.099 --> 00:31:36.720
ideas. Contingencies within contingencies. Well,
00:31:36.720 --> 00:31:39.259
he picks up people. I mean, everyone says that.
00:31:39.319 --> 00:31:42.779
Yeah. But apparently Duran was inspired to write
00:31:42.779 --> 00:31:45.779
the book by this thing going on in Solano County
00:31:45.779 --> 00:31:48.539
in California, where basically they're trying
00:31:48.539 --> 00:31:50.799
to take it over as a private enterprise. And
00:31:50.799 --> 00:31:54.440
the residents are objecting, surprisingly enough.
00:31:57.680 --> 00:32:01.059
We visited quite a few of those micro states
00:32:01.059 --> 00:32:04.579
last year. I think we covered two or three of
00:32:04.579 --> 00:32:07.579
those. um places that i've been trying to make
00:32:07.579 --> 00:32:10.619
yes separate sovereign states so they can go
00:32:10.619 --> 00:32:14.259
all out capitalist nuts so with as much cryptocurrency
00:32:14.259 --> 00:32:16.299
i think the difference is then it was sort of
00:32:16.299 --> 00:32:19.319
considered some sort of cryptocurrency scam whereas
00:32:19.319 --> 00:32:22.799
that isn't the point anymore no no i sort of
00:32:22.799 --> 00:32:25.079
was telling a friend of mine a 17 year old english
00:32:25.079 --> 00:32:26.880
girl about about this book and she said well
00:32:26.880 --> 00:32:29.980
you know kind of what's the relevance You know,
00:32:29.980 --> 00:32:32.000
and I said, well, you know, they're talking about
00:32:32.000 --> 00:32:35.720
charter, you know, these microstates. Think Rishi
00:32:35.720 --> 00:32:38.700
Sunak and all these free ports he wanted to build.
00:32:39.019 --> 00:32:44.380
Yeah. I mean, it's weird how we were just talking
00:32:44.380 --> 00:32:46.980
about the evolution of the company town and how
00:32:46.980 --> 00:32:48.640
that that's now something. And it's like and
00:32:48.640 --> 00:32:50.339
it's almost like they went, how about a company
00:32:50.339 --> 00:32:52.700
town? But none of those poor people show up and
00:32:52.700 --> 00:32:54.339
it's just us rich people and we get to do whatever
00:32:54.339 --> 00:32:56.240
we want because we're all rich. Or you have the
00:32:56.240 --> 00:32:58.799
poor people there to do things like... Yes, the
00:32:58.799 --> 00:33:00.359
poor people have to be there. No, you till your
00:33:00.359 --> 00:33:03.500
garden and clean your house. Well, to bring it
00:33:03.500 --> 00:33:06.019
back to Cory Dottrow, they're the reverse centaurs.
00:33:06.359 --> 00:33:09.920
Right, but you do what they do in Dubai, which
00:33:09.920 --> 00:33:12.319
is you don't let them have any rights. Yeah,
00:33:12.359 --> 00:33:15.079
they just round the back and do the work. Was
00:33:15.079 --> 00:33:16.819
it Douglas Rushkoff who said he got paid a lot
00:33:16.819 --> 00:33:19.339
of money to go and... He got paid a lot of money
00:33:19.339 --> 00:33:22.279
by some billionaires. Yeah, to go and talk to
00:33:22.279 --> 00:33:24.599
these billionaires who were asking him things
00:33:24.599 --> 00:33:26.940
like, how do I keep my private security force
00:33:26.940 --> 00:33:32.920
loyal to me? After the event. People see in the
00:33:32.920 --> 00:33:35.359
Apple TV series, Silo, because I just realized
00:33:35.359 --> 00:33:37.000
how on point it is, but I don't want to spoil
00:33:37.000 --> 00:33:41.259
anything. Yeah, that's basically the same. I
00:33:41.259 --> 00:33:44.880
mean, honestly, I saw Silo and I've seen some
00:33:44.880 --> 00:33:48.380
of it and my brain just went, this is just grim,
00:33:48.400 --> 00:33:53.769
dark fallout. Grimdark Fallout. Is that an actual
00:33:53.769 --> 00:33:57.150
thing? Well, Grimdark is a thing and Fallout
00:33:57.150 --> 00:33:59.230
is a video game that now got turned into an actually
00:33:59.230 --> 00:34:01.630
really good TV series. But it's also it's the
00:34:01.630 --> 00:34:04.210
same premise as that super rich people. They
00:34:04.210 --> 00:34:07.029
have a similar premise of vaults in the ground
00:34:07.029 --> 00:34:10.090
where people go to survive out the survive out
00:34:10.090 --> 00:34:12.139
the nuclear whatever. And then it turns out,
00:34:12.139 --> 00:34:14.360
oh, they're actually experiments. Grim Dark was
00:34:14.360 --> 00:34:16.599
the name for Warhammer 40 ,000 and other genre
00:34:16.599 --> 00:34:19.079
similarities. I actually wanted to write something
00:34:19.079 --> 00:34:23.940
like that around the year 2000 when... You know,
00:34:23.960 --> 00:34:25.559
there were all these guys on the Internet talking
00:34:25.559 --> 00:34:27.239
about how they had to go live in bunkers for
00:34:27.239 --> 00:34:29.840
10 years or something to be safe. And I was,
00:34:29.840 --> 00:34:32.679
you know, it was sort of 2010. I was thinking,
00:34:32.760 --> 00:34:35.679
you know, it'd be pretty amusing to write something
00:34:35.679 --> 00:34:38.079
about the people emerging from these bunkers.
00:34:38.159 --> 00:34:40.639
And not only is the world peaceful and continuing,
00:34:40.739 --> 00:34:42.599
but everybody has these weird devices and you
00:34:42.599 --> 00:34:45.940
have missed, you know, 10 generations of technology
00:34:45.940 --> 00:34:48.139
and you have no idea how to make the world work.
00:34:48.519 --> 00:34:51.519
I think Laura Penny did write a short story on
00:34:51.519 --> 00:34:54.559
the sort of similar. point about people these
00:34:54.559 --> 00:34:57.000
rich people actually going in there and actually
00:34:57.000 --> 00:34:59.000
being a war and actually not surviving because
00:34:59.000 --> 00:35:01.300
of course yeah they have lots of money that has
00:35:01.300 --> 00:35:04.519
no value which is the original point i think
00:35:04.519 --> 00:35:06.199
the most insane thing these nuclear bunkers is
00:35:06.199 --> 00:35:08.260
like you know people like sam altman kind of
00:35:08.260 --> 00:35:10.260
brag about theirs and like why is anyone taking
00:35:10.260 --> 00:35:12.659
advice about the future from a billionaire who's
00:35:12.659 --> 00:35:14.960
got their own private nuclear book absolutely
00:35:14.960 --> 00:35:19.820
like the stakes are just different for us Speaking
00:35:19.820 --> 00:35:22.039
of Peter Thiel, did you hear that he'd given
00:35:22.039 --> 00:35:24.099
up on New Zealand and he's gone to Argentina?
00:35:25.139 --> 00:35:29.480
No. Yeah, because New Zealand's too woke. That's
00:35:29.480 --> 00:35:34.440
fucking on brand, isn't it? My nuclear bunker
00:35:34.440 --> 00:35:38.500
is too woke. I was more thinking about the Argentina
00:35:38.500 --> 00:35:43.920
part, but that's a deep cut. Wow. Well, he may
00:35:43.920 --> 00:35:49.739
come to regret it. Who knows? All right, next
00:35:49.739 --> 00:35:53.739
up is an article about trying to get all of your
00:35:53.739 --> 00:35:57.739
data from companies and how they replied by deleting
00:35:57.739 --> 00:36:01.880
it instead. Ian? Yeah, so this is a story from
00:36:01.880 --> 00:36:05.900
Wired Magazine. And ultimately, a journalist
00:36:05.900 --> 00:36:10.699
goes around trying to get their data. They kind
00:36:10.699 --> 00:36:13.900
of want to understand what the companies are
00:36:13.900 --> 00:36:17.369
basing their algorithms on. trying to understand
00:36:17.369 --> 00:36:19.630
what kind of data that they've been collecting.
00:36:20.369 --> 00:36:26.750
And as the story title kind of suggests, they
00:36:26.750 --> 00:36:29.530
turn back and go, oh, we haven't got any data.
00:36:29.829 --> 00:36:32.710
Yeah, we've deleted it all. Most of them even
00:36:32.710 --> 00:36:35.230
say they just don't have any data. And then they
00:36:35.230 --> 00:36:37.429
kind of say, well, we've deleted it. They have
00:36:37.429 --> 00:36:42.690
no data? Yeah. This is, and the reason why I
00:36:42.690 --> 00:36:46.849
bring this up is because as someone who... requested
00:36:46.849 --> 00:36:50.630
the GDPR data, you know, when GDPR, when we were
00:36:50.630 --> 00:36:53.989
still in the EU, to be fair, you know, just like
00:36:53.989 --> 00:36:56.949
requesting it and trying to get the data just
00:36:56.949 --> 00:36:59.769
to try and understand what, especially some of
00:36:59.769 --> 00:37:04.150
the dating sites, what they were using, how they
00:37:04.150 --> 00:37:09.989
saw me, all those things. This speaks volumes
00:37:09.989 --> 00:37:13.909
exactly what was happening back then. So I remember,
00:37:14.050 --> 00:37:19.289
for example, oh what was the site called um one
00:37:19.289 --> 00:37:21.650
of the sites i can't which one it is now you
00:37:21.650 --> 00:37:24.590
know they went over there 34 days over 30 days
00:37:24.590 --> 00:37:26.630
because in gpr you've got a certain amount of
00:37:26.630 --> 00:37:30.690
time and they um said oh we don't actually have
00:37:30.690 --> 00:37:33.309
any data on you and i was like well i have an
00:37:33.309 --> 00:37:39.210
account i closed the account but and just this
00:37:39.210 --> 00:37:41.929
is the kind of thing that happens now this journalist
00:37:41.929 --> 00:37:45.480
story goes through all this and a lot more, but
00:37:45.480 --> 00:37:48.480
it's worth a read. But I was wondering how many
00:37:48.480 --> 00:37:51.099
of the tech firms have actually requested their
00:37:51.099 --> 00:37:59.460
data and what you've had in response. Have you
00:37:59.460 --> 00:38:01.880
had a response? And then even if you had a response,
00:38:02.079 --> 00:38:05.619
did you actually get any data at all? I know
00:38:05.619 --> 00:38:09.079
stuff like most of us have probably moved from
00:38:09.079 --> 00:38:13.869
Twitter. That's the only example I can think
00:38:13.869 --> 00:38:16.309
of. Yeah. I'm wondering about some of the other
00:38:16.309 --> 00:38:18.710
ones. A lot of the sites do not, they literally
00:38:18.710 --> 00:38:22.630
ignore you. And then some of them say, oh, well,
00:38:22.750 --> 00:38:25.789
actually, you're based in the UK and we're based
00:38:25.789 --> 00:38:29.969
in America, so we don't respond to GDPR requests,
00:38:30.150 --> 00:38:32.650
which is like, no, that's not the way the law
00:38:32.650 --> 00:38:36.250
works. And then, yeah, a lot of them just kind
00:38:36.250 --> 00:38:38.289
of say, well, we don't have any data on you.
00:38:39.210 --> 00:38:41.469
And then that's it. And then some of your accounts
00:38:41.469 --> 00:38:44.570
close down. Or I'm just wondering what other
00:38:44.570 --> 00:38:46.469
people's thoughts are and what other people's
00:38:46.469 --> 00:38:49.610
experiences are of it. Because I feel like this
00:38:49.610 --> 00:38:55.670
story speaks volumes about what I had back 10
00:38:55.670 --> 00:38:59.070
years ago, or just under 10 years ago, and it's
00:38:59.070 --> 00:39:02.670
still happening now. So it's more of a question
00:39:02.670 --> 00:39:06.409
to you all. There's some complications with that
00:39:06.409 --> 00:39:08.809
story. And you actually made them clear, maybe
00:39:08.809 --> 00:39:11.469
accidentally, because you said data and then
00:39:11.469 --> 00:39:13.250
you said algorithm. Now, they don't have to store
00:39:13.250 --> 00:39:16.110
or give you an algorithm, of course. Some of
00:39:16.110 --> 00:39:19.289
the data may mean absolutely nothing. Doesn't
00:39:19.289 --> 00:39:21.550
have to be in English. Doesn't have to, in fact,
00:39:21.570 --> 00:39:26.170
be recognisable, readable data. So it's possibly
00:39:26.170 --> 00:39:30.469
when you say my data, you mean the things that
00:39:30.469 --> 00:39:33.380
are just related to you. that haven't been post
00:39:33.380 --> 00:39:37.159
-processed into nothingness. It's not always
00:39:37.159 --> 00:39:40.099
obvious what it is that they could or should
00:39:40.099 --> 00:39:41.980
be keeping, even though you know, you've got
00:39:41.980 --> 00:39:45.739
to count. You're right. Sorry, I'll be very clear
00:39:45.739 --> 00:39:50.360
as well. When you do a GDPR request, you're asking
00:39:50.360 --> 00:39:54.619
for the information that you put into the system.
00:39:55.440 --> 00:39:59.500
You never get the process. So, for example, I
00:39:59.500 --> 00:40:04.579
would love to know how... you know, my spelling
00:40:04.579 --> 00:40:08.960
mistakes have altered the algorithm, but they're
00:40:08.960 --> 00:40:10.159
never going to show you that because that's a
00:40:10.159 --> 00:40:12.059
secret source. And that was something that they,
00:40:12.139 --> 00:40:14.780
you know, they fight a lot about, you know, you
00:40:14.780 --> 00:40:19.880
don't want to share the secret source. They don't
00:40:19.880 --> 00:40:22.079
let you see the inferences they make from your
00:40:22.079 --> 00:40:25.079
data. No, they do not. That's definitely, that's
00:40:25.079 --> 00:40:27.699
a secret source, but you should be able to get
00:40:27.699 --> 00:40:32.440
exactly what you put in. back out. And what I
00:40:32.440 --> 00:40:36.159
found was even with that, you don't even get
00:40:36.159 --> 00:40:38.639
that. A lot of times you just get, sometimes
00:40:38.639 --> 00:40:40.820
there has been in the past, I should have found
00:40:40.820 --> 00:40:43.599
the examples where people have requested the
00:40:43.599 --> 00:40:46.639
data back and all they've got is a blank piece
00:40:46.639 --> 00:40:49.500
of paper back. Like they're literally taking
00:40:49.500 --> 00:40:53.099
the piss. And I think those early days of GDPR,
00:40:53.179 --> 00:40:55.719
they kind of just were like, but it's a legal
00:40:55.719 --> 00:40:59.630
request. So this is where I'm thinking about
00:40:59.630 --> 00:41:02.289
an article we had like a year ago where someone
00:41:02.289 --> 00:41:05.829
was talking about how that the rise of agentic
00:41:05.829 --> 00:41:08.829
handling of things may end up like apparently
00:41:08.829 --> 00:41:11.250
one of the things that's going on is a lot of
00:41:11.250 --> 00:41:16.070
local councils and other services have been quietly
00:41:16.070 --> 00:41:19.550
relying. on making the access to their services
00:41:19.550 --> 00:41:24.409
more and more labyrinthine and relying on the
00:41:24.409 --> 00:41:26.190
fact that the more difficult you make it the
00:41:26.190 --> 00:41:27.869
more percentage of people will not make it through
00:41:27.869 --> 00:41:29.630
or they'll lose patience or they won't have the
00:41:29.630 --> 00:41:31.869
yeah you know executive functioning to go through
00:41:31.869 --> 00:41:34.110
it and there's i i'm hopeful that there's a version
00:41:34.110 --> 00:41:36.329
of this which is like if it is truly a legal
00:41:36.329 --> 00:41:38.969
requirement and they keep ignoring it they they
00:41:38.969 --> 00:41:40.630
are going it's going to be more of a pain for
00:41:40.630 --> 00:41:44.309
you to sue us than it will be for us to just
00:41:44.309 --> 00:41:47.739
ignore you and guess what ai can do for you there
00:41:47.739 --> 00:41:49.920
i think we're already there i was going to say
00:41:49.920 --> 00:41:52.380
i've never done a subject access request i think
00:41:52.380 --> 00:41:54.780
it's called but i have done foi requests and
00:41:54.780 --> 00:41:57.219
that's a weird game too because everyone's right
00:41:57.219 --> 00:41:59.219
apparently all the foi departments everywhere
00:41:59.219 --> 00:42:02.519
are just swamped right now and again it's like
00:42:02.519 --> 00:42:04.639
a lot of bad actors who are using them to try
00:42:04.639 --> 00:42:07.380
and do things with right i saw something about
00:42:07.380 --> 00:42:09.719
um i don't think it's going to come to anything
00:42:09.719 --> 00:42:12.019
but you know there was some like far right organizers
00:42:12.019 --> 00:42:14.519
who are trying to they were claiming that they're
00:42:14.519 --> 00:42:16.380
going to use it to find all the information on
00:42:16.380 --> 00:42:17.920
all the asylum processing and where everyone
00:42:17.920 --> 00:42:20.260
lives and who's part of the asylum system and
00:42:20.260 --> 00:42:22.579
they're clearly not it's just like a threat but
00:42:22.579 --> 00:42:24.719
like i can't kind of imagine the amount of like
00:42:24.719 --> 00:42:29.679
you know bad faith um FOI requests that are being
00:42:29.679 --> 00:42:31.820
sent out at the moment on kind of every subject.
00:42:32.500 --> 00:42:34.239
It's a denial of service attack, which is what
00:42:34.239 --> 00:42:36.579
you get in any situation. Any situation where
00:42:36.579 --> 00:42:38.679
it's easier, where you can do something that
00:42:38.679 --> 00:42:40.739
causes something else to have to do a whole lot
00:42:40.739 --> 00:42:42.940
of effort, that you can do a little bit of an
00:42:42.940 --> 00:42:45.699
effort. They do have the ability to dub things
00:42:45.699 --> 00:42:49.860
vexatious requests. Yes. So for something like
00:42:49.860 --> 00:42:52.219
that with the far right folks could be. But if
00:42:52.219 --> 00:42:53.699
you're doing this from the perspective of like
00:42:53.699 --> 00:42:55.659
this article where it's like I asked 100 companies
00:42:55.659 --> 00:42:58.760
about myself. Yeah, that's a different. I don't
00:42:58.760 --> 00:43:01.039
know. I don't think you can claim that for GDPR.
00:43:01.199 --> 00:43:02.880
I know you claim that for taxations, but it would
00:43:02.880 --> 00:43:05.300
be a good way to be like, hey, so it turns out
00:43:05.300 --> 00:43:07.519
now it's really easy for me to keep track of
00:43:07.519 --> 00:43:09.719
these 100 requests. And I know exactly what the
00:43:09.719 --> 00:43:11.079
legal requirements are. And it's really easy
00:43:11.079 --> 00:43:12.800
for me to track all of this because now I have
00:43:12.800 --> 00:43:16.369
this. bot that'll do it for me so let's kick
00:43:16.369 --> 00:43:18.269
off this and out of these hundred ones i can
00:43:18.269 --> 00:43:19.869
play a numbers game and if the company doesn't
00:43:19.869 --> 00:43:21.889
give me what i want i can and also you know starting
00:43:21.889 --> 00:43:24.769
to figure out things like what are the What are
00:43:24.769 --> 00:43:27.909
the low effort ways I can try to punish you for
00:43:27.909 --> 00:43:29.809
not giving me this data that don't take me a
00:43:29.809 --> 00:43:31.230
lot of effort, but that will result in money
00:43:31.230 --> 00:43:33.130
like the in the US? They have something called
00:43:33.130 --> 00:43:35.469
a small claims court, which people will sometimes
00:43:35.469 --> 00:43:36.829
take advantage of because they know that if they,
00:43:36.889 --> 00:43:38.269
you know, I'm not going to be able to sue you
00:43:38.269 --> 00:43:39.570
for a million dollars and I'm not going to have
00:43:39.570 --> 00:43:40.989
to pay the lawyers for this. But if I ask you
00:43:40.989 --> 00:43:42.610
for four thousand, sometimes you won't even give
00:43:42.610 --> 00:43:44.230
a shit, won't show up. And then, boom, I get
00:43:44.230 --> 00:43:47.730
four thousand dollars. I guess I feel like the
00:43:47.730 --> 00:43:49.690
writing might be on the wall for FOI, to be honest.
00:43:49.889 --> 00:43:52.150
I would not be completely shocked if pretty soon
00:43:52.150 --> 00:43:53.650
they try and start putting things through the
00:43:53.650 --> 00:43:55.789
courts to repeal it on the basis that it's too
00:43:55.789 --> 00:43:57.650
time -consuming. I'm a little bit worried about
00:43:57.650 --> 00:44:00.969
it. I just want to quickly make the distinction
00:44:00.969 --> 00:44:03.110
you did make it, Kim, between an FOI request
00:44:03.110 --> 00:44:04.750
that's treated differently from a subject request.
00:44:05.269 --> 00:44:09.349
Rather important. I just want to add one last
00:44:09.349 --> 00:44:12.469
thing, and I think... And so the wired piece
00:44:12.469 --> 00:44:16.369
is obviously, not obviously, it is an American
00:44:16.369 --> 00:44:22.070
perspective, but there is a constant thing, especially
00:44:22.070 --> 00:44:24.670
when, for example, the dating site, which I can't
00:44:24.670 --> 00:44:25.909
remember the name of off the top of my head,
00:44:26.030 --> 00:44:30.869
went over their time limit, which was 30 days,
00:44:30.949 --> 00:44:35.090
and they went to 34 days before they actually
00:44:35.090 --> 00:44:39.309
responded to me. That is a breach. And I remember...
00:44:39.639 --> 00:44:43.559
Yeah, forwarding this to, oh God, what's the,
00:44:43.800 --> 00:44:48.280
the Ofcom? For the Ombudsman or Ofcom? Yeah,
00:44:48.340 --> 00:44:52.219
the Ombudsman. And the Ombudsman also are flooded
00:44:52.219 --> 00:44:55.500
with this stuff. And there is a large request
00:44:55.500 --> 00:44:59.500
from a lot of organisations, including the Open
00:44:59.500 --> 00:45:03.920
Rights Group, which I am an advisor to, that
00:45:03.920 --> 00:45:08.940
we need to change. um the way that that's done
00:45:08.940 --> 00:45:11.960
because they're obviously over flooded with this
00:45:11.960 --> 00:45:16.239
stuff and you know it's not helping that these
00:45:16.239 --> 00:45:18.579
companies are getting away with with kind of
00:45:18.579 --> 00:45:22.400
just breaking the law and that's a problem so
00:45:22.400 --> 00:45:26.079
i thought it was a very good uh you know article
00:45:26.079 --> 00:45:33.769
to to highlight all this And next up is another
00:45:33.769 --> 00:45:36.210
article from Ian that's coming from TechRadar.
00:45:36.210 --> 00:45:38.809
It's about the BBC versus Trump case, and they're
00:45:38.809 --> 00:45:42.610
using a VPN argument. Go ahead, Ian. Okay, so
00:45:42.610 --> 00:45:46.989
the BBC, as you might have remembered, that there
00:45:46.989 --> 00:45:53.789
was a problem where they edited part of Trump's
00:45:53.789 --> 00:45:58.469
speech, and Trump didn't like this, or Trump
00:45:58.469 --> 00:46:02.889
understood. The people around Trump didn't like
00:46:02.889 --> 00:46:06.070
this and they decided that they're going to sue
00:46:06.070 --> 00:46:12.329
the BBC. This was a little while ago. So recently
00:46:12.329 --> 00:46:14.570
the BBC has kind of come back and said, look,
00:46:14.630 --> 00:46:22.650
we're going to use the VPN as a major part of
00:46:22.650 --> 00:46:27.989
this because the show was not shown outside the
00:46:27.989 --> 00:46:32.300
UK. So anyone that watched it, because the part
00:46:32.300 --> 00:46:36.820
of the thing that Trump said was anyone in America
00:46:36.820 --> 00:46:41.860
who saw that, you know, well, BBC said that,
00:46:41.880 --> 00:46:46.519
sorry, the BBC is saying that anyone outside
00:46:46.519 --> 00:46:50.239
of the UK who saw this must have used a VPN.
00:46:50.719 --> 00:46:52.739
And they actually have some statistics about
00:46:52.739 --> 00:46:56.639
this or some tracking about this. Trump's saying
00:46:56.639 --> 00:47:02.530
that. that that change, that edit, influenced
00:47:02.530 --> 00:47:09.250
how people felt about what he said. He's saying
00:47:09.250 --> 00:47:12.570
it was reputational damage, right? Yes, exactly.
00:47:13.030 --> 00:47:15.889
So I think it's interesting because I was wondering
00:47:15.889 --> 00:47:20.570
how the BBC was going to stand against this.
00:47:20.750 --> 00:47:22.889
And I think the VPN one's interesting because
00:47:22.889 --> 00:47:26.969
it does cut the number of... or the limitation
00:47:26.969 --> 00:47:32.789
of damage down to quite a small amount. I can't
00:47:32.789 --> 00:47:35.409
remember the exact number. In the document or
00:47:35.409 --> 00:47:40.929
in the article, it has the numbers, but it's
00:47:40.929 --> 00:47:42.650
a very interesting way of going about things.
00:47:42.750 --> 00:47:49.449
It also opens up the issue with VPNs, and it
00:47:49.449 --> 00:47:53.929
does open up a whole number of questions about...
00:47:54.760 --> 00:47:58.619
If you watch this through a VPN, who's responsible
00:47:58.619 --> 00:48:04.480
for that? I actually think it might be a positive
00:48:04.480 --> 00:48:09.179
thing, but I could see it being jumped on by
00:48:09.179 --> 00:48:12.659
other governments going, this is why we need
00:48:12.659 --> 00:48:15.059
to ban all VPNs or something like that. So I
00:48:15.059 --> 00:48:17.320
was just interested in what the rest of the tech
00:48:17.320 --> 00:48:19.920
grumps thought about it. Well, you know, the
00:48:19.920 --> 00:48:24.070
first thing that occurs to me... There was a
00:48:24.070 --> 00:48:27.429
case recently in Texas where the court ordered
00:48:27.429 --> 00:48:33.010
ICANN to lock the domain of this porn site that
00:48:33.010 --> 00:48:34.929
was based, I forget where it was based, but it
00:48:34.929 --> 00:48:39.230
wasn't based in the States. And actually, you
00:48:39.230 --> 00:48:42.349
know, a court could order ICANN to lock the BBC's
00:48:42.349 --> 00:48:46.690
domain. Can you say a little bit more about what
00:48:46.690 --> 00:48:50.409
locking it means? Makes it inaccessible worldwide.
00:48:51.989 --> 00:48:55.429
Oh. Yes, they can disconnect it. ICANN is a central
00:48:55.429 --> 00:48:59.050
point of failure. Yeah, for the visible internet,
00:48:59.230 --> 00:49:01.610
yeah. They can just suddenly decide to redirect
00:49:01.610 --> 00:49:03.070
it to a way, nope, you don't get to see this
00:49:03.070 --> 00:49:07.650
page. Yeah, yeah, I mean, so it's an interesting
00:49:07.650 --> 00:49:10.929
argument, but the reason, the thing that it led
00:49:10.929 --> 00:49:13.869
me to was, well, actually, you could turn, you
00:49:13.869 --> 00:49:18.909
know, a U .S. court could actually make the BBC
00:49:18.909 --> 00:49:25.119
inaccessible. Inaccessible, yeah. I don't think
00:49:25.119 --> 00:49:29.000
you can use the VPN argument to say you saw it,
00:49:29.019 --> 00:49:31.659
but you saw it over a VPN, so the problem is
00:49:31.659 --> 00:49:36.519
with a VPN. That won't stand up in court. I know
00:49:36.519 --> 00:49:39.480
what you're saying, David. I think they're saying,
00:49:39.719 --> 00:49:44.599
from the BBC point of view, they're saying we
00:49:44.599 --> 00:49:50.739
put this out only on these places. Anyone that
00:49:50.739 --> 00:49:55.219
saw it, because Trump is saying that that influence
00:49:55.219 --> 00:49:58.920
people outside of the UK. If Trump is arguing
00:49:58.920 --> 00:50:02.320
repressional damage within the US, the BBC can
00:50:02.320 --> 00:50:04.820
say no one in the US could have watched it. Yeah.
00:50:04.960 --> 00:50:06.539
We make sure they don't watch it. And anyone
00:50:06.539 --> 00:50:09.820
who did watch it, that's on them. No, no, no.
00:50:09.820 --> 00:50:12.260
That doesn't work. You can't say it didn't exist
00:50:12.260 --> 00:50:14.719
because it was legal. What they can say is that's
00:50:14.719 --> 00:50:19.440
a different issue. What they can say is that
00:50:19.440 --> 00:50:23.380
the potential audience in America for the program
00:50:23.380 --> 00:50:29.760
is not 350 Americans. It's the few thousands
00:50:29.760 --> 00:50:32.059
or hundreds, tens of thousands or hundreds of
00:50:32.059 --> 00:50:36.659
thousands who would use a VPN to... It says in
00:50:36.659 --> 00:50:39.380
the article it was played 34 times. That's right,
00:50:39.519 --> 00:50:45.159
yeah. It's not a large number. So, I mean, the
00:50:45.159 --> 00:50:48.559
point is that if the court decided the BBC was
00:50:48.559 --> 00:50:52.900
guilty, the potential damage is far less if it's
00:50:52.900 --> 00:50:56.860
34 people. Yeah. Oh, sorry. Let me rephrase that.
00:50:56.920 --> 00:50:59.539
It says the 34 Florida streams out. There was
00:50:59.539 --> 00:51:02.159
570 ,000 requests. Yeah, I've not read this article
00:51:02.159 --> 00:51:04.480
either. It's a bit confusing. I suppose they're
00:51:04.480 --> 00:51:06.219
saying they were trying to buy it in Florida.
00:51:07.039 --> 00:51:11.099
That might limit the damage. It's still a lot
00:51:11.099 --> 00:51:14.039
cheaper if they find the BBC guilty. Absolutely.
00:51:15.079 --> 00:51:18.480
I find it interesting, but I do also think about
00:51:18.480 --> 00:51:27.480
what precedent this sets about VPNs beyond this
00:51:27.480 --> 00:51:31.039
corporate. I think that if the BBC go with this
00:51:31.039 --> 00:51:36.440
argument, as Wendy said, it limits the... you
00:51:36.440 --> 00:51:40.739
know the damage um quite a lot but i also wonder
00:51:40.739 --> 00:51:45.539
about um what else this could could mean as well
00:51:45.539 --> 00:51:49.039
i don't i i i think legally it's interesting
00:51:49.039 --> 00:51:51.519
i think it's a big mistake to presume that any
00:51:51.519 --> 00:51:55.179
legal precedent is ever going to be set by a
00:51:55.179 --> 00:51:57.639
trump lawsuit because they are they are they
00:51:57.639 --> 00:52:01.599
are by i mean yes yeah but also It's an established
00:52:01.599 --> 00:52:04.539
pattern of harassment for the sake of harassment.
00:52:05.099 --> 00:52:08.059
That's not even, yes. But what is interesting
00:52:08.059 --> 00:52:13.800
is that, you know, the BBC geofences these things
00:52:13.800 --> 00:52:17.300
because, among other things, they don't, sometimes
00:52:17.300 --> 00:52:20.760
they use music that they only have the license
00:52:20.760 --> 00:52:25.139
for in the UK. And so there's a whole sort of
00:52:25.139 --> 00:52:28.539
copyright thing, you know. some sort of intellectual
00:52:28.539 --> 00:52:31.300
property control, which is that they did not
00:52:31.300 --> 00:52:34.639
publish it to the US. And yet these Americans
00:52:34.639 --> 00:52:37.360
have access to it. So there's a whole copyright
00:52:37.360 --> 00:52:39.960
thing in there somewhere. Yeah, but I mean, that
00:52:39.960 --> 00:52:41.840
won't get far. It's like saying, don't worry
00:52:41.840 --> 00:52:43.719
about porn. You can only get it in this country.
00:52:45.500 --> 00:52:47.460
There's a limit to how far you can use that.
00:52:48.320 --> 00:52:50.639
I mean, there's no question the amount of people
00:52:50.639 --> 00:52:52.619
who could actually see it is a solid argument.
00:52:52.840 --> 00:52:56.260
But beyond that, no. So, David, I know what you're
00:52:56.260 --> 00:53:00.559
saying. I think Wendy's actually right. So, as
00:53:00.559 --> 00:53:03.380
someone who's done this before for the BBC, sorry,
00:53:03.559 --> 00:53:06.659
and had to choose certain types of music, the
00:53:06.659 --> 00:53:11.260
lawyers have agreed that there's a certain level
00:53:11.260 --> 00:53:18.960
of people who, or, like I said, damages or, you
00:53:18.960 --> 00:53:24.099
know, use that. It's acceptable. That's just
00:53:24.099 --> 00:53:26.420
best effort. And if you're told to use best effort,
00:53:26.480 --> 00:53:28.900
that's fine. And obviously, if you're supposed
00:53:28.900 --> 00:53:31.260
to limit it, if you've only got the license for
00:53:31.260 --> 00:53:32.920
the UK, and as far as you're concerned, you're
00:53:32.920 --> 00:53:35.880
only releasing to the UK, you've achieved best
00:53:35.880 --> 00:53:39.059
effort. You can only use that so far. You certainly
00:53:39.059 --> 00:53:41.639
can't say, I've done my best effort to stop this
00:53:41.639 --> 00:53:44.679
horse porn being released. If it's seen outside
00:53:44.679 --> 00:53:51.239
of that area, that isn't going to work. And now
00:53:51.239 --> 00:53:55.000
we've got Kim. Coming with an article from MSN
00:53:55.000 --> 00:53:58.119
about a Berlin cyber attack and the hackers leaking
00:53:58.119 --> 00:54:00.860
highly sensitive data across the dark web. Go,
00:54:00.960 --> 00:54:04.559
Kim. Cool. Yeah, so there's a story that's been
00:54:04.559 --> 00:54:06.280
covered a little bit. It happened a couple of
00:54:06.280 --> 00:54:09.659
days ago by a hacker group called Rise SIDA,
00:54:09.820 --> 00:54:12.460
who have published six terabytes of data from
00:54:12.460 --> 00:54:15.440
Berlin State Administration. And I read a few
00:54:15.440 --> 00:54:18.730
articles around this, but it sounds like. To
00:54:18.730 --> 00:54:20.289
be honest, a lot of what I've heard about German
00:54:20.289 --> 00:54:23.250
bureaucracy is now coming to the front. I think
00:54:23.250 --> 00:54:25.269
they're running a lot of Windows 95 computers,
00:54:25.489 --> 00:54:27.349
Windows 98 computers that haven't been security
00:54:27.349 --> 00:54:31.329
patched for decades. They just didn't pay the
00:54:31.329 --> 00:54:33.829
ransom and it got leaked. And a lot of it just
00:54:33.829 --> 00:54:36.690
seems like this pigheadedness around, oh, we
00:54:36.690 --> 00:54:38.070
obviously won't get hacked with the government
00:54:38.070 --> 00:54:40.429
and now it has got hacked. But like a lot of
00:54:40.429 --> 00:54:42.329
it's really scary stuff. There's things in it
00:54:42.329 --> 00:54:45.829
like the chemical, biological, radiological and
00:54:45.829 --> 00:54:49.599
nuclear threats plans. And I mean, I saw a story
00:54:49.599 --> 00:54:50.860
a couple of weeks ago, you know, they're still
00:54:50.860 --> 00:54:53.460
finding bombs that are unexploded from World
00:54:53.460 --> 00:54:57.699
War II around Berlin. So it seems a kind of crazy
00:54:57.699 --> 00:54:59.699
situation. I think I accidentally talked about
00:54:59.699 --> 00:55:02.320
this in the AI chunk at the start as well. But
00:55:02.320 --> 00:55:05.059
yeah, I think it feels like at the moment, just
00:55:05.059 --> 00:55:06.940
these things are going to be on the rise as they
00:55:06.940 --> 00:55:08.719
get easier to do, right? It's kind of been the
00:55:08.719 --> 00:55:10.679
theme of the podcast that it's now a lot easier
00:55:10.679 --> 00:55:12.659
for bad actors to do bad things very quickly.
00:55:13.440 --> 00:55:15.159
but yeah it's kind of terrifying it's like i've
00:55:15.159 --> 00:55:16.920
got a lot of friends there and it's like is all
00:55:16.920 --> 00:55:19.440
their stuff in public now or that you know the
00:55:19.440 --> 00:55:21.420
stuff to do their civil servants birth certificates
00:55:21.420 --> 00:55:24.920
telephone numbers i just have the weirdest thought
00:55:24.920 --> 00:55:28.940
like i wonder if it's possible there's going
00:55:28.940 --> 00:55:31.440
to be a secondary market for training ai on this
00:55:31.440 --> 00:55:34.300
data because it is new data like they're like
00:55:34.300 --> 00:55:37.639
pay us the money or we sell it to we sell it
00:55:37.639 --> 00:55:41.179
to you know anthropic the interesting thing is
00:55:41.179 --> 00:55:43.570
though it's illegal to access So I've seen quite
00:55:43.570 --> 00:55:45.429
a lot of the reporting's just been like, I'm
00:55:45.429 --> 00:55:46.869
not going to link you to it and you shouldn't
00:55:46.869 --> 00:55:50.010
look for it because it's illegal. So I, yeah,
00:55:50.829 --> 00:55:53.409
it's hard to think about. And yet that weirdly
00:55:53.409 --> 00:55:55.210
almost makes me think more like that would be
00:55:55.210 --> 00:55:57.329
the perfect secondary market because not like
00:55:57.329 --> 00:55:58.769
they ever gave a shit in the first place with
00:55:58.769 --> 00:56:02.190
trying to train on that on legal data. But yeah,
00:56:02.230 --> 00:56:08.630
that is, that is, it, I wonder how much more
00:56:08.630 --> 00:56:13.909
of that we're going to see. A lot more. I feel
00:56:13.909 --> 00:56:20.349
like this is just the start of a really horrible
00:56:20.349 --> 00:56:27.739
place we're going to be. It's sad to say. it
00:56:27.739 --> 00:56:30.099
honestly i start putting in fake information
00:56:30.099 --> 00:56:32.519
on principle and sites if they're not i shouldn't
00:56:32.519 --> 00:56:34.400
say this out loud probably but you know if i'm
00:56:34.400 --> 00:56:36.139
like making like a something like a dating app
00:56:36.139 --> 00:56:38.599
profile or like some stupid recipe app once my
00:56:38.599 --> 00:56:40.400
date of birth i've just started lying on principle
00:56:40.400 --> 00:56:42.599
because i feel like any service you're on it's
00:56:42.599 --> 00:56:44.340
a certainty at some point it's getting leaked
00:56:44.340 --> 00:56:47.480
and obviously like states can force you to have
00:56:47.480 --> 00:56:49.460
this information so some of it's there but like
00:56:49.460 --> 00:56:55.059
it really makes you scared to put anything online
00:56:55.059 --> 00:56:58.510
right I'm with you. I'm with you. I, I just,
00:56:58.610 --> 00:57:04.650
I cannot put in real data. I just, I don't trust
00:57:04.650 --> 00:57:10.309
them to hold the data, to control the data. And
00:57:10.309 --> 00:57:13.150
I could just see floods of it just appearing
00:57:13.150 --> 00:57:15.769
all over the place. I have to put in fake data.
00:57:16.349 --> 00:57:20.190
It's horrible to say, but yeah. Also, I have,
00:57:20.190 --> 00:57:24.099
I have a weirdly. worth i the similar similar
00:57:24.099 --> 00:57:26.099
result but i'm talking of you which is like anytime
00:57:26.099 --> 00:57:28.159
somebody asks me for real data i'm like you pay
00:57:28.159 --> 00:57:31.460
me i know it's value right like like i'm not
00:57:31.460 --> 00:57:33.219
going to give you that for free like every time
00:57:33.219 --> 00:57:35.880
like i'll like people who all have like people
00:57:35.880 --> 00:57:38.800
from high school who will be like hey vote for
00:57:38.800 --> 00:57:40.820
us in this competition on facebook and i'm like
00:57:40.820 --> 00:57:42.719
you're gonna it's gonna make me sign up and give
00:57:42.719 --> 00:57:45.679
them all my data and i'm not gonna give you like
00:57:45.679 --> 00:57:48.570
that's not worth That that's that's way too high
00:57:48.570 --> 00:57:51.070
a cost for letting me vote in a stupid poll online
00:57:51.070 --> 00:57:52.849
where they know that really what they're doing
00:57:52.849 --> 00:57:55.550
is like, hey, if you want to win this, just,
00:57:55.610 --> 00:57:58.409
you know, hashtag us and make a public post on
00:57:58.409 --> 00:58:00.969
Instagram. I'm like, I'm not your free hype machine.
00:58:01.130 --> 00:58:04.269
Piss off like my way more than that. And it's
00:58:04.269 --> 00:58:06.469
weird how both of those attitudes have kind of
00:58:06.469 --> 00:58:08.099
come to the same thing where it's like, no. No,
00:58:08.139 --> 00:58:09.900
we've, we've come to both know our value, but
00:58:09.900 --> 00:58:12.860
also understand how deep down the rabbit hole,
00:58:12.940 --> 00:58:15.260
the potential downside of giving that goes, you
00:58:15.260 --> 00:58:17.000
know, it's like, well, it's just this name, buddy.
00:58:18.099 --> 00:58:20.380
Berlin. Like, seriously, take a, take a look
00:58:20.380 --> 00:58:24.659
at this particular what they did. You're just,
00:58:24.679 --> 00:58:26.039
you're going to take that. You're going to stick
00:58:26.039 --> 00:58:28.400
it in some little zip drive somewhere. And then
00:58:28.400 --> 00:58:32.159
two, you know, two years later. My God, you just
00:58:32.159 --> 00:58:38.090
dated yourself. You know what's really funny?
00:58:38.170 --> 00:58:39.730
A lot of people are going to think when I say
00:58:39.730 --> 00:58:42.010
zip drive that they're just talking about a USB
00:58:42.010 --> 00:58:46.170
drive, but I'm not. No, he's not. What are you
00:58:46.170 --> 00:58:47.389
talking about, a car? What are those things,
00:58:47.389 --> 00:58:50.269
hold 100 megabytes or something? They hold 100
00:58:50.269 --> 00:58:53.469
meg, and they had one of my favorite obscure
00:58:53.469 --> 00:58:57.269
points is that they were one of the fascinating
00:58:57.269 --> 00:59:01.050
examples of an actual physical hardware computer
00:59:01.050 --> 00:59:05.050
virus that was contagious. Yeah. Zip drive death.
00:59:05.769 --> 00:59:09.389
because the zip drive clicked out. Sorry, to
00:59:09.389 --> 00:59:13.710
get back to the point, it's hard because I can
00:59:13.710 --> 00:59:17.570
literally see it. I can see, I was thinking about
00:59:17.570 --> 00:59:21.489
the cost of securing the data, the cost of not
00:59:21.489 --> 00:59:25.090
and shitifying it. I just don't think there's
00:59:25.090 --> 00:59:28.349
any, there's very few companies that actually
00:59:28.349 --> 00:59:33.110
make that balance and they just don't do that.
00:59:33.309 --> 00:59:35.420
They're just kind of like, well, Yeah, we'll
00:59:35.420 --> 00:59:36.900
hold it for a certain amount of time and then
00:59:36.900 --> 00:59:38.360
we'll get bought by somebody else and then it'll
00:59:38.360 --> 00:59:41.559
get kind of sold on or it'll get leaked and we'll
00:59:41.559 --> 00:59:45.880
pay the fine and we'll just work by it. I just
00:59:45.880 --> 00:59:49.659
don't trust companies. That last point is really
00:59:49.659 --> 00:59:51.559
important. That thing about they'll get bought
00:59:51.559 --> 00:59:54.380
is one of the things where it's like, even if
00:59:54.380 --> 00:59:58.159
I believe that you are sincere, even if I believe
00:59:58.159 --> 01:00:01.420
that you are sincere, you will get bought. Is
01:00:01.420 --> 01:00:04.099
anyone working on solutions to this? i periodically
01:00:04.099 --> 01:00:06.980
look up that tim berners -lee solid protocol
01:00:06.980 --> 01:00:13.219
and then the other one was i heard like um oh
01:00:13.219 --> 01:00:15.300
which eastern european country is it that's supposed
01:00:15.300 --> 01:00:18.039
to have amazing blockchain -based public internet
01:00:18.039 --> 01:00:21.940
infrastructure obviously estonia i'm like Because
01:00:21.940 --> 01:00:23.659
it just seems crazy, you know, and there's a
01:00:23.659 --> 01:00:25.420
lot of chat of like digital sovereignty at the
01:00:25.420 --> 01:00:26.960
moment, which always seems to mean like state
01:00:26.960 --> 01:00:29.019
sovereignty or European sovereignty. It's never
01:00:29.019 --> 01:00:31.800
like personal sovereignty of your own data. But
01:00:31.800 --> 01:00:33.500
it really feels like we're in this place now
01:00:33.500 --> 01:00:36.179
where like, it's exactly like you just said,
01:00:36.199 --> 01:00:38.039
if there's a web form that's asking your demographic
01:00:38.039 --> 01:00:40.760
detail, you know, I used to fill that in without
01:00:40.760 --> 01:00:42.619
even thinking about it. But nowadays, I think
01:00:42.619 --> 01:00:45.300
also we have to fill in so many now, right? I
01:00:45.300 --> 01:00:47.099
feel like we're doing all of our own admin and
01:00:47.099 --> 01:00:49.690
everything's manual and like... You know, I've
01:00:49.690 --> 01:00:50.949
just been trying to get a blood test for the
01:00:50.949 --> 01:00:52.889
last four months and it's just like filling in
01:00:52.889 --> 01:00:55.130
the same stupid doctor form five times over.
01:00:55.449 --> 01:01:01.250
The data creep of them asking for way more information
01:01:01.250 --> 01:01:02.929
than they clearly need for that just because
01:01:02.929 --> 01:01:05.090
it's going to be valuable for them, not for you.
01:01:05.389 --> 01:01:07.570
And it takes more of my time and more of my effort.
01:01:08.809 --> 01:01:10.690
And it's pretty easy to quantify too, right?
01:01:10.750 --> 01:01:13.210
I mean, if anyone's ever had a test, if anyone's
01:01:13.210 --> 01:01:15.380
ever been in Tesco. You don't want to shop in
01:01:15.380 --> 01:01:17.420
Tesco without giving them all your data, which
01:01:17.420 --> 01:01:20.059
they call a Tesco club card. It's like it's got
01:01:20.059 --> 01:01:24.800
a name. You know, this is where it's an advantage.
01:01:25.139 --> 01:01:27.019
This is where rich people have the advantage.
01:01:27.159 --> 01:01:29.760
They can afford not to have the Tesco club card.
01:01:30.219 --> 01:01:33.219
Yeah. I don't know if I have the advantage where
01:01:33.219 --> 01:01:36.940
they can. I did the old phrase of fines are just
01:01:36.940 --> 01:01:40.380
fines just mean their laws for poor people. Like.
01:01:42.480 --> 01:01:45.320
If I mean, and especially in a corporate in a
01:01:45.320 --> 01:01:47.840
corporate stakeholder, you know, shareholder
01:01:47.840 --> 01:01:50.179
methodology where corporate is, it's built in
01:01:50.179 --> 01:01:52.500
that you have to make the decision that returns
01:01:52.500 --> 01:01:55.519
the greatest value. If the equation is putting
01:01:55.519 --> 01:01:57.659
babies on spikes will earn us one hundred and
01:01:57.659 --> 01:01:59.519
one million dollars, but will cost us a million
01:01:59.519 --> 01:02:01.559
dollars and a hundred millions in fines. They'll
01:02:01.559 --> 01:02:03.280
look they'll look at that and go, well, babies
01:02:03.280 --> 01:02:05.119
on spikes equals a million dollars profit. Let's
01:02:05.119 --> 01:02:08.599
go. Yeah, absolutely. Cost of the spikes, too.
01:02:10.460 --> 01:02:13.280
Babies of spikes. We did. That's one of my go
01:02:13.280 --> 01:02:22.960
-to metaphors. And we got one more article. We've
01:02:22.960 --> 01:02:26.940
got Wendy's come with an interesting story about
01:02:26.940 --> 01:02:30.519
the Trump administration has kind of taken out
01:02:30.519 --> 01:02:34.079
an Italian service provider, Astici. I think
01:02:34.079 --> 01:02:37.519
the full name is Astici Inventati. And I had
01:02:37.519 --> 01:02:40.880
not heard of it until this case. The Trump administration
01:02:40.880 --> 01:02:44.360
has dubbed it a terrorist, put it on a terrorist
01:02:44.360 --> 01:02:47.019
watch list. And apparently the claim is that
01:02:47.019 --> 01:02:51.099
by providing secure email and services to, I
01:02:51.099 --> 01:02:53.840
think it's about 20 ,000 activist organizations
01:02:53.840 --> 01:02:56.639
around the world, that it's providing material
01:02:56.639 --> 01:03:02.940
comfort or something to terrorists. And the consequence
01:03:02.940 --> 01:03:06.179
of this that's interesting is it took within
01:03:06.179 --> 01:03:09.139
like two or three days. The service announced
01:03:09.139 --> 01:03:11.699
it was completely shutting down because basically
01:03:11.699 --> 01:03:16.159
it was completely crippled. Its domain was suspended
01:03:16.159 --> 01:03:19.599
and it can't get any payments from anybody. But
01:03:19.599 --> 01:03:21.820
I think its real reason is that it can't keep
01:03:21.820 --> 01:03:24.719
people safe the way it promised. So it's gone.
01:03:25.440 --> 01:03:30.980
And, you know, it's not a good, it's not, whatever
01:03:30.980 --> 01:03:32.920
you think about the site or the organizations
01:03:32.920 --> 01:03:37.500
it helped support, that doesn't matter. compared
01:03:37.500 --> 01:03:43.139
to the problem that one country can use the internet
01:03:43.139 --> 01:03:47.099
infrastructure to basically kill off anything
01:03:47.099 --> 01:03:51.199
it doesn't like so kim you had some comments
01:03:51.199 --> 01:03:53.940
yeah no i was at the the anarchist studies network
01:03:53.940 --> 01:03:56.099
conference was in manchester last weekend and
01:03:56.099 --> 01:03:58.239
i was out there doing a talk about kind of this
01:03:58.619 --> 01:04:00.800
the need to move from corporate tech to kind
01:04:00.800 --> 01:04:02.739
of our own things again and it was a very spooky
01:04:02.739 --> 01:04:04.940
thing to be happening in the background because
01:04:04.940 --> 01:04:07.079
my you know my my understanding of this group
01:04:07.079 --> 01:04:08.619
which i don't think i've ever really said out
01:04:08.619 --> 01:04:11.400
loud but i still kind of fail is like you know
01:04:11.400 --> 01:04:13.559
they they run no blogs which a lot of sort of
01:04:13.559 --> 01:04:15.760
like anarchist blogs run on they provide email
01:04:15.760 --> 01:04:18.480
for a few people i know but the idea of thinking
01:04:18.480 --> 01:04:20.099
of them as anything like a threat is sort of
01:04:20.099 --> 01:04:22.500
absurd i mean you know it's like riseup .net
01:04:22.500 --> 01:04:25.699
would be the sort of you know um english language
01:04:25.699 --> 01:04:28.019
speaking equivalent or there's other ones in
01:04:28.019 --> 01:04:31.320
you know like activix and it's just a very scary
01:04:31.320 --> 01:04:34.800
thing because yeah i think on top of the um the
01:04:34.800 --> 01:04:36.699
american threat it's also just like you know
01:04:36.699 --> 01:04:40.179
where's the lack of defense from european legislature
01:04:40.179 --> 01:04:43.380
towards this kind of thing and and it sort of
01:04:43.380 --> 01:04:45.860
feels very much like you know you're talking
01:04:45.860 --> 01:04:48.079
about test cases ian and where things might go
01:04:48.079 --> 01:04:50.420
but it really feels like a test case to me because
01:04:50.420 --> 01:04:52.039
i feel like the direction we're going in at the
01:04:52.039 --> 01:04:56.940
moment is that the the web's been colonized by
01:04:56.940 --> 01:04:59.900
massive corporations and actually like they're
01:04:59.900 --> 01:05:01.820
colluding with governments to make it so that
01:05:01.820 --> 01:05:04.360
not being a gigantic corporation makes it very
01:05:04.360 --> 01:05:06.599
hard to publish stuff online so i think this
01:05:06.599 --> 01:05:08.199
has now been enforced with things like the online
01:05:08.199 --> 01:05:11.780
safety act which basically subject things like
01:05:11.780 --> 01:05:15.139
cycling forums to the same kind of public safeguards
01:05:15.139 --> 01:05:17.019
that facebook do there's you know it's still
01:05:17.019 --> 01:05:19.699
kind of out in the open but this sort of feels
01:05:19.699 --> 01:05:21.800
like part of this project to me it feels like
01:05:21.800 --> 01:05:24.309
there's kind of an attack on the open web And
01:05:24.309 --> 01:05:26.170
where they're always going to start is with the
01:05:26.170 --> 01:05:28.570
sort of more out there political stuff, you know?
01:05:28.730 --> 01:05:32.150
So I'm very scared where this one's leading,
01:05:32.289 --> 01:05:35.429
less so than the other one, I think. What's that
01:05:35.429 --> 01:05:37.889
famous thing? They came for me, they came for
01:05:37.889 --> 01:05:39.969
you, or I don't know, they came for all of us
01:05:39.969 --> 01:05:40.969
or something like that? I can't remember what
01:05:40.969 --> 01:05:42.530
it is, but yeah. I think you've got that exactly
01:05:42.530 --> 01:05:46.570
backwards, but yes. Okay, sorry. Sorry, don't
01:05:46.570 --> 01:05:48.250
make me laugh. But yeah, no, I completely agree.
01:05:51.610 --> 01:05:56.170
When you talked about sovereignty before, yeah,
01:05:57.590 --> 01:06:01.809
I just, it's really hard because I also, I did
01:06:01.809 --> 01:06:04.849
work with solid and personal data stores and
01:06:04.849 --> 01:06:06.690
stuff like that. And there's an emphasis on like
01:06:06.690 --> 01:06:09.090
putting all the pressure on you to do all the
01:06:09.090 --> 01:06:14.469
work. And it feels like there's got to be a better
01:06:14.469 --> 01:06:19.940
way, full stop. And I think that applies. in
01:06:19.940 --> 01:06:24.059
what you personally do, but also the sites that
01:06:24.059 --> 01:06:26.880
you use, the sites that are community driven.
01:06:27.880 --> 01:06:33.219
I just feel like, I do agree that the web has
01:06:33.219 --> 01:06:36.739
become this kind of, this corporate playground.
01:06:37.840 --> 01:06:42.039
And it's a real shame. There are still corners
01:06:42.039 --> 01:06:45.019
of the web where, but they are being pushed aside
01:06:45.019 --> 01:06:50.269
or kind of like. legalized, pushed out by legal
01:06:50.269 --> 01:06:56.469
things. Has anyone tried to self -host email
01:06:56.469 --> 01:07:01.329
recently? You kind of can't. I gave up three
01:07:01.329 --> 01:07:06.769
years ago, finally. I like the idea that we use
01:07:06.769 --> 01:07:09.190
a rise -up pad, therefore we're terrorists. This
01:07:09.190 --> 01:07:12.070
is a terrorist podcast. That's the thing. With
01:07:12.070 --> 01:07:16.420
a service provider like... what you would have
01:07:16.420 --> 01:07:19.480
expected is if there was a terrorist organization
01:07:19.480 --> 01:07:23.820
that they were serving, that that organization
01:07:23.820 --> 01:07:27.679
would be, you know, they'd be asked to take that
01:07:27.679 --> 01:07:31.519
organization down, right? And that organization
01:07:31.519 --> 01:07:34.199
would be the target of the action, not the service
01:07:34.199 --> 01:07:36.400
provider. I mean, you don't see anybody going
01:07:36.400 --> 01:07:41.409
after Facebook. You know, you don't see anybody
01:07:41.409 --> 01:07:43.670
trying to take out all of Facebook because there's
01:07:43.670 --> 01:07:45.869
a few people on it that, you know. Oh, but I
01:07:45.869 --> 01:07:47.829
mean, they position themselves as service provision
01:07:47.829 --> 01:07:51.110
for activists and anarchists. So I assume that
01:07:51.110 --> 01:07:54.250
they're basically Trump enemies by definition.
01:07:54.769 --> 01:07:58.710
The way they're going after. Normally you would
01:07:58.710 --> 01:08:02.829
have due process for each organization. You wouldn't
01:08:02.829 --> 01:08:06.349
be aiming at the service provider. Unless the
01:08:06.349 --> 01:08:10.480
service provider itself was really doing. Right.
01:08:10.480 --> 01:08:12.639
And I assume that's the argument they're making,
01:08:12.760 --> 01:08:15.360
that, look, you're basically like WordPress .com
01:08:15.360 --> 01:08:17.640
but for terrorism, you know, which is just such
01:08:17.640 --> 01:08:19.060
an insane thing to say. Wow, it's a niche market.
01:08:19.199 --> 01:08:21.899
I don't see the problem. It's like if you read
01:08:21.899 --> 01:08:24.619
those blogs on there, it's like it's impenetrable
01:08:24.619 --> 01:08:27.180
anarchist essays about how we should be organising
01:08:27.180 --> 01:08:28.960
things that I don't get. I think terrorists need
01:08:28.960 --> 01:08:32.359
to write bad documents too. Right, yeah. They
01:08:32.359 --> 01:08:41.840
absolutely do. Yeah. Yeah, I... Okay, so I know
01:08:41.840 --> 01:08:44.979
it's really easy to kind of like just put our
01:08:44.979 --> 01:08:47.420
heads down and go, oh, my God, it's doom and
01:08:47.420 --> 01:08:49.560
gloom. And I know that TechCrunch is meant to
01:08:49.560 --> 01:08:53.380
be that. But I do think that there is, I have
01:08:53.380 --> 01:08:59.659
seen some lights where, yeah, a lot of the focus
01:08:59.659 --> 01:09:04.500
around peer -to -peer systems. around sovereign,
01:09:04.640 --> 01:09:06.960
you know, not just sovereignty as in, I know
01:09:06.960 --> 01:09:08.859
what you're saying, Kim, where there's a focus
01:09:08.859 --> 01:09:11.119
on like sovereignty. It feels like sovereignty
01:09:11.119 --> 01:09:14.180
to a country, but there are a lot more systems
01:09:14.180 --> 01:09:16.859
which are much more focused on, you know, sovereignty
01:09:16.859 --> 01:09:20.640
to a group. I think that that has started to
01:09:20.640 --> 01:09:23.399
grow. I noticed that there's a lot more projects
01:09:23.399 --> 01:09:27.880
in the open source world, which are focused around
01:09:27.880 --> 01:09:30.979
this. I see frameworks, which are like build
01:09:30.979 --> 01:09:34.100
it. top of disk. This is how to do it properly
01:09:34.100 --> 01:09:37.300
rather than make your own kind of like peer -to
01:09:37.300 --> 01:09:39.560
-peer scrappy thing, which is completely open.
01:09:39.819 --> 01:09:43.880
How many of these things are dependent on the
01:09:43.880 --> 01:09:48.460
domain name system? Yeah. Okay. That's not actually,
01:09:48.560 --> 01:09:52.199
I've seen a few people that have said that they're
01:09:52.199 --> 01:09:57.220
deliberately not. I mean, things like IRC are
01:09:57.220 --> 01:09:58.920
not actually dependent. I don't think they're
01:09:58.920 --> 01:10:02.359
dependent on the domain name system. But the
01:10:02.359 --> 01:10:06.239
web is. Anything on the web is. Yeah, I've seen,
01:10:06.380 --> 01:10:09.199
I know what you're saying, Wendy, and I think
01:10:09.199 --> 01:10:12.819
I have seen somewhere that deliberately not using
01:10:12.819 --> 01:10:15.659
domain system because of the exact point you're
01:10:15.659 --> 01:10:19.920
making. But I do think that they're still kind
01:10:19.920 --> 01:10:22.300
of under the radar and they're still not the
01:10:22.300 --> 01:10:27.140
adoption. I do worry that a lot of developers
01:10:27.140 --> 01:10:29.359
just kind of go, right, I'm going to solve this
01:10:29.359 --> 01:10:32.920
problem again, rather than, Look and try and
01:10:32.920 --> 01:10:36.779
find systems that do the thing correctly. So,
01:10:36.859 --> 01:10:42.239
for example, and then I'll end, is when they
01:10:42.239 --> 01:10:45.699
were talking about the European social stack,
01:10:45.819 --> 01:10:49.859
which I talked about a few tech months ago, rather
01:10:49.859 --> 01:10:53.220
than kind of focus on creating a new kind of
01:10:53.220 --> 01:10:57.600
stack, they used existing technologies, and the
01:10:57.600 --> 01:11:01.609
one for messaging was Matrix. Because Matrix
01:11:01.609 --> 01:11:03.949
is actually pretty good and it's open and it
01:11:03.949 --> 01:11:07.329
has all the things rather than trying to build
01:11:07.329 --> 01:11:09.630
something different. And I think we need more
01:11:09.630 --> 01:11:12.729
of that. We need more people to kind of not roll
01:11:12.729 --> 01:11:16.590
their own, but actually use ones or contribute
01:11:16.590 --> 01:11:21.449
to ones that are doing the right thing. Peace
01:11:21.449 --> 01:11:24.369
out. Do we think the EU can make a good Matrix
01:11:24.369 --> 01:11:31.640
client? Let's find out. And that's the last of
01:11:31.640 --> 01:11:34.460
our news stories for here on Tech Grumps. The
01:11:34.460 --> 01:11:36.300
only thing left for us to do is the Any Other
01:11:36.300 --> 01:11:38.159
Business. And the first thing on the list is
01:11:38.159 --> 01:11:41.470
Wendy with Coyote vs. Acme. I haven't seen it
01:11:41.470 --> 01:11:43.810
yet. I'm just really pleased that it's actually
01:11:43.810 --> 01:11:46.449
been released, given that Warner Brothers wanted
01:11:46.449 --> 01:11:48.590
to take it as a tax write off and was going to
01:11:48.590 --> 01:11:52.010
take a completed movie and say, fuck you to everybody
01:11:52.010 --> 01:11:54.829
who worked on it and refused to release it because
01:11:54.829 --> 01:11:57.069
they wanted a tax write off. I mean, where is
01:11:57.069 --> 01:11:59.510
Douglas Adams when you need him? Yeah, I mean,
01:11:59.510 --> 01:12:02.090
it's giving me it's actually giving me joy. I've
01:12:02.090 --> 01:12:03.529
been it's been making me legitimately happy to
01:12:03.529 --> 01:12:05.609
see that they won this battle. And the buzz is
01:12:05.609 --> 01:12:08.979
good. Yeah, people are saying it's a perfectly
01:12:08.979 --> 01:12:12.859
good film. Very funny, very well done. Ian, you've
01:12:12.859 --> 01:12:15.979
got something about Cory Doctorow? Yeah, so Cory's
01:12:15.979 --> 01:12:21.500
in Manchester, Manchester, UK, not Alabama. And
01:12:21.500 --> 01:12:26.960
he's talking about the reverse centaur. But I
01:12:26.960 --> 01:12:30.779
want people to know and go along. I think it's
01:12:30.779 --> 01:12:34.500
a highly accessible way of talking about AI.
01:12:34.760 --> 01:12:37.800
And I think more people should go along. Okay,
01:12:37.859 --> 01:12:40.779
it's the London version. And I've got something
01:12:40.779 --> 01:12:42.220
that I just noticed, a little thing that gave
01:12:42.220 --> 01:12:44.939
me joy this week was watching an evolution of
01:12:44.939 --> 01:12:47.439
monetization sabotage, which was there's been
01:12:47.439 --> 01:12:50.159
this thing about how to deal with basically content
01:12:50.159 --> 01:12:52.199
botherers, people who do conflict content and
01:12:52.199 --> 01:12:53.659
want to walk up on you. And one of the things
01:12:53.659 --> 01:12:55.500
that I used to hear people say is things like,
01:12:55.520 --> 01:12:58.220
just start singing a Disney tune. because then
01:12:58.220 --> 01:12:59.439
they can't put the video out because of that.
01:12:59.500 --> 01:13:00.920
And someone came up with an even better idea,
01:13:01.020 --> 01:13:02.640
which is the moment they do that, they go, this
01:13:02.640 --> 01:13:06.159
video was sponsored by Squarespace. And so, and
01:13:06.159 --> 01:13:09.340
apparently not only does this work, but the TikTok
01:13:09.340 --> 01:13:11.859
that I saw of the person who talked about this
01:13:11.859 --> 01:13:14.119
had to come back with a second TikTok and going,
01:13:14.140 --> 01:13:19.010
so. i managed to get my own video taken down
01:13:19.010 --> 01:13:21.069
by using that technique because they also they
01:13:21.069 --> 01:13:23.590
didn't realize i was saying it in sarcasm and
01:13:23.590 --> 01:13:24.989
so they took down my video thinking that it was
01:13:24.989 --> 01:13:26.829
not that was an unlisted advertisement and i
01:13:26.829 --> 01:13:29.529
was like i love that i love that it's so effective
01:13:29.529 --> 01:13:32.010
you shot yourself like i've i've admitted a new
01:13:32.010 --> 01:13:35.279
gun shoot themselves in the foot also good sign
01:13:35.279 --> 01:13:37.680
that it works so hey next time you want to deal
01:13:37.680 --> 01:13:39.119
with someone who's making content in front of
01:13:39.119 --> 01:13:40.340
you they're not sure you're consenting about
01:13:40.340 --> 01:13:42.079
just start talking about how they're sponsored
01:13:42.079 --> 01:13:44.060
by squarespace or whatever else you want because
01:13:44.060 --> 01:13:46.000
apparently that's the new way to make the money
01:13:46.000 --> 01:13:49.159
to make to make the video non -monetizable and
01:13:49.159 --> 01:13:53.060
that gave me delight have you just screwed up
01:13:53.060 --> 01:13:56.550
no I was just going to say, is there anything
01:13:56.550 --> 01:13:58.989
more unhinged than intellectual property law
01:13:58.989 --> 01:14:01.130
right now? I feel like there is nothing more
01:14:01.130 --> 01:14:03.369
inconsistent in the world than the way IP law
01:14:03.369 --> 01:14:06.289
is being applied. I honestly think anyone trying
01:14:06.289 --> 01:14:08.930
to make an IP claim in the age of AI stealing
01:14:08.930 --> 01:14:12.449
everything is just like the hubris is astonishing.
01:14:13.750 --> 01:14:16.949
100%. And that's it for another Tech Grumps.
01:14:17.920 --> 01:14:19.859
Please remember, if you'd like to join us on
01:14:19.859 --> 01:14:22.600
Tech Grumps, please just use the hashtag Tech
01:14:22.600 --> 01:14:25.760
Grumps hashtag. I said it twice. We will find
01:14:25.760 --> 01:14:28.479
you on Mastodon and other locations. The only
01:14:28.479 --> 01:14:31.000
thing left to do is to say goodbye from Ian.
01:14:32.409 --> 01:14:40.010
Bye -bye. Wendy. Bye. David. Goodbye. Kim. Goodbye.
01:14:40.409 --> 01:14:42.890
Thank you for joining us, Kim, and myself. I'm
01:14:42.890 --> 01:14:44.989
Ryan. Thank you for joining us for another Tech
01:14:44.989 --> 01:14:45.329
Grumps.