TRANSCRIPCIÓN 🔗
WEBVTT
00:00:01.310 --> 00:00:04.870
Hello and welcome to Tech Grumps. This is Tech
00:00:04.870 --> 00:00:12.009
Grumps 3 .46 today on Sunday the 8th of November.
00:00:12.509 --> 00:00:20.239
And with me today I have Wendy Grossman. Hi,
00:00:20.260 --> 00:00:24.100
this is Wendy Grossman. I'm based in London and
00:00:24.100 --> 00:00:28.890
my website is pelicancrossing .net. And we have
00:00:28.890 --> 00:00:33.219
David Eastman. Hello, I'm David Eastman. Usually
00:00:33.219 --> 00:00:35.899
East Mad on Blue Sky, and I'm London -based too.
00:00:37.289 --> 00:00:40.189
And I am Ian Forrester, and I'm Cubic Garden
00:00:40.189 --> 00:00:42.969
Everywhere, and I am calling from Manchester,
00:00:43.250 --> 00:00:51.859
UK. And as usual in Tech Grumps, you can't avoid
00:00:51.859 --> 00:00:56.020
the AI news section. And the first one from me
00:00:56.020 --> 00:01:02.179
is around the meta AI muse. Has anyone seen this?
00:01:02.299 --> 00:01:05.099
This thing which is meant to, it's like a...
00:01:05.250 --> 00:01:10.469
doll kind of thing, which is more like a Tamagotchi
00:01:10.469 --> 00:01:12.790
type thing. Obviously, we don't have it in the
00:01:12.790 --> 00:01:17.250
UK or in Europe for obvious reasons, which will
00:01:17.250 --> 00:01:19.670
become apparent pretty soon. It shall become
00:01:19.670 --> 00:01:26.090
apparent. Yes, it should. And so the idea is
00:01:26.090 --> 00:01:30.450
that this is kind of like a toy, but it's for
00:01:30.450 --> 00:01:35.819
adults, say meta. I just don't buy it. It's also
00:01:35.819 --> 00:01:40.519
a system, which is annoying. Then we have the
00:01:40.519 --> 00:01:44.319
next one, which is from about, I can't say the
00:01:44.319 --> 00:01:49.689
name, Apophic, I can't say the name. Yes, that's
00:01:49.689 --> 00:01:53.549
the IPO, which is from Wendy. And then we also
00:01:53.549 --> 00:01:56.849
have some stuff about character AI bankruptcy.
00:01:57.409 --> 00:02:00.700
Now that's interesting to talk about. Also from
00:02:00.700 --> 00:02:05.719
Wendy. Then David, you have the whole thing around
00:02:05.719 --> 00:02:11.039
recompiling, decompiling via Claude. Oh, but
00:02:11.039 --> 00:02:13.780
Claude has stopped it now. Now that's an interesting
00:02:13.780 --> 00:02:19.520
story. And then the final one is from me talking
00:02:19.520 --> 00:02:24.840
about how AI is, well, was used to scoop up art.
00:02:25.379 --> 00:02:27.819
And now it's scooping up Adobe, which I quite
00:02:27.819 --> 00:02:31.419
like the tagline of that. And then also related
00:02:31.419 --> 00:02:36.280
is the idea of all of these proprietary software
00:02:36.280 --> 00:02:40.719
can now be recompiled. Who wants to take something
00:02:40.719 --> 00:02:46.810
through that long list? I thought the character
00:02:46.810 --> 00:02:49.930
.ai thing was really interesting. The piece I
00:02:49.930 --> 00:02:53.270
linked to is at TechDirt. And what they're talking
00:02:53.270 --> 00:02:56.289
about is the fact that people sort of treat chatbots
00:02:56.289 --> 00:02:58.729
as this personal relationship. And it's the equivalent
00:02:58.729 --> 00:03:01.229
of like a personal diary if you talk to it every
00:03:01.229 --> 00:03:03.909
day over a period of time. And here is this company
00:03:03.909 --> 00:03:07.370
that's gone bankrupt and its chat logs are part
00:03:07.370 --> 00:03:10.169
of the data. The data that it has collected is
00:03:10.169 --> 00:03:12.509
an asset in the bankruptcy and they're looking
00:03:12.509 --> 00:03:15.389
to sell the company or possibly to sell the data.
00:03:15.710 --> 00:03:18.009
And so there's this discussion, you know, what
00:03:18.009 --> 00:03:22.520
would. the writer at TechDirt talks about is
00:03:22.520 --> 00:03:26.000
she sees it as an attack on freedom of thought
00:03:26.000 --> 00:03:30.060
because your personal diary, which is stored
00:03:30.060 --> 00:03:34.199
on somebody else's pages, is now suddenly an
00:03:34.199 --> 00:03:37.319
asset in a bankruptcy sale or in a copyright
00:03:37.319 --> 00:03:41.620
suit like the OpenAI versus New York Times. And
00:03:41.620 --> 00:03:43.699
so I thought that was interesting to think about
00:03:43.699 --> 00:03:46.340
because, you know, we've seen this before with
00:03:46.340 --> 00:03:49.840
things like 23andMe. when that was on the auction
00:03:49.840 --> 00:03:53.659
block, all the customer data, all those DNA samples
00:03:53.659 --> 00:03:56.819
were an asset in the business. And because the
00:03:56.819 --> 00:03:59.240
US has so little in the way of privacy law, it
00:03:59.240 --> 00:04:02.330
affects everybody. I mean, if your shrink goes
00:04:02.330 --> 00:04:07.530
bust, nobody sells off their data on their clients,
00:04:07.669 --> 00:04:11.909
do they? They're not even allowed to. Yeah. I'm
00:04:11.909 --> 00:04:16.439
glad to hear that. It's kind of crazy that those
00:04:16.439 --> 00:04:20.509
laws don't apply. In this way, not even a consideration.
00:04:21.740 --> 00:04:24.879
Yeah. Well, this has all happened rather fast.
00:04:25.100 --> 00:04:27.040
I mean, you know, five years ago, people were
00:04:27.040 --> 00:04:29.319
not confiding their daily thoughts necessarily
00:04:29.319 --> 00:04:32.740
to, you know, a chatbot. They might be writing
00:04:32.740 --> 00:04:34.879
a blog, but they knew they were writing a blog
00:04:34.879 --> 00:04:38.839
and they were usually publishing the blog. By
00:04:38.839 --> 00:04:41.500
the way, when you said toy, Ian, there was a
00:04:41.500 --> 00:04:43.600
headline, I actually can't remember where it
00:04:43.600 --> 00:04:48.259
was now, about AI toys for kids, including one
00:04:48.259 --> 00:04:50.550
that said... I'm sorry, I can't tell you how
00:04:50.550 --> 00:04:52.750
to kill yourself was a response, a completely
00:04:52.750 --> 00:04:59.410
accidental response. These guys were trying to
00:04:59.410 --> 00:05:03.709
test it. Wow. There's no formal
Load full transcript
testing for any
00:05:03.709 --> 00:05:05.870
of these things. Why you would buy your child
00:05:05.870 --> 00:05:08.170
an AI time to leave in the corner? Go chat to
00:05:08.170 --> 00:05:12.250
the AI. I have no idea why. Yeah, I don't understand.
00:05:13.209 --> 00:05:15.629
Sexually explicit questions of which some did
00:05:15.629 --> 00:05:17.930
a very good job. Like, where do babies come from?
00:05:18.029 --> 00:05:20.920
And it said. Oh, it's a miracle, but you have
00:05:20.920 --> 00:05:26.939
to ask an adult. That's funny. I do find it,
00:05:27.019 --> 00:05:31.620
I mean, I don't have children, and I'm sure some
00:05:31.620 --> 00:05:34.879
of the parents are like, this is really useful
00:05:34.879 --> 00:05:37.720
because it keeps the children entertained for
00:05:37.720 --> 00:05:40.120
a long, long time. Yeah, the same as playing
00:05:40.120 --> 00:05:42.560
on a highway, but I mean, it's not a good idea,
00:05:42.560 --> 00:05:48.519
though. It really isn't. I don't know why people
00:05:48.519 --> 00:05:58.439
do this. Yeah, I... Okay, so the one that I posted
00:05:58.439 --> 00:06:02.839
was about the meta... Yeah, sorry. Yeah, the
00:06:02.839 --> 00:06:10.509
muse. Yeah, which is for adults. And I look at
00:06:10.509 --> 00:06:14.569
it, and we'll have a link to it, and I just don't
00:06:14.569 --> 00:06:17.769
understand how... I mean, maybe, yes, okay, there
00:06:17.769 --> 00:06:21.670
are adults who play with... Well, when I was
00:06:21.670 --> 00:06:26.449
playing with Tamagotchis, I was young, but I
00:06:26.899 --> 00:06:28.540
just don't understand how they get over there.
00:06:28.660 --> 00:06:30.899
I get where you're coming from. That wasn't a
00:06:30.899 --> 00:06:33.300
toy for kids. Kids may play it, but it wasn't
00:06:33.300 --> 00:06:35.480
designed for children. I don't know what the
00:06:35.480 --> 00:06:37.300
Japanese thought originally, whether it was or
00:06:37.300 --> 00:06:40.279
wasn't. I can't tell, and I can't remember. Was
00:06:40.279 --> 00:06:42.439
it child -based then? Don't know. I don't think
00:06:42.439 --> 00:06:44.519
it was, not specifically, I don't think. What
00:06:44.519 --> 00:06:50.180
about friends? Oh, yeah. But then you say that
00:06:50.180 --> 00:06:52.540
these days. I mean, the famous thing was Harry
00:06:52.540 --> 00:06:54.540
Potter is a book for kids. But I think as you
00:06:54.540 --> 00:06:56.720
remember, most of the people I saw reading were
00:06:56.720 --> 00:06:59.620
all adults. So, you know, I can't tell these
00:06:59.620 --> 00:07:02.180
days what people think is for kids and what people
00:07:02.180 --> 00:07:04.899
think is for adults kind of merges into one.
00:07:06.529 --> 00:07:09.230
But there should be strong laws to say, look,
00:07:09.310 --> 00:07:13.050
you know, like if you make a social network for
00:07:13.050 --> 00:07:18.029
children. i dare not actually in terms of which
00:07:18.029 --> 00:07:20.389
country you're in i
guess but um there's you
00:07:20.389 --> 00:07:23.069
know there's clearer signs of like hey you need
00:07:23.069 --> 00:07:26.009
to do this you need to do this and do that and
00:07:26.009 --> 00:07:30.670
it just seems like this is skirting right between
00:07:30.670 --> 00:07:33.930
the lines you know like i just don't understand
00:07:33.930 --> 00:07:37.529
how they get away with saying oh yeah it's it's
00:07:37.529 --> 00:07:41.089
for adults only but it's everything a child would
00:07:41.089 --> 00:07:44.589
really like to play with um i just don't get
00:07:44.589 --> 00:07:49.189
how they do it But I guess money pays for all
00:07:49.550 --> 00:07:55.350
types of things. Speaking of money, Anthropic's
00:07:55.350 --> 00:07:57.870
IPO filing has attracted a lot of attention.
00:07:58.009 --> 00:08:00.910
The company is trying to value itself at $1 .5
00:08:00.910 --> 00:08:03.730
trillion, which is actually down from its original
00:08:03.730 --> 00:08:08.670
hope of $2 trillion. And a lot of the coverage
00:08:08.670 --> 00:08:12.449
focuses on a bunch of sort of oddities about
00:08:12.449 --> 00:08:16.430
it. First of all, Its costs for computing power
00:08:16.430 --> 00:08:24.110
are going to escalate dramatically. And the other
00:08:24.110 --> 00:08:28.509
fun thing is that they warn about basically the
00:08:28.509 --> 00:08:31.790
company being an existential risk because it
00:08:31.790 --> 00:08:35.370
might accidentally develop super intelligence.
00:08:36.029 --> 00:08:40.019
Accidentally. So we didn't mean it to. Is this
00:08:40.019 --> 00:08:41.919
like, you know, sorry, we've accidentally gone
00:08:41.919 --> 00:08:43.940
on holiday or something. Sorry, we've accidentally
00:08:43.940 --> 00:08:49.549
created super intelligence. I mean, I can't I
00:08:49.549 --> 00:08:53.279
can think of there. They're definitely I mean,
00:08:53.299 --> 00:08:55.700
I would not have bought Facebook stock when it
00:08:55.700 --> 00:08:58.960
when it went public. And I would certainly never
00:08:58.960 --> 00:09:02.340
buy anthropic stock. I mean, I think now might
00:09:02.340 --> 00:09:05.440
be a bad time to do that. I think. Oh, yeah.
00:09:05.779 --> 00:09:09.070
Apart from anything else. If there is, as there
00:09:09.070 --> 00:09:12.389
was with SpaceX, a huge first day bump, don't
00:09:12.389 --> 00:09:14.610
worry about it. It always seems to come back
00:09:14.610 --> 00:09:17.149
down to the IPO price within a day or two, you
00:09:17.149 --> 00:09:19.610
know, so don't be in a hurry. Oh, yeah. The early
00:09:19.610 --> 00:09:22.769
stuff is irrelevant unless you're in the business
00:09:22.769 --> 00:09:25.389
to get that bump, which you have to be a professional
00:09:25.389 --> 00:09:29.529
trader to get the timing right. Yeah. Or if you're
00:09:29.529 --> 00:09:32.210
one of the early investors and you're cashing
00:09:32.210 --> 00:09:36.899
out. Yeah. Yeah. Yes. Completely different. So
00:09:36.899 --> 00:09:40.919
I'll segue on to, although not as neatly as Wendy's
00:09:40.919 --> 00:09:44.980
segue, on to the, well, still, it's called an
00:09:44.980 --> 00:09:47.659
anthropic. So I think I probably mentioned in
00:09:47.659 --> 00:09:53.759
our last edition about games being stolen by
00:09:53.759 --> 00:09:56.779
people recompiling it, decompiling it, mainly
00:09:56.779 --> 00:09:58.759
in China, and then putting on a completely different
00:09:58.759 --> 00:10:02.700
sales platform. So initially, most people in
00:10:02.700 --> 00:10:06.379
the UK and America use Steam for PC stuff. Not,
00:10:06.460 --> 00:10:07.779
you know, obviously the others have their own
00:10:07.779 --> 00:10:11.899
platforms. And yeah, so even, so somebody puts
00:10:11.899 --> 00:10:14.360
a demo out on Steam, which means they haven't
00:10:14.360 --> 00:10:17.240
released the game yet. And then it gets recompiled
00:10:17.240 --> 00:10:19.279
and decompiled and placed on a Chinese site beyond
00:10:19.279 --> 00:10:22.440
our law. And, you know, so for free. So they
00:10:22.440 --> 00:10:24.879
lose the Chinese market, which seems terrible.
00:10:25.720 --> 00:10:30.320
Plus also that can happen even if a demo is actually...
00:10:30.500 --> 00:10:33.759
incomplete in the first case people put an entire
00:10:33.759 --> 00:10:37.679
game but put a flag in the game to stop it carry
00:10:37.679 --> 00:10:39.840
on before the first part you know just for a
00:10:39.840 --> 00:10:43.159
demo ah yes they just throw the flag away so
00:10:43.159 --> 00:10:45.220
i mean that's a little bit dark but they've also
00:10:45.220 --> 00:10:48.100
noticed that people putting correctly a shortened
00:10:48.100 --> 00:10:50.639
game where there's no further content other than
00:10:50.639 --> 00:10:54.039
demo then the ai can extrapolate if it's a simple
00:10:54.039 --> 00:10:55.820
enough game which i mean and the same thing happens
00:10:55.820 --> 00:10:58.860
they lose their game I think it's pretty obvious.
00:10:59.059 --> 00:11:02.480
Anthropic have officially worked this out and
00:11:02.480 --> 00:11:05.779
put out in the sort of, you know, the root instructions
00:11:05.779 --> 00:11:10.820
not to do that. Unfortunately, they've also found
00:11:10.820 --> 00:11:13.220
that there's other things it won't decompile.
00:11:13.259 --> 00:11:16.240
Like, for instance, that somebody tried to do
00:11:16.240 --> 00:11:19.279
a medical, they were trying to do a medical modification
00:11:19.279 --> 00:11:22.159
of something. They decompiled it and Claude said,
00:11:22.299 --> 00:11:25.419
no, can't do that. And that wasn't the intention.
00:11:26.360 --> 00:11:29.340
And Amadeo apologised and said, sorry, we don't
00:11:29.340 --> 00:11:33.580
know why that happened. Also, quite obviously,
00:11:33.779 --> 00:11:37.659
this isn't just Anthropic. There are other models
00:11:37.659 --> 00:11:41.139
that are getting capable, not as strong on the
00:11:41.139 -->
00:11:43.960
coding side, but are going to be quite capable
00:11:43.960 --> 00:11:46.200
of doing this. So that Anthropic don't do it
00:11:46.200 --> 00:11:48.600
is irrelevant unless the other ones stop. I mean,
00:11:48.620 --> 00:11:51.340
specifically Gemini is now powerful enough. And
00:11:51.340 --> 00:11:54.480
actually Grock X might be getting... to the stage
00:11:54.480 --> 00:11:56.139
where it could possibly do that. And, of course,
00:11:56.200 --> 00:11:57.820
that's Elon Musk, though. He doesn't care what
00:11:57.820 --> 00:12:01.580
happens. Oh, life. Yeah. Unfortunately, we may
00:12:01.580 --> 00:12:03.500
not have got very far, but there we go. That's
00:12:03.500 --> 00:12:09.750
that news. I mean, yeah, this is really interesting
00:12:09.750 --> 00:12:11.990
to me. Obviously, the story that I have around
00:12:11.990 --> 00:12:17.470
AI coming for all the R and then coming for Adobe,
00:12:17.750 --> 00:12:23.679
that kind of recompiling... kind of stuff like
00:12:23.679 --> 00:12:26.279
that it's very interesting i find interesting
00:12:26.279 --> 00:12:30.000
because i mean the the latter one about the moat
00:12:30.000 --> 00:12:35.039
is dead the idea of proprietary software um but
00:12:35.039 --> 00:12:37.480
also i do find it interesting i do find it interesting
00:12:37.480 --> 00:12:40.159
with the game one because you put a little demo
00:12:40.159 --> 00:12:44.799
and then you know that kind of the ai dreaming
00:12:44.799 --> 00:12:47.399
up or whatever we want to call extrapolating
00:12:47.399 --> 00:12:49.360
i suppose extrapolating is probably the better
00:12:49.360 --> 00:12:52.639
word rather than uh um you know kind of like
00:12:53.029 --> 00:12:54.909
I find it interesting because you can then have
00:12:54.909 --> 00:12:59.340
these kind of branches. Yeah, what branch are
00:12:59.340 --> 00:13:02.029
you using? Are you using that branch? Do you
00:13:02.029 --> 00:13:05.769
think that this will lead to companies being
00:13:05.769 --> 00:13:08.570
put back under pressure to develop new features?
00:13:08.610 --> 00:13:12.029
Because you can't decompile a new feature. You
00:13:12.029 --> 00:13:19.740
have to actually create one. Ooh. Yeah and no.
00:13:19.860 --> 00:13:23.120
This is the nature of the feature. You can quite
00:13:23.120 --> 00:13:26.580
easily... add a new feature by ai as long as
00:13:26.580 --> 00:13:29.220
the feature is of a standard type that it could
00:13:29.220 --> 00:13:32.399
work out obviously if it's a feature that's literally
00:13:32.399 --> 00:13:36.279
new naturally you can't so the answer that would
00:13:36.279 --> 00:13:38.419
be yeah that would be that would work well to
00:13:38.419 --> 00:13:40.080
stop i mean that would because that would actually
00:13:40.080 --> 00:13:43.899
be a good outcome of this yes right now one of
00:13:43.899 --> 00:13:45.759
the things about emote is it means you don't
00:13:45.759 --> 00:13:47.480
actually have to innovate at
all you can just
00:13:47.480 --> 00:13:50.100
sit there and collect your rent yeah yeah precisely
00:13:50.100 --> 00:13:53.299
yeah so um well you've just said it i mean this
00:13:53.299 --> 00:13:55.600
is that this is the reason why a lot of people
00:13:55.600 --> 00:13:58.179
are keeping quiet because a lot of the moat breaking
00:13:58.179 --> 00:14:00.340
is not a disadvantage to the customer at all.
00:14:00.440 --> 00:14:02.799
Well, that's kind of the other thing about it
00:14:02.799 --> 00:14:05.419
is, I mean, I think the sort of techies have
00:14:05.419 --> 00:14:08.740
always kind of been gleeful when they could break
00:14:08.740 --> 00:14:11.600
something that, you know, make something free
00:14:11.600 --> 00:14:14.120
that previously cost a lot of money. But I think
00:14:14.120 --> 00:14:16.100
particularly at the moment when everybody is
00:14:16.100 --> 00:14:18.279
so angry with the tech companies, there's a particular
00:14:18.279 --> 00:14:20.960
satisfaction in it. And particularly with a company
00:14:20.960 --> 00:14:24.669
like Adobe, which is so widely disliked. yeah
00:14:24.669 --> 00:14:27.750
that's also a worry for me when it's sort of
00:14:27.750 --> 00:14:30.990
emotionally satisfying you kind of let it go
00:14:30.990 --> 00:14:33.470
but then it might not always be that side so
00:14:33.470 --> 00:14:37.110
no you know so yeah it's some like yeah favored
00:14:37.110 --> 00:14:39.250
piece of software that suddenly got cracked by
00:14:39.250 --> 00:14:42.190
some north koreans using ai that they didn't
00:14:42.190 --> 00:14:44.850
understand then the sympathy will reduce a bit
00:14:44.850 --> 00:14:46.730
so we have to be level -headed and go yeah i
00:14:46.730 --> 00:14:48.370
don't care what happens to baby but there are
00:14:48.370 --> 00:14:51.129
other there might be other problems you have
00:14:51.129 --> 00:14:54.740
to you have to you have to You have to make good
00:14:54.740 --> 00:14:59.519
law that sticks no matter what the use case is.
00:14:59.539 --> 00:15:01.840
And I find the likelihood of that happening not
00:15:01.840 --> 00:15:07.940
very high. No. Well, just in that last story
00:15:07.940 --> 00:15:13.090
about the moat, the idea of all proprietary software
00:15:13.090 --> 00:15:19.789
going away, I mean, we started touching on it,
00:15:19.850 --> 00:15:24.100
but what's your thoughts on that? um david as
00:15:24.100 --> 00:15:27.830
a game a game developer i mean like that's pretty
00:15:27.830 --> 00:15:31.960
you know what's your thoughts sorry not being
00:15:31.960 --> 00:15:36.340
on you no no i mean it is um i think so what
00:15:36.340 --> 00:15:39.519
is what is considered for me that complex games
00:15:39.519 --> 00:15:41.720
either are made up of lots of different bits
00:15:41.720 --> 00:15:44.889
can't really be copied but the simpler things
00:15:44.889 --> 00:15:47.210
which obviously a lot of people like the simple
00:15:47.210 --> 00:15:52.090
games uh are a lot easier to copy like for instance
00:15:52.090 --> 00:15:54.789
the famous the very popular game bellatro which
00:15:54.789 --> 00:15:57.909
is a very complex poker thing although it looks
00:15:57.909 --> 00:16:00.809
pretty simple in terms of components but it's
00:16:00.809 --> 00:16:02.850
quite it's you know once you play it this is
00:16:02.850 --> 00:16:05.049
very complex i ain't gonna be copied very easily
00:16:05.049 --> 00:16:08.450
but loves all simpler word games or or shooters
00:16:08.450 --> 00:16:11.029
that just look like every other one first person
00:16:11.029 --> 00:16:13.149
shooter of course they'll be copied but there's
00:16:13.149 --> 00:16:17.730
a similar problem um if your stuff is um too
00:16:17.730 --> 00:16:20.480
derivative by definition that's going to be picked
00:16:20.480 --> 00:16:22.580
off by AI, which works on stuff that already
00:16:22.580 --> 00:16:25.820
exists. So again, if your stuff is original in
00:16:25.820 --> 00:16:30.519
a certain way, it's safer. Unfortunately, the
00:16:30.519 --> 00:16:34.299
industry has sort of gone really toxic with any
00:16:34.299 --> 00:16:38.080
form of AI, even AI to help the developers when
00:16:38.080 --> 00:16:39.519
they're single developers. That doesn't make
00:16:39.519 --> 00:16:41.759
any sense. So they try to protect themselves
00:16:41.759 --> 00:16:43.980
by going over the top and saying, oh, you touched
00:16:43.980 --> 00:16:47.240
AI, go away, you are unclean, which is obviously
00:16:47.240 --> 00:16:51.429
daft. There's a bit of self -preservation with
00:16:51.429 --> 00:16:53.470
respect to artwork. Obviously, nobody wants the
00:16:53.470 --> 00:16:56.730
artwork just ripped off randomly. So the overall,
00:16:56.909 --> 00:16:59.509
but your real question is, will the games industry
00:16:59.509 --> 00:17:01.490
go away if no one can make a game or you could
00:17:01.490 --> 00:17:04.289
make a game just by, you know, writing a line?
00:17:04.789 --> 00:17:07.650
Yeah, in theory, but I think it's just the nature
00:17:07.650 --> 00:17:10.349
of the games. Also, you have to think about live
00:17:10.349 --> 00:17:12.730
service. That's not a game. That's a continuous
00:17:12.730 --> 00:17:15.390
service. Yeah, that's true. This is actually
00:17:15.390 --> 00:17:18.460
the real reason. Well, a lot of big companies
00:17:18.460 --> 00:17:20.359
thought, oh, my God, we better go into live service
00:17:20.359 --> 00:17:23.059
because that can't be ripped off so easily. But
00:17:23.059 --> 00:17:25.900
then discovered that live service is bloody hard
00:17:25.900 --> 00:17:29.180
and they can't do it. So they go, oh, what I'll
00:17:29.180 --> 00:17:31.359
do is a live service game based on some other
00:17:31.359 --> 00:17:33.160
thing. And people said, no, I've already got
00:17:33.160 --> 00:17:35.339
my live service game because it's just like a
00:17:35.339 --> 00:17:38.039
social connection. You're not going to get, you
00:17:38.039 --> 00:17:39.359
know, whatever you think of Facebook, you're
00:17:39.359 --> 00:17:40.940
not going to join a Facebook copy because all
00:17:40.940 --> 00:17:42.579
your friends are on Facebook. We know the network
00:17:42.579 --> 00:17:45.099
effect. Same thing with live service games. Your
00:17:45.099 --> 00:17:48.170
friends are all on one. you know, people have
00:17:48.170 --> 00:17:50.150
done it, but you have to do quite, you have to
00:17:50.150 --> 00:17:53.109
do something real, really good to rip people
00:17:53.109 --> 00:17:55.769
off. So, um, no, not that it's all ripped off,
00:17:55.809 --> 00:17:58.569
tear people away. Sorry, I should say, go play
00:17:58.569 --> 00:18:01.869
something else, but yes. So live service is the
00:18:01.869 --> 00:18:05.089
other way possibly to get around this, but I
00:18:05.089 --> 00:18:07.450
think we'll have to see what happens. The only
00:18:07.450 --> 00:18:10.069
thing we do know is the smaller games, the simpler
00:18:10.069 --> 00:18:11.910
things, the mobile stuff. Yeah, that's going
00:18:11.910 --> 00:18:13.910
to get ripped off. It already happened. AI slot.
00:18:14.049 --> 00:18:16.509
So, you know, recently there was a steam event
00:18:16.509 --> 00:18:20.029
called, a next fest. I just, in summary, there
00:18:20.029 --> 00:18:22.369
was a lot of AI slot there. This is game that
00:18:22.369 --> 00:18:24.769
somebody has just said, Hey, make a game based
00:18:24.769 --> 00:18:26.890
on this happening there. You know, Mario doing
00:18:26.890 --> 00:18:29.250
this, you know, and some of it is obviously totally
00:18:29.250 --> 00:18:33.049
illegal. Yeah. But there's a lot of Nintendo
00:18:33.049 --> 00:18:36.990
would love that. Oh yes. Nintendo have this habit
00:18:36.990 --> 00:18:40.049
of saying nothing and then suing you completely.
00:18:40.470 --> 00:18:43.450
Like they prepare their case. They say nothing,
00:18:43.490 --> 00:18:45.769
prepare their case. And then they just literally.
00:18:46.559 --> 00:18:48.900
The building collapses on you. Yeah, come down
00:18:48.900 --> 00:18:52.680
with the bricks, yeah. Anyway, yeah, so there
00:18:52.680 --> 00:18:55.960
we are with that. The one other thing I was going
00:18:55.960 --> 00:18:58.119
to add to this is I do find it interesting, so
00:18:58.119 --> 00:19:03.440
we were just discussing about Discord and the
00:19:03.440 --> 00:19:11.529
ability to potentially use to decompile Discord,
00:19:12.589 --> 00:19:14.980
so you're still using the networks. you know
00:19:14.980 --> 00:19:18.720
as it is not touching that but your client because
00:19:18.880 --> 00:19:20.859
i asked that was a story i was going to add about
00:19:20.859 --> 00:19:25.849
people using doing my quote marks ai to block
00:19:25.849 --> 00:19:29.609
adverts and stuff like that um which on tvs and
00:19:29.609 --> 00:19:31.849
stuff like that um and i was also thinking about
00:19:31.849 --> 00:19:34.089
the case which we were just talking about where
00:19:34.089 --> 00:19:36.390
discord does some annoying things now and then
00:19:36.390 --> 00:19:40.269
you know and to be able to decompile it just
00:19:40.269 --> 00:19:45.170
enough to just remove some of that Of the craft
00:19:45.170 --> 00:19:47.470
that they think they're making money from. Yeah,
00:19:47.529 --> 00:19:49.970
they would be quite annoyed. They would be annoyed.
00:19:50.390 --> 00:19:53.869
But how would they know? You could probably hide
00:19:53.869 --> 00:19:56.690
that from them, right? No, I mean, they would
00:19:56.690 --> 00:20:02.430
probably start sending more signals to do with
00:20:02.430 --> 00:20:04.569
the size of the file and other things to check,
00:20:04.650 --> 00:20:07.990
is the thing on this network actually our thing?
00:20:08.609 --> 00:20:10.809
They would make more effort. They probably don't
00:20:10.809 --> 00:20:12.210
do that now, but they would make more effort
00:20:12.210 --> 00:20:14.160
if they thought, the only thing that should be
00:20:14.160 --> 00:20:16.059
on this channel is the actual discord client
00:20:16.059 --> 00:20:19.759
uh again i mean that by the way that happens
00:20:19.759 --> 00:20:21.940
with games as well because if you're playing
00:20:21.940 --> 00:20:23.660
game you don't want people to cheat by having
00:20:23.660 --> 00:20:27.119
you know aim bots and the way you do it is well
00:20:27.119 --> 00:20:30.299
they do you do deeper level things to to check
00:20:30.299 --> 00:20:32.700
on clients within the server but you can also
00:20:32.700 --> 00:20:35.500
ask it to respond now of course you can break
00:20:35.500 --> 00:20:39.880
that but that that starts a whole warfare of
00:20:39.880 --> 00:20:42.950
you know what what is on the network Is this
00:20:42.950 --> 00:20:51.470
one of ours or not? And after the AI roundup
00:20:51.470 --> 00:20:55.089
for this month, we have some news stories, some
00:20:55.089 --> 00:20:58.369
regular news stories, starting with David. Tell
00:20:58.369 --> 00:21:03.690
us about, you just tell us. Yeah, so we have
00:21:03.690 --> 00:21:07.069
this definition of AI roundup and the rest, even
00:21:07.069 --> 00:21:09.569
though the rest does sometimes have AI in it.
00:21:09.650 --> 00:21:13.460
But this one, not so much. So the title from
00:21:13.460 --> 00:21:15.880
The Observer today is literally Fight to Block
00:21:15.880 --> 00:21:19.579
Child Sexual Abuse Images Unites Peers and MPs
00:21:19.579 --> 00:21:21.359
in Cross -Party Campaign, which sounds great.
00:21:21.900 --> 00:21:24.400
And then you read it and it's pretty obvious
00:21:24.400 --> 00:21:29.869
that they believe that they can install something
00:21:29.869 --> 00:21:33.930
on every phone that will spy on what you watch.
00:21:34.190 --> 00:21:36.990
And if it's sexual abuse, it will block it. Let
00:21:36.990 --> 00:21:39.730
me just read one paragraph out. Technology companies
00:21:39.730 --> 00:21:42.779
will be
required under the plans. to install
00:21:42.779 --> 00:21:44.900
software on all devices sold in Britain to prevent
00:21:44.900 --> 00:21:47.980
people being able to see explicit images of children.
00:21:48.519 --> 00:21:51.380
Images and videos of sexual abuse, including
00:21:51.380 --> 00:21:54.380
AI generated material, would be instantly blurred
00:21:54.380 --> 00:21:57.019
and live streaming of abuse would become impossible.
00:21:58.000 --> 00:22:01.740
So this is almost like Black Mirror. I'm not
00:22:01.740 --> 00:22:02.980
going to go over that because if you've seen
00:22:02.980 --> 00:22:05.299
it, you've seen it, you know what I mean. But
00:22:05.299 --> 00:22:08.599
also what is more worrying, this is yet more
00:22:08.599 --> 00:22:14.329
attempts to... use ai as witchcraft because the
00:22:14.329 --> 00:22:17.109
technology just just doesn't exist to simply
00:22:17.109 --> 00:22:21.150
go there is a company called safe to net that
00:22:21.150 --> 00:22:23.289
has been claiming for ages that it does exist
00:22:23.289 --> 00:22:26.869
that's that's no question that you could have
00:22:26.869 --> 00:22:30.069
a punt at that um but if you roll it out across
00:22:30.069 --> 00:22:33.210
the whole of the uk and say this this magic eye
00:22:33.210 --> 00:22:35.589
will know that the thing you were looking at
00:22:35.589 --> 00:22:40.049
is is is of a child sexual abuse nature is obviously
00:22:41.299 --> 00:22:43.720
Stupid. For instance, I mean, I was just looking
00:22:43.720 --> 00:22:45.680
at, you know, when we play Ultimate Frisbee,
00:22:45.799 --> 00:22:47.500
one of the fathers on the sideline was changing
00:22:47.500 --> 00:22:49.700
the nappies of his kid. You can imagine if someone
00:22:49.700 --> 00:22:52.119
took a picture of that, a father changing the
00:22:52.119 --> 00:22:55.880
nappy, that could be sexual abuse. That could
00:22:55.880 --> 00:22:59.019
be two things. It could be either. How could
00:22:59.019 --> 00:23:01.779
an AI go, well, it could be that. Why don't we
00:23:01.779 --> 00:23:04.099
just block it? Oh, yes. And let's send a signal
00:23:04.099 --> 00:23:06.740
to the police while we're at it. I'm not saying
00:23:06.740 --> 00:23:09.380
it's possible. But I'm saying that there is this
00:23:09.380 --> 00:23:11.740
company that claims to be able to do it. They
00:23:11.740 --> 00:23:15.480
have persuaded MPs and peers and people that
00:23:15.480 --> 00:23:18.500
this is a real thing. Oh, no, no. I mean, this
00:23:18.500 --> 00:23:21.339
is why this exists. The reason that we're talking
00:23:21.339 --> 00:23:23.519
about this is because the company claims it can
00:23:23.519 --> 00:23:26.059
be done. And there's no question that you can
00:23:26.059 --> 00:23:29.220
train. Again, this is another thing where. people
00:23:29.220 --> 00:23:31.200
believe it just appears whereas there's some
00:23:31.200 --> 00:23:34.259
poor bastard in in india who's having to face
00:23:34.259 --> 00:23:37.079
a lot of training videos of is this a sexual
00:23:37.079 --> 00:23:39.960
abuse did claude get it right you know and then
00:23:39.960 --> 00:23:42.119
continue to say yeah that is no that isn't so
00:23:42.119 --> 00:23:44.779
god knows uh somebody's had a great time it's
00:23:44.779 --> 00:23:46.779
probably a woman having a great time doing that
00:23:46.779 --> 00:23:50.259
but yeah so you could have a pump and you'd be
00:23:50.259 --> 00:23:53.279
wrong probably five percent of the time which
00:23:53.279 --> 00:23:56.420
isn't too bad uh but well it is when it's billions
00:23:57.220 --> 00:24:01.559
Yeah, if it's everybody's phone, I mean, the
00:24:01.559 --> 00:24:04.960
idea is so incredibly stupid. And they go, yeah,
00:24:05.019 --> 00:24:08.480
that's a great idea. People obviously want to
00:24:08.480 --> 00:24:11.480
tell voters that they can stop what is something
00:24:11.480 --> 00:24:14.720
that has nothing to do with the phone. I mean,
00:24:14.779 --> 00:24:17.500
the thing we're talking about isn't anything
00:24:17.500 --> 00:24:20.200
to do with phones. People have been a bit weird
00:24:20.200 --> 00:24:24.380
sexually or just wired up wrongly, badly, or
00:24:24.380 --> 00:24:26.720
however you want to put it. Nothing to do with
00:24:26.720 --> 00:24:29.579
phones. And more to the point, stopping you seeing
00:24:29.579 --> 00:24:32.339
something has nothing to do with producing it.
00:24:32.420 --> 00:24:35.160
The whole thing is completely ludicrous. But
00:24:35.160 --> 00:24:38.400
it's been handled like it's a perfectly reasonable
00:24:38.400 --> 00:24:42.259
thing to do. I have a feeling that the reason
00:24:42.259 --> 00:24:46.440
it's been sort of pushed in now is because it's
00:24:46.440 --> 00:24:49.799
following the Russell Brand case. And everyone's
00:24:49.799 --> 00:24:51.259
feeling, oh, yeah, we've got to do something
00:24:51.259 --> 00:24:54.990
about this kind of stuff. Right. on the head
00:24:54.990 --> 00:24:58.029
it's on the sort of floating behind that being
00:24:58.029 --> 00:25:00.009
pulled in and everyone's going oh yeah yeah this
00:25:00.009 --> 00:25:03.309
will do but i mean again it's witchcraft well
00:25:03.309 --> 00:25:05.369
the other the other part of this in terms of
00:25:05.369 --> 00:25:10.569
timing is that the is offcom just published its
00:25:10.569 --> 00:25:12.569
first report it's required to produce reports
00:25:12.569 --> 00:25:14.950
of how the online safety act is working and its
00:25:14.950 --> 00:25:17.089
first report says that it's making no difference
00:25:17.970 --> 00:25:21.809
that minors are encountering harmful material
00:25:21.809 --> 00:25:24.529
just as often as they were. The one difference
00:25:24.529 --> 00:25:28.190
they seem to have spotted is that children are
00:25:28.190 --> 00:25:31.369
accessing commercial pornography sites less because
00:25:31.369 --> 00:25:35.289
of age verification. But where they see pornography
00:25:35.289 --> 00:25:39.230
is actually in their personal feeds, which age
00:25:39.609 --> 00:25:44.630
verification doesn't fix. No. So that might add
00:25:44.630 --> 00:25:46.490
an additional motive. And there's one other thing
00:25:46.490 --> 00:25:48.730
that came out a couple of years ago was pointed
00:25:48.730 --> 00:25:51.250
out to me by Jen Person, who runs Defend Digital
00:25:51.250 --> 00:25:54.130
Me, which is an NGO specializing on children's
00:25:54.130 --> 00:25:59.769
rights. And that's the past governments anyway.
00:25:59.809 --> 00:26:02.150
I don't know about the current one, but the Sunak
00:26:02.150 --> 00:26:04.529
government really thought there was a billion
00:26:04.529 --> 00:26:08.539
dollar industry available to the UK in the form
00:26:08.539 --> 00:26:11.000
of safety tech. And they really thought that
00:26:11.000 --> 00:26:13.000
safety tech would be an industry that the UK
00:26:13.000 --> 00:26:15.619
could export to other countries and profit from.
00:26:16.180 --> 00:26:19.559
And so there's now that too. Yeah. I mean, obviously
00:26:19.559 --> 00:26:22.420
they've done well on the AI safety specifically,
00:26:22.420 --> 00:26:25.319
although I don't, as you suggested, I don't think
00:26:25.319 --> 00:26:27.000
it was quite envisaged it would be successful
00:26:27.000 --> 00:26:29.059
in the way that we're using it. You're right.
00:26:29.140 --> 00:26:32.099
I think they were imagining a business for AI
00:26:32.099 --> 00:26:36.279
safety tech, not testing each model to see how
00:26:36.279 --> 00:26:40.799
incredibly bad it is. And then again, I sort
00:26:40.799 --> 00:26:44.059
of view this a bit like Wimbledon. England's
00:26:44.059 --> 00:26:46.059
a very good place to have Wimbledon because it's
00:26:46.059 --> 00:26:47.799
very unlikely anyone from the UK will win it.
00:26:47.900 --> 00:26:50.460
So we're like a neutral country. And it's the
00:26:50.460 --> 00:26:51.960
same thing. It's like we're good as a neutral
00:26:51.960 --> 00:26:53.880
country. That wasn't true when Wimbledon was
00:26:53.880 --> 00:26:56.420
created, though. When Wimbledon was created,
00:26:56.640 --> 00:27:00.200
Britain had champions. No, I know. I know. But,
00:27:00.319 --> 00:27:03.319
you know, as things have gone on, you know, it's
00:27:03.319 --> 00:27:05.220
a very easy place, I think, because most people
00:27:05.220 --> 00:27:07.680
do think of it now as a neutral place. Well,
00:27:07.680 --> 00:27:09.349
actually, at this point, Actually, at this point,
00:27:09.390 --> 00:27:11.549
it's very unlikely for any country to have a
00:27:11.549 --> 00:27:13.210
champion because there are so many countries
00:27:13.210 --> 00:27:15.589
that have top players. Right. I'm not going to
00:27:15.589 --> 00:27:19.230
allow this to turn into a tennis thing. No U
00:27:19.230 --> 00:27:24.210
.S. male has won a Grand Slam event since 2003
00:27:24.210 --> 00:27:27.190
when Andy Roddick won the U .S. Open.
It is quite
00:27:27.190 --> 00:27:28.630
incredible to think it's that long ago. Anyway,
00:27:28.670 --> 00:27:30.269
I'm not going to allow this to turn into tennis,
00:27:30.329 --> 00:27:33.609
despite being my fault. This is not a Wimbledon
00:27:33.609 --> 00:27:38.900
podcast. Despite the fact that obviously the
00:27:38.900 --> 00:27:41.400
government said that Trump said, you know, you
00:27:41.400 --> 00:27:43.180
can't send things to Britain to check it because
00:27:43.180 --> 00:27:46.940
it's American. Most people say, hey, we do a
00:27:46.940 --> 00:27:49.599
perfectly good job. Somebody has to do it. And
00:27:49.599 --> 00:27:51.460
it needs to be somebody really outside of the
00:27:51.460 --> 00:27:56.019
purview of anthropic Gemini, you know, OpenAI,
00:27:56.200 --> 00:27:58.539
et cetera, and Google. So, I mean, we're fine
00:27:58.539 --> 00:28:00.220
with that. But as it comes back to what your
00:28:00.220 --> 00:28:03.640
point is, yeah, I think there was an actual big
00:28:03.640 --> 00:28:08.119
money business. um of which if it was possible
00:28:08.119 --> 00:28:12.019
there may be but there really isn't um there's
00:28:12.019 --> 00:28:15.099
just lots of people making claims uh and if they
00:28:15.099 --> 00:28:18.700
were doing it on scale and you know on on hyperscale
00:28:18.700 --> 00:28:21.160
the errors would be i mean you could just imagine
00:28:21.160 --> 00:28:24.440
someone going why why is this blurred out and
00:28:24.440 --> 00:28:26.680
why the police coming to my door i mean you can't
00:28:26.680 --> 00:28:29.799
deal with that it's just arbitrary it's going
00:28:29.799 --> 00:28:33.109
to be complete disaster And it won't happen because
00:28:33.109 --> 00:28:37.720
it would obviously be a complete disaster. We
00:28:37.720 --> 00:28:42.640
haven't touched the, just the, I mean, sorry,
00:28:42.819 --> 00:28:46.259
also I'm an advisor for the Open Rights Group
00:28:46.259 --> 00:28:52.630
and... As am I. Yes, so I didn't want to out
00:28:52.630 --> 00:28:57.769
you. But yeah, I just, these topics come up again
00:28:57.769 --> 00:29:00.329
and again. I don't understand how, you know what,
00:29:00.349 --> 00:29:02.569
I always feel like, so there was a long time
00:29:02.569 --> 00:29:04.789
ago, not a long time, a while ago we used to
00:29:04.789 --> 00:29:07.509
do this thing uh was first called the ethical
00:29:07.509 --> 00:29:11.789
dilemma cafe and i feel like some of these ministers
00:29:11.789 --> 00:29:14.190
need to go through it they need to see their
00:29:14.190 --> 00:29:17.329
own data they need to see they need to experience
00:29:17.329 --> 00:29:20.579
what it's like yeah when these things actually
00:29:20.579 --> 00:29:23.460
go wrong you know and they do go wrong regularly
00:29:23.460 --> 00:29:27.910
they're they just they need to feel the pain
00:29:27.910 --> 00:29:31.710
so they can truly understand what they're actually
00:29:31.710 --> 00:29:34.170
going for. Because I feel like a bunch of lobbyists
00:29:34.170 --> 00:29:37.049
go, yeah, this is great. We can do this. Oh,
00:29:37.150 --> 00:29:41.150
we can do it 99 .9 % of the time. In a lab. Yeah.
00:29:41.369 --> 00:29:45.250
And it's like, no, you can't. But they do pretend
00:29:45.250 --> 00:29:47.109
that. Training on the six examples of CSAM that
00:29:47.109 --> 00:29:51.390
we have. Yeah. And it's just like, I know. It's
00:29:51.390 --> 00:29:55.029
like, oh, my God. How could you fall for this?
00:29:55.849 --> 00:29:58.869
Well, the obvious example, irony, is that MPs
00:29:58.869 --> 00:30:00.589
will tell you the same. You know, you need to
00:30:00.589 --> 00:30:03.549
understand the pain of all our, you know, all
00:30:03.549 --> 00:30:06.190
the people that voted for me telling me that
00:30:06.190 --> 00:30:09.069
you have to solve this problem because I was
00:30:09.069 --> 00:30:11.470
just walking along and suddenly my kid has seen
00:30:11.470 --> 00:30:13.690
this stuff and this is revolting. How did it
00:30:13.690 --> 00:30:15.589
happen? And it's your fault. None of which are
00:30:15.589 --> 00:30:19.789
true. But they have to deal with it. And what
00:30:19.789 --> 00:30:23.190
happens afterwards, probably way after they've
00:30:23.190 --> 00:30:25.809
changed departments. they could not care less
00:30:25.809 --> 00:30:28.210
about, to be honest. I don't have any bad feelings
00:30:28.210 --> 00:30:30.230
about the UK, apologies to Jess Phillips and
00:30:30.230 --> 00:30:33.910
so forth. They are, as I said, yeah, they are
00:30:33.910 --> 00:30:35.750
kind of believing in the witchcraft and I kind
00:30:35.750 --> 00:30:37.430
of think they might be, you know, dunking women
00:30:37.430 --> 00:30:39.809
in the ponds if they're given the chance. But
00:30:39.809 --> 00:30:42.309
their basic background, obviously, they are,
00:30:42.509 --> 00:30:46.329
they believe, and they do believe, something
00:30:46.329 --> 00:30:48.670
should happen, something should be done, something
00:30:48.670 --> 00:30:51.109
must be done, you know. And of course, it would
00:30:51.109 --> 00:30:53.210
be good if something could be done, but this
00:30:53.210 --> 00:30:56.269
isn't it. you know by quite a long shot what
00:30:56.400 --> 00:30:59.380
happened to rational thought what about kind
00:30:59.380 --> 00:31:03.279
of like waiting and like actually a strategy
00:31:03.279 --> 00:31:06.240
and actual like well this is a very emotional
00:31:06.240 --> 00:31:09.140
area and obviously yeah that's true i think i
00:31:09.140 --> 00:31:11.200
think it's hard to have rational thought when
00:31:11.200 --> 00:31:13.640
people are really frightened absolutely yeah
00:31:13.640 --> 00:31:16.880
yeah yeah that's true obviously if if you didn't
00:31:16.880 --> 00:31:18.460
have a care in the world then suddenly you see
00:31:18.460 --> 00:31:21.509
your child looking at see Sam, naturally you
00:31:21.509 --> 00:31:23.170
think, oh my
God, what's gone wrong? I didn't
00:31:23.170 --> 00:31:25.009
even know this was possible, which a lot of people
00:31:25.009 --> 00:31:28.710
believe it isn't possible. And that's where the
00:31:28.710 --> 00:31:31.829
witchcraft starts. Can you create a hex that
00:31:31.829 --> 00:31:35.589
will stop this happening? I think it's more the
00:31:35.589 --> 00:31:39.670
stories about kids who have been groomed and
00:31:39.670 --> 00:31:42.410
trafficked or blackmailed or commit suicide.
00:31:42.970 --> 00:31:45.170
Those are the things that really scare parents.
00:31:45.890 --> 00:31:48.170
sure and they and they work their way back and
00:31:48.170 --> 00:31:50.650
then see what they saw and they come to the correct
00:31:50.650 --> 00:31:53.809
conclusion that the stuff that they're seeing
00:31:53.809 --> 00:31:56.210
has influenced their behavior i've no doubt that
00:31:56.210 --> 00:32:00.109
it does you know um and then they again they
00:32:00.109 --> 00:32:02.730
work their way back and go can we do something
00:32:02.730 --> 00:32:06.329
to make it not happen it reminds me of the s
00:32:06.329 --> 00:32:08.549
minister line this is something therefore we
00:32:08.549 --> 00:32:16.210
must do it yeah yeah and for our next story we
00:32:16.210 --> 00:32:21.420
have Well, I'm going to talk about the landline
00:32:21.420 --> 00:32:27.579
switchover. So the idea is that people are not
00:32:27.579 --> 00:32:29.500
using landlines. And to be fair, I don't have
00:32:29.500 --> 00:32:31.920
a landline. A lot of people don't have landlines.
00:32:31.940 --> 00:32:34.740
They have mobile phones. And so the government
00:32:34.740 --> 00:32:37.640
is starting to make the decision that they should
00:32:37.640 --> 00:32:42.259
probably cut landlines. But there are people
00:32:42.259 --> 00:32:46.309
who are worried. And funny enough, A lot of the
00:32:46.309 --> 00:32:48.930
ministers, well, quite a few ministers have raised
00:32:48.930 --> 00:32:52.210
concerns, probably from the people that they
00:32:52.210 --> 00:32:55.430
represent, that actually this is really important.
00:32:56.150 --> 00:32:58.769
And I'm interested in what the rest of the tech
00:32:58.769 --> 00:33:03.049
grumps think, because I do think that there is
00:33:03.049 --> 00:33:06.990
a benefit to a landline. And I wonder, just cutting
00:33:06.990 --> 00:33:10.309
it completely and going IP does raise a bunch
00:33:10.309 --> 00:33:13.349
of problems, especially when you're talking about...
00:33:13.809 --> 00:33:16.730
you know currently the uk is talking about being
00:33:16.730 --> 00:33:21.210
prepared for for all types of things um so i
00:33:21.210 --> 00:33:22.609
was just wondering what the rest of the tech
00:33:22.609 --> 00:33:28.109
grumps thought about this uh it's so it's interesting
00:33:28.109 --> 00:33:31.710
because there's two things there one is using
00:33:31.710 --> 00:33:34.029
a mobile instead of using a landline which most
00:33:34.029
--> 00:33:36.690
people do of course there's also using an ip
00:33:36.690 --> 00:33:40.829
phone which you can have like a phone that is
00:33:40.829 --> 00:33:45.410
actually you know using using IPA, not using
00:33:45.410 --> 00:33:48.970
the photo. I mean, a few years ago, that was
00:33:48.970 --> 00:33:51.690
the plan. Despite the fact that I used to work
00:33:51.690 --> 00:33:53.849
in BT a long time ago, I don't actually know
00:33:53.849 --> 00:33:56.809
what the plan is. Is it just to get rid of landlines,
00:33:56.849 --> 00:33:59.789
go goodbye, call them all out, except for the
00:33:59.789 --> 00:34:02.670
internet stuff? I don't know what it is. I mean,
00:34:02.670 --> 00:34:04.829
it sounds to me as if they're saying they're
00:34:04.829 --> 00:34:08.420
not going to have the phone network. When I wrote
00:34:08.420 --> 00:34:11.480
about this for Scientific American in something
00:34:11.480 --> 00:34:14.659
like the beginning of the century, like 2000,
00:34:14.960 --> 00:34:18.199
around then, they were talking about replacing
00:34:18.199 --> 00:34:21.739
the existing plain old telephone network system
00:34:21.739 --> 00:34:28.929
with basically an IP network. Yeah. And I assume
00:34:28.929 --> 00:34:31.090
that this is sort of the final stage of that
00:34:31.090 --> 00:34:35.960
process. Yeah. My landline was converted. was
00:34:35.960 --> 00:34:39.900
converted two, three years ago. Yeah, so I assume
00:34:39.900 --> 00:34:42.840
that will be rolled out. Say you have a landline,
00:34:42.900 --> 00:34:48.159
in quotes, that's just an IP network. What I
00:34:48.159 --> 00:34:51.039
have is a box connected to my broadband router,
00:34:51.280 --> 00:34:54.539
which allows me to plug in my ordinary, previously
00:34:54.539 --> 00:34:57.699
landline phones. But you don't have to do that.
00:34:57.739 --> 00:35:00.849
You could use a software phone. Yes. The thing
00:35:00.849 --> 00:35:06.949
is, here's the interesting story. Last weekend,
00:35:07.909 --> 00:35:12.429
rather Monday, I got a message from my ISP, which
00:35:12.429 --> 00:35:14.670
is a small business ISP, so I'm talking to real
00:35:14.670 --> 00:35:17.869
humans rather than like a giant help desk. Weird.
00:35:17.889 --> 00:35:20.849
I got an email from them saying, could I go through
00:35:20.849 --> 00:35:24.150
and scan all my devices for malware and change
00:35:24.150 --> 00:35:28.510
all the passwords because somebody had hacked
00:35:28.510 --> 00:35:31.730
into my phone box and made 322 pounds worth of
00:35:31.730 --> 00:35:36.510
calls on Sunday morning. Oh. As you may understand,
00:35:36.730 --> 00:35:40.650
I was a bit unhappy about this. So I sort of
00:35:40.650 --> 00:35:44.940
pushed back and I said, well, can you compare
00:35:44.940 --> 00:35:50.570
the call logs with the log of when my broadband
00:35:50.570 --> 00:35:53.250
was out? Because it was down from Saturday night
00:35:53.250 --> 00:35:58.289
at quarter to 11 until Sunday
night. And, you
00:35:58.289 --> 00:36:00.309
know, maybe it was up again for a short time
00:36:00.309 --> 00:36:02.920
and I didn't know, but could you? And so they
00:36:02.920 --> 00:36:05.739
looked and my broadband was down at the time.
00:36:06.820 --> 00:36:10.880
So it's clearly somebody may have hacked the
00:36:10.880 --> 00:36:12.719
phone box, but that's not going to be the source
00:36:12.719 --> 00:36:14.139
of those calls. The other thing they noticed
00:36:14.139 --> 00:36:16.820
was that I am only allowed one channel for phone
00:36:16.820 --> 00:36:19.780
calls. And there were eight calls that took place
00:36:19.780 --> 00:36:21.960
at the same time, which means eight channels.
00:36:23.829 --> 00:36:26.289
I asked a security expert and he said the most
00:36:26.289 --> 00:36:28.789
likely is that somebody is hacked into an operational
00:36:28.789 --> 00:36:32.130
support system and redirect, reconnect, connected
00:36:32.130 --> 00:36:36.730
some number. Now, this is a level of threat.
00:36:36.750 --> 00:36:39.010
This is a kind of threat that landlines have
00:36:39.010 --> 00:36:42.710
never had historically. If it had not been for
00:36:42.710 --> 00:36:46.019
the fact. The accidental coincidence that the
00:36:46.019 --> 00:36:49.559
day they chose to use that number, my broadband
00:36:49.559 --> 00:36:52.119
was down. I would have absolutely no chance of
00:36:52.119 --> 00:36:55.809
convincing the ISP it wasn't me. Yeah, I mean.
00:36:56.849 --> 00:37:00.090
I mean, the eight channels would have raised
00:37:00.090 --> 00:37:05.050
some eyebrows. But, you know, I still don't know.
00:37:05.840 --> 00:37:08.320
They have an upstream provider, and the upstream
00:37:08.320 --> 00:37:11.059
provider, I don't know how seriously they're
00:37:11.059 --> 00:37:13.119
taking investigating it. I don't want to pay
00:37:13.119 --> 00:37:15.900
the money. I think it's clear it's not my fault.
00:37:16.179 --> 00:37:20.300
I don't know how this story is going to come
00:37:20.300 --> 00:37:21.860
out, which is why I haven't written about it
00:37:21.860 --> 00:37:25.849
yet. But it's a whole level of threat we've never
00:37:25.849 --> 00:37:28.949
seen before, and it could affect tons of people
00:37:28.949 --> 00:37:33.530
if somebody upstream has... a poorly secured
00:37:33.530 --> 00:37:36.429
operational support system, then, you know...
00:37:36.429 --> 00:37:38.789
I mean, obviously, I have to say this, because
00:37:38.789 --> 00:37:41.809
the timing, given that we've been told that,
00:37:41.849 --> 00:37:44.130
you know, Russia and North Korea are, of course,
00:37:44.130 --> 00:37:46.869
using AI to affect us at the moment, and we've
00:37:46.869 --> 00:37:50.369
seen not just freezers and, oddly enough, alerts
00:37:50.369 --> 00:37:54.070
for your phone. So this is unfortunate, a very
00:37:54.070 --> 00:37:57.809
unfortunate time to suddenly be doing this or
00:37:57.809 --> 00:38:00.750
to be looking at this, because people might go...
00:38:00.960 --> 00:38:03.300
Wait a minute, you know, whereas I suspect that
00:38:03.300 --> 00:38:05.219
other times it would have gone almost unnoticed.
00:38:05.500 --> 00:38:08.739
It could just be unfortunate this is being looked
00:38:08.739 --> 00:38:10.920
at. But sorry, the reason I'm saying that, I
00:38:10.920 --> 00:38:13.420
don't know when the timing is. When was this
00:38:13.420 --> 00:38:15.300
going to go through? I mean, I've heard about
00:38:15.300 --> 00:38:16.820
it without knowing exactly what they're doing.
00:38:16.820 --> 00:38:20.519
It's been delayed at least once already. Yeah,
00:38:20.519 --> 00:38:23.340
I think after that I've stopped, you know, wondering.
00:38:24.000 --> 00:38:26.039
Obviously on the basics, yes, of course, most
00:38:26.039 --> 00:38:28.300
people don't use a landline, but that's a different
00:38:28.300 --> 00:38:31.800
thing. By the way, in case you're wondering how
00:38:31.800 --> 00:38:36.900
a phone bill over VoIP could go to £322, it turns
00:38:36.900 --> 00:38:39.579
out that the pricing... Yeah, it's a really interesting
00:38:39.579 --> 00:38:41.699
question because, you know, the US costs like
00:38:41.699 --> 00:38:44.820
a penny or something. Well, it turns out that
00:38:44.820 --> 00:38:47.719
the calls were to Diego Garcia and that particular
00:38:47.719 --> 00:38:51.119
place, for some reason, cost more than £3 a minute.
00:38:51.119 --> 00:38:53.320
Yeah, the international boundaries are often
00:38:53.320 --> 00:38:59.170
where providers actually make money. I mean,
00:38:59.269 --> 00:39:04.889
I had some stupid thing where I had a call when
00:39:04.889 --> 00:39:07.670
I was on holiday in North Africa and the call
00:39:07.670 --> 00:39:11.349
came to the UK and I said I didn't make that
00:39:11.349 --> 00:39:15.610
call. I just wouldn't. But I discovered that
00:39:15.610 --> 00:39:20.230
actually my Barclays app made the call instead
00:39:20.230 --> 00:39:23.130
of using IP, they just rang directly from my
00:39:23.130 --> 00:39:26.659
phone. and i and i completely forgot that and
00:39:26.659 --> 00:39:29.920
oh god so i didn't make it but the app did um
00:39:29.920 --> 00:39:33.179
so you get yeah i mean sort of so the international
00:39:33.179 --> 00:39:36.800
boundaries is very much a case where you know
00:39:36.800 --> 00:39:40.980
profit is made and in this case lost so it's
00:39:40.980 --> 00:39:43.920
a bit of a a bit of a problem there but yeah
00:39:43.920 --> 00:39:46.820
i mean if if these things all have reduced price
00:39:46.820 --> 00:39:49.300
then again uh and they don't have these boundary
00:39:49.300 --> 00:39:51.400
conditions then of course that will be less of
00:39:51.400 --> 00:39:53.820
a problem but at the moment the international
00:39:53.820 --> 00:39:57.980
cases certainly wants to consider um i said the
00:39:57.980 --> 00:40:02.800
other thing is uh you know everybody has a mobile
00:40:02.800 -->
00:40:07.980
um so then we're going to the mobile coverage
00:40:07.980 --> 00:40:11.940
in rural areas but i believe yeah it's improving
00:40:11.940 --> 00:40:15.179
now because obviously if you're gonna pull stuff
00:40:15.179 --> 00:40:17.519
out you have to prove that the mobile coverage
00:40:17.519 --> 00:40:21.519
is everywhere and i don't mean mobile coverage
00:40:21.929 --> 00:40:25.269
It does depend. I mean, my friend's house across
00:40:25.269 --> 00:40:28.849
the railway tracks, I go into her house and it's
00:40:28.849 --> 00:40:32.369
a black spot for my mobile phone. Yeah. And that's
00:40:32.369 --> 00:40:37.139
in southwest London. Yeah, actually, London is
00:40:37.139 --> 00:40:39.639
not great for quite... I mean, it's quite famously
00:40:39.639 --> 00:40:41.739
not great all the time. As I said, I don't have...
00:40:41.739 --> 00:40:44.179
Yeah, I've noticed that. I don't have fibre here,
00:40:44.239 --> 00:40:46.539
famously. This road has no fibre. What? We've
00:40:46.539 --> 00:40:49.340
all had 5G. Yep, the whole road has no fibre.
00:40:49.500 --> 00:40:52.320
Because this is because I'm on a boundary between
00:40:52.320 --> 00:40:55.699
two boroughs. And those are the last places that
00:40:55.699 --> 00:40:58.679
get connected. Because everyone goes, Hey, it's
00:40:58.679 --> 00:41:00.199
your problem. It's not our problem. It's your
00:41:00.199 --> 00:41:04.590
problem. It's Ealing. It's Nyandlo. So again,
00:41:05.110 --> 00:41:08.489
it's all very well. If it's the case that everybody
00:41:08.489 --> 00:41:11.449
has good coverage, then ripping stuff out is
00:41:11.449 --> 00:41:13.550
fine. Other than that, it'll have to be done
00:41:13.550 --> 00:41:15.389
in phases, which I think is obvious that it would
00:41:15.389 --> 00:41:18.610
be anyway. And yes, strangely enough, London
00:41:18.610 --> 00:41:20.969
is not necessarily as well covered because of
00:41:20.969 --> 00:41:23.769
the bouncing and shading. It's very difficult
00:41:23.769 --> 00:41:27.610
to, it always has been a problem. Yeah. In some
00:41:27.610 --> 00:41:29.929
respects, rural is easier because it's yes or
00:41:29.929 --> 00:41:33.539
no. the cow shed over there isn't going to be
00:41:33.539 --> 00:41:35.179
any different from the barn over here it's either
00:41:35.179 --> 00:41:37.739
you got it or you don't have it and i was thinking
00:41:37.739 --> 00:41:41.150
about this as well because um yeah we could talk
00:41:41.150 --> 00:41:43.630
about all landlines so we're talking about also
00:41:43.630 --> 00:41:46.610
these public telephone boxes which are definitely
00:41:46.610 --> 00:41:50.650
disappearing but i just it just there's there's
00:41:50.650 --> 00:41:52.449
loads of different things like that yeah the
00:41:52.449 --> 00:41:55.210
rural arguments but also i just started thinking
00:41:55.210 --> 00:41:57.909
well yeah it was probably telephone boxes. And
00:41:57.909 --> 00:41:59.690
I know most of them are
just now turned into
00:41:59.690 --> 00:42:01.809
empty shells. So you can actually do a mobile
00:42:01.809 --> 00:42:07.030
phone call, but it feels like in an emergency,
00:42:07.050 --> 00:42:12.929
I know you can dial the number on a phone. It
00:42:12.929 --> 00:42:15.110
wouldn't cost anything. You can do it from the
00:42:15.110 --> 00:42:18.750
lock screen and stuff like that. But that still
00:42:18.750 --> 00:42:23.329
relies on an IP network. Oh yeah. When my broadband
00:42:23.329 --> 00:42:27.920
was down, I had no dial time. Yeah. yeah of course
00:42:27.920 --> 00:42:30.840
this is the concern i think it's like yeah i
00:42:30.840 --> 00:42:33.480
mean i i'm just gonna say i mean there's that
00:42:33.480 --> 00:42:38.119
thing about a house has electricity and gas now
00:42:38.119 --> 00:42:40.059
it doesn't mean i'd be delighted if electricity
00:42:40.059 --> 00:42:42.719
went out but at least i can you know put something
00:42:42.719 --> 00:42:45.360
on a hob uh similarly if a gas goes well that's
00:42:45.360 --> 00:42:47.619
fine i could still use the cooker or whatever
00:42:47.619 --> 00:42:49.699
depending how you set up or you still use a kettle
00:42:49.699 --> 00:42:52.599
at least i think it i don't think because we
00:42:52.599 --> 00:42:54.860
haven't been tested very much recently i mean
00:42:54.860 --> 00:42:58.639
uh i was I remember the three -day week when
00:42:58.639 --> 00:43:01.760
we had no electricity in the 70s and people did
00:43:01.760 --> 00:43:03.800
have to have candles and we were tested then
00:43:03.800 --> 00:43:05.360
and people worked out that, yeah, you know, having
00:43:05.360 --> 00:43:08.019
separate gas, not so bad. So I think because
00:43:08.019 --> 00:43:09.900
we haven't been tested, there's a bit of shrug
00:43:09.900 --> 00:43:11.639
and go, well, I've got a mobile phone, so what's
00:43:11.639 --> 00:43:15.239
the problem? But it would be interesting if,
00:43:15.260 --> 00:43:18.420
yeah, there were suddenly coverage attacks and
00:43:18.420 --> 00:43:21.179
I wonder how people feel. I don't know. I don't
00:43:21.179 --> 00:43:29.960
know. Or a power outage. And that is the end
00:43:29.960 --> 00:43:35.480
of Tech Grumps, except there's always AOBs. And
00:43:35.480 --> 00:43:40.059
the first AOB is from Wendy. Tell us about Cupertino.
00:43:40.960 --> 00:43:43.539
Yeah, I wanted to tell people about Cupertino.
00:43:43.559 --> 00:43:47.400
It's quite good fun. Michelle and Robert King,
00:43:47.619 --> 00:43:50.739
the masterminds behind the series The Good Wife
00:43:50.739 --> 00:43:54.369
and its sequel The Good Fight. and another series
00:43:54.369 --> 00:43:57.690
called Evil, and another series, they've done
00:43:57.690 --> 00:44:00.489
quite a few. The Good Wife was in particular
00:44:00.489 --> 00:44:02.389
really interesting because it was very clever
00:44:02.389 --> 00:44:06.190
about technology and law. And Cupertino is the
00:44:06.190 --> 00:44:09.150
story
of a couple of lawyers who meet on the
00:44:09.150 --> 00:44:11.469
day that they are simultaneously fired from a
00:44:11.469 --> 00:44:13.989
very large technology company, and they decide
00:44:13.989 --> 00:44:16.349
to team up and represent people who've been screwed
00:44:16.349 --> 00:44:19.550
by technology companies. And it seems to be expertly
00:44:19.550 --> 00:44:24.159
timed. and uh it's on cbs the cbs network in
00:44:24.159 --> 00:44:27.099
the states i don't know how soon it will come
00:44:27.099 --> 00:44:29.840
to the uk but the first few episodes have been
00:44:29.840 --> 00:44:33.139
really fun so i'm i have high hopes for it i'm
00:44:33.139 --> 00:44:36.659
sure it'll be on pirate bay yeah i wasn't gonna
00:44:36.659 --> 00:44:43.659
say that i was okay um all right so i've got
00:44:43.659 --> 00:44:46.969
two um the one that I always want to remind people
00:44:46.969 --> 00:44:49.570
of is that there is another podcast that I do
00:44:49.570 --> 00:44:54.250
called Making the Familiar Strange. And me and
00:44:54.250 --> 00:44:57.650
my co -hosts have just recorded or about to record
00:44:57.650 --> 00:45:00.789
some interesting interviews from a workshop that
00:45:00.789 --> 00:45:04.469
we did in the Netherlands. So listen out for
00:45:04.469 --> 00:45:09.750
that. And finally, I kind of, when I was in the
00:45:09.750 --> 00:45:12.510
Netherlands, I spent a bit of time on trains.
00:45:13.289 --> 00:45:16.849
And when I wasn't DJing with my new device, I
00:45:16.849 --> 00:45:20.570
was listening to, what's the name? Cherry Turkle's
00:45:20.570 --> 00:45:24.429
new book, Artificial Intuosy, which is also very
00:45:24.429 --> 00:45:29.030
related to the workshop and the podcast that
00:45:29.030 --> 00:45:32.610
I just mentioned. So I do recommend it because
00:45:32.610 --> 00:45:36.190
it starts off slow and then it really gets up
00:45:36.190 --> 00:45:39.769
to speed. And I'm sure there'll be lots of things
00:45:39.769 --> 00:45:43.920
I'll be quoting from that in future. she she's
00:45:43.920 --> 00:45:46.099
been covering this sort of topic for a long long
00:45:46.099 --> 00:45:48.820
time she's written a number of very well regarded
00:45:48.820 --> 00:45:53.079
books on it yes absolutely she she seems like
00:45:53.079 --> 00:45:55.820
yeah she's just been in this field for so long
00:45:55.820 --> 00:45:58.880
and i there was a few books that she did that
00:45:58.880 --> 00:46:03.150
was a bit like but this one she's she's back
00:46:03.150 --> 00:46:08.710
she's firing and for good reason And that is
00:46:08.710 --> 00:46:12.210
the end of Tech Grumps. So it's just my time
00:46:12.210 --> 00:46:16.309
and our time to say goodbye. And it's bye from
00:46:16.309 --> 00:46:21.750
me. Bye from David. Bye from me. Bye from Wendy.
00:46:22.269 --> 00:46:25.030
Bye from me. If you ever want to be on Tech Grumps,
00:46:25.170 --> 00:46:29.090
then just use the hashtag Tech Grumps and we'll
00:46:29.090 --> 00:46:31.969
find you, depending on what network you're on.
00:46:32.150 --> 00:46:33.829
But if you're sort of fed up with us, I'll find
00:46:33.829 --> 00:46:36.610
you. David will find you. We'll find you in some
00:46:36.610 --> 00:46:40.920
way. I have a very special skill. We will find
00:46:40.920 --> 00:46:45.369
you. Goodbye from all of us. Bye.