Aug 31, 2024 ยท 2024 #29. Read the transcript grouped by speaker, inspect word-level timecodes, and optionally turn subtitles on for direct video playback
Edit labels for this show, save them in this browser, or download a JSON override for the production folder.
Transcript Playback
Irrational Liability
Human Transcript
Timed transcript
Blocks are grouped by speaker for readability. Expand a block to inspect word-level timing.
Speaker 2
Hello, everybody. It is Saturday, August the 31st, 2024, last day, of course, of August. It's been quite a week in terms of technology and politics. Three major stories. Pavel Durov, the CEO of Telegram, was arrested in Paris, created a great deal of controversy. A lot of people very unhappy, others more ambivalent. Then yesterday, Brazil blocked X after Elon Musk ignored court orders. Surprise, surprise. Brazilian judge ordered the suspension of X. And finally, Mark Zuckerberg published a letter on what he calls Facebook censorship by the Biden administration when it came to COVID-19. Lots of material then for Keith's tears. That was the week newsletter. And of course, he leads with the juror of arrest. He asks in his editorial, irrational liability is telegram. a criminal plot for regular viewers of the show and for people who know Keith. You won't be surprised to hear that Keith is not very happy with Durov's arrest. I'm more ambivalent. Keith, what's the problem? Why shouldn't we be celebrating the arrest of Durov?
Well, he isn't a criminal, is the first thing. This is an administrative arrest in pursuit of a political idea. And that political idea is that when the human race uses social media, the owners of the platform should be liable for what the people publish. And that is a political idea that is at odds with the entire history of Western civilization, which is a history of increasingly allowing individual freedom to say...
I don't know, was it Gandhi or Nehru who, when he was asked about Western civilization, his response was, that would be a good idea. So I'm not entirely sure what Western civilization is, but let's not go down that rabbit hole. How... What were these three things, the Musk, the X case in Brazil, the Jurov case, and maybe the Zuckerberg case, connected? It seems to me as if the governments are getting into the business of getting rid of Section 230 in an abstract sense and simply saying, look, we're going to determine internet policy. Is there any truth to that, whether it's Brazil or France or the U.S. ?
Well, there's a lot of truth to that. That's exactly what's happening. The UK, by the way, is leading the pack with the new Labour government imprisoning people involved in the recent riots for statements they made of opinion on social media, often for several years in prison. So this has been a trend for a long time. We've talked about it here on the show. We started to highlight it a few weeks ago, but we've talked about it for a few years.
Yeah, a couple of weeks ago, you asked if Europe was dying, although I'm sure your editorial at that time was about innovation in Europe rather than government policy. You mentioned Western civilization, Keith, which, as I said, is a rather abstract notion. But some people might argue that Western civilization is built on responsible, strong government. And without that, you wouldn't have Western civilization.
You remember, the government is of the people, for the people, by the people. So the concept of strong government is usually against the people. And I don't really see any threat of democracy here. What I see is a threat of fear. Uh, and it's really driven by something quite new and it's, it's worth pointing this out. There were, there were really no global technologies until the 1990s, uh, before the 90s.
Yeah, but it wasn't, uh, there were, you know, people didn't publish books in every language simultaneously globally. So, uh, it was really national in character, the internet. didn't change that initially. The internet was initially pretty much in English, and so governments that don't have English-speaking populations didn't fear it. But little by little, and really accelerated by the rise of mobile, where you've got two operating systems in every country in the world, either Android or iOS, the same apps everywhere in the world, unless the government blocks them, suddenly you have this global canvas that each individual government is now faced with a new challenge, things being published outside of its control. And that is new. And governments always react badly to losing control. So I think this is essentially a control-driven set of events. And the law is used as a weapon to exercise that control. And in this case, the Telegram founder is a victim of that, but there'll be many more. If this passes muster, why isn't Zuckerberg exactly the same? Why isn't Musk exactly the same? What is the difference? Well,
Musk, yeah, I mean, and as I was saying to you before, you and I, we agree on some things, but on this front, we're completely in disagreement. I would actually... If X shut down tomorrow, I think it would benefit the world. It would probably be the beginning rather than the end of what you call Western civilization. I think it's a corrosive platform that does no good to anyone. But I mean, you know what I think. I know what you think. So I don't necessarily think it's a bad thing for X to be shut down by Brazil or any other government. The Telegram case is a particularly interesting one because, of course, it does raise the issue of accountability. So if you're doing something illegal on Telegram, selling drugs, building nuclear weapons, should the platform owner be liable? Is that the French case against Juro?
Well, post-feudalism, individual freedom became a thing. And democracy, meaning every individual is equal as a citizen and has rights, is a thing. Those are not just... Right,
but I think Nero's point might have been that along with all that, you had Western colonialization and mass murder and all the rest. But let's not get stuck into that, Keith. But let's talk about Telegram and whether or not... So back to privacy then, Andrew, back to privacy.
Yeah. Now, it isn't fully encrypted. Public conversations are not encrypted. But private conversations are. And Telegram supports groups very similar to WhatsApp. Right. And unlike Apple... Telegram doesn't throw away the keys so it does have the keys available to unlock encrypted conversations and Therefore governments have come to it and said we want to know what these private conversations are about Now it turns out that telegram took a slightly different approach than Apple what it did is it distributed the keys and globally into different domiciles, so that no single government has jurisdiction to require it to hand over the keys. In order to hand over the keys, the government would have to get jurisdiction in multiple nations.
When it comes to the French case, are the French authorities, is there a particular case they're pursuing? Is there a broader complaint against Telegram generally?
I think there's a lot of speculation, and my idea there is equally speculative. But the belief is that the French police are pursuing a number of cases. Child pornography appears to be one of them. Surprise, surprise.
Would it be fair to say... No, let me... Okay, go on. I didn't answer your question yet. So basically what's happened is they've come to Telegram saying, we want you to unlock these things. And Telegram is saying, you have no legal right to ask us that. Telegram's in the right there. And French government, you must pursue this by getting each government where the keys are to agree to giving them to you. And no one has ever succeeded in that. Today, Telegram hasn't ever handed over a single bite to any authority anywhere. And so the French, in their frustration, have arrested the leader. Now, the leader also doesn't... He's not a leader, he's the founder.
Well, whatever you want to call him. It's important to distinguish. He's not a leader. He's not Macron. He's not Big Brother. So what does it represent that they arrested him? It represents their failure to be able to find legal means to get to that content. And in their frustration, they're using bullying, basically.
Well, bullying, I don't know whether you call the state bullying. I don't really understand why. Maybe we can do a whole show on it. I don't know what your problem is with the law or authorities. Maybe you got in prison when you were a young man. I did, twice. Well, that probably explains why you're so suspicious of the state. I like people going to jail when they've done wrong.
Well, on the Durov case, I think I'm not necessarily... in support of him going to jail. But what's interesting, it seems, and in this sense, it's like the Brazil case, is the governments are taking it into their own hands. And they have every right on French soil or on Brazilian soil. They can't, you know, the Brazilian government can't close X in the US or France, but they can in Brazil.
No, but what I'm saying is that here we have, and you and I have talked about this a lot on the show over the years, this growing conflict you see in historical terms between a networked age, a post-nation state age, and the old nation state. I don't necessarily see it in those terms, but it's certainly a major conflict brewing.
Well, let me finish my sentence. Go on. What is on social media other than the people? I'm not running. Durov isn't making any posts on Telegram. It's his users. They are the people. And what the government objects to is what the people are doing. And they're saying to social media, tell us what they're doing, spy on them. And just like if that was happening in every cafe in Paris, that you had to put microphones in place to spy on conversations, we'd all be up in arms.
Well, but if, hold on, first it's not the people, because not everybody's on it. I'm not on social media. Well, whoever is on it is part of the people, though. Well, but that's another question. But if everyone in French cafes were dealing drugs and child pornography and laundering money and doing whatever everybody else is doing on the dark web, then you would have microphones and police in every French cafe.
I will, I will. The real reason in Brazil is that the political class want to silence their opponents. And they're accusing X of facilitating the opponents in affecting the outcome of the election. So Brazil, it's a very clear-cut case of state censorship against political opponents. And the judge, by the way, who passed this thing is essentially a dictator in brazil well compared to i'm guessing that musk and
bolsonaro are somewhat friendly so yeah this is a more of a leftist so that depends on your politics but come okay so i take your point i'm sure there's some truth to that about brazil come back to the issue of french cafes and people doing illegal things in them wouldn't that trigger lots of cameras and microphones, surveillance in those French cafes?
I think London's a very unusual, extreme example of surveillance. I agree with you. But I do think that... lots of legal private things happen both in cafes and on social media. For example, people having affairs outside their marriage is a good example. That's what I said, legal and private. Now it's, it's almost impossible, especially in France.
I mean, I think it's illegal not to have an affair, isn't it?
Words and timings
Imean,Ithinkit'sillegalnottohaveanaffair,isn'tit?
Speaker 1
And, and, and, and moving right on. Um, uh, uh, if you insist on spying on illegal private things, by definition, you have to spy on legal private things as well. There's no ability ahead of the surveillance to understand what the content's going to be. And by the way, that's in the US, there are laws against the secret services spying on its own citizens.
And that comes back to section 230 and the post office and whether the state has the right to open people's letters and all the rest of it. But it is interesting the way in which, I mean, I guess if you had to guess two governments that would take the law into their own hands, you would certainly guess the French and maybe the Brazilian as well. What about on the Telegram case? Obviously, Durov is Russian. He's based, I think, in the Gulf.
Well, Durov is the typical anti-Putin hero. Putin sees Durov's prior company because Durov refused to disclose to Putin things that were happening on that platform. So Durov is consistent. He basically supports his users' privacy. And he's done the same here. Now, here's a guy, and this really shows the trend, where Putin and Macron are wearing the same hat. They want to break privacy.
I do not. I mean, and again, this is where you're right. I don't break the law. If I'm dealing in child pornography or selling or buying illegal drugs.
Other people might be. But many, many people are breaking the law on Telegram. So whether or not Durov is responsible, that's more complicated legally and morally. But your argument suggests that these platforms, Telegram, WhatsApp, that they should have a kind of carte blanche, that the governments have no right to access them.
Actually, I'm not saying that. I'm saying the platforms have every legal right to refuse. I'm not saying governments shouldn't try. Just like Police and agencies can seek out criminals within nation states or across borders even, but they have to be successful on their own terms. And in this case, there is no legal right for the French government to break the encryption keys of telegrams. And so they are circumventing the law by arresting him. And they are the ones who are seeking to invade the privacy of legal users of Telegram, not just illegal. Of course, illegality hides in plain sight everywhere in the world. I'm not in favor of the illegals getting away with their illegality, but I'm also in favor of private citizens being private. So I wonder whether Telegram is...
sort of in the gray area between the dark, and I mean that literally and metaphorically, in the gray area between the dark web and what most people, you and I certainly, and most of our listeners, I guess, think of as the internet. Will this case and cases like this result in a deepening, a darkening of the dark web? And then an internet where you don't have platforms like Telegram, which might not be such a good thing.
Do you see my point? I do see your point. And I think to answer it, one has to kind of step back a bit and look at the tools available to criminals to pursue private criminality. Telegram isn't even a good one. Signal would be way better. And there are other self, you know, software is easy to write. I'm pretty sure really sophisticated criminals have their own software that isn't even known to us. So it's very weird. to claim that Telegram is the center of criminal conspiracy. I really doubt that because the criminals would have to be stupid because there are better tools out there.
Well, there are some stupid criminals. The other interesting thing about the Durov and the X case is they're such different platforms. X is in the business of visibility. You don't go on X to do anything privately. It's always public. whereas Telegram is premised on privacy, on secrecy. I don't know whether that's... I guess what people find distasteful about X is the kinds of things which are said publicly, whereas what governments are troubled with on Telegram are the kinds of things that are said privately.
Yeah, I mean, if you really want to start with the idea that the French government has some right on its side here... you are compelled to look at WhatsApp, which is end-to-end encrypted, even in public. You're compelled to look at Twitter direct messages, which are private. You're compelled to look at almost any service that provides encrypted end-to-end, including iMessage, Apple's...
But you would have a right if there was, God forbid, another terrible... bombing in France, and the French authorities got hold of the WhatsApp or X accounts of the bombers, wouldn't they have a right to access that data?
Well, let's assume that the criminals in that case were clever enough to use iMessage. Apple doesn't have the keys. It's actually incapable of breaking its own encryption. So that is Apple making... Okay, but that's Apple. But what about the others? I don't know the specifics of each case, but some of them will have the keys. And some of them, by the way, as Zuckerberg proved this week, do concede to governments what the government is asking for.
So what was Zuckerberg up to here? Was he schmoozing with Trump in his all-too-public letter about the Biden administration pressuring the company to censor content related to COVID?
You know, hard to tell. I think maybe he was having a mea culpa moment personally. And in the context of what's going on in Brazil, maybe also with what's happening in France, was triggered to go on the record about something he's believed for a while, possibly. But certainly what he said is he regrets having been so subservient to government requests in the past, because in retrospect, both on COVID and on the Hunter Biden case, it turned out that the Biden government was wrong.
You know, you can't really trust him to be that strong because he's shown in the past he's prepared to blow in the wind. So we'll see. But yes, on the face of it, yes, I think that is probably the tone of it.
It's all very interesting, Keith. And meanwhile, in California, where you and I live, there's a growing storm over regulation over AI, a Californian bill that I think Newsom has or will sign. It's called SB 1047. It's reached his desk. What do you make of this? Is this part of the this imminent storm between government regulation and internet companies, especially with AI, which is, of course, the new, new thing?
Yeah, I think this is a slightly different framing is required to understand this, but they do overlap. I think this is really to do with, on the face of it, safety. AI safety, in quotes, which is broadly defined as meaning what AI says and also what it can or cannot do.
It's watermarks. And what's interesting is that OpenAI, Adobe, and Microsoft all support it too. Do you think there are some companies, some internet companies, maybe Microsoft, even Apple, which are actually celebrating what's happening in Paris and Brazil. And they think it's a good thing. It benefits them.
I doubt that. I think encryption is so deep now. You know, one of the big issues in the French case is you need to have a license from the French government for encrypted communications, which is a new fact. Nobody knew that before. I assume it's true. And, you know, encryption software is open source. So anyone can build.
The old Eric Schmidt remark about privacy and secrecy says, I don't have any problem revealing everything I do because I don't break the law. Isn't that? A fair point. I mean, if you're not breaking the law, then what problem would you have the government seeing what you're saying?
Well, Eric Schmidt is a bit of a hypocrite because Eric Schmidt is a well-known philanderer who does lots of private things he would not want to be disclosed.
I would say most human beings have parts of their life that they absolutely want to be private. And there's nothing wrong with that. It's reasonable. I once heard you, Andrew, give a speech that you'd rather live in a city than a village because you want to be anonymous. And anonymity and privacy go together. You can't really be anonymous unless you can be private. So I can't remember which book you were launching when you said that, but it was some book launch.
Yeah, although those are two different things. I'd rather be in a city in the sense that one has the freedom to invent and reinvent oneself, but it doesn't give one the freedom to break the law.
I'm not sure that's the case, though. Are you suggesting that the... Coming back to this, and then we need to leave it and move on to some other stories, but are you suggesting the French government, for example, is using this as an excuse to destroy private communications?
No question about it. Every government wants to do that. Every government hates... Why? Because... Private citizens being able to engage in encrypted conversations breaks every secret service in the world.
What about letters then? I mean, physical letters. Does every government want to open every letter we send to one another? Or historically, they didn't want to do that. They haven't changed.
Well, there's lots of evidence that they... But you still need the law. I mean, leaving aside the French case, you still... Look at Watergate, Andrew. Watergate was a president... breaking the law against his competitors. The state is quite clearly capable of wanting to surveil legal opponents.
Look at Snowden. Snowden exposed the fact that the U.S. Secret Service, using foreign governments, was snooping on every American citizen. Every American citizen. Every phone call.
Anyway, let's move on. Maybe we'll come back to this. Honestly, Andrew, you have to read it. They use keyword technology to trigger deeper surveillance. It's all real. So I don't think there's anything wrong with wanting to protect privacy. And if you use criminality... Of course, none of us want child pornography anywhere, and we want those people locked up. But if the price of doing that is everyone's privacy goes out the window, that's a big price to pay from a societal point of view.
Yeah, I am not convinced, spoken like a true libertarian, but this will be a subject we come back to. Lots of other news this week, especially associated with AI, OpenAI, according to the journal, is in talks for a funding round valuing it above $100 billion. You predicted a few months, maybe a year ago, Keith, that OpenAI would be worth, what, $5, $10 trillion? What was your prediction?
And you, I've always been ambivalent, but maybe you're right on this one. You think OpenAI is ahead of everybody else, head of Google, head of Anthropic. You still think that?
I think Anthropic is certainly neck and neck now. Anthropic's really good. And Amazon just announced a big distribution deal where Anthropic's system is going to be behind ALEXA, which I don't want to say because my ALEXA will respond if I do. So Anthropic is really doing a good job of being in the race as well.
But is this, I'm thinking from your expertise as an investor and valuations, this $100 billion valuation doesn't seem, if OpenAI is indeed one of the leaders, if not the leader in AI, it's not an unrealistic valuation, is it?
It's 30 times their revenue, according to the rumors of their revenue being at $3 billion now. So it's certainly within reason for a high growth startup to be valued at 30 times revenue. And if their revenue gets where I think it will get, which is more in the hundreds of billions, then those higher valuations are going to happen.
And does this mean that... all those legal corporate shenanigans of last year with Altman getting thrown out and then coming back, that that's all been resolved and that OpenAI now is just another standard for-profit startup, but all the non-profit, the effective altruism roots of OpenAI have been done away with.
I think that's still in transition. There's still quite a few structural legacies that make the not-for-profit part and the board of that still having quite a lot of power. But it is under Altman's control now. This new money, by the way, a lot of it's coming from Apple, apparently.
Yeah, which is another of your pieces that you connect with. Of course, Apple is investing in open AI. Is that Apple waving the white flag and acknowledging that it can't do AI itself?
Not just Apple. I mean, Amazon's wave the white flag as well. It is quite astounding that the companies that started down this path, I don't know, five or more years ago, are way ahead of big tech. And Google's the exception. I think Google's systems are pretty good. Not yet fully there, but catching, for sure, catching up. But Google's got a problem, which is it can't really deploy its success without damaging its search. business model. So Google's in a bit of a difficult situation.
And Google have another problem this week, an antitrust lawsuit against it, which was filed by Yelp, which you connect to. Is this significant? Yelp's always been a critic of Google. They've always been resentful, I guess, of Google's power and reach.
Yeah. You know, I watched the Yelp CEO on CNBC. Is that still Stoppelman or is it somebody else? I can't remember his name, so I'm not sure of the answer to that. Uh, uh, any, anyway, they're just being opportunistic in light of this court finding that we discussed last week. They're saying, okay, well, if Google has a monopoly, then it also has a monopoly in local search and we're impacted by it. So we're going to file a lawsuit sitting on the shoulders of the findings of the previous one. Of course, the previous one is going to be, uh, resisted by Google, they'll appeal it. And so we're several years away from anything meaningful happening here.
Well, at least it gives those lazy lawyers at Google something to do, right? They need something to do. Keep them off the streets. We are moving on, Keith, to Startup of the Week. Surprise, surprise, an AI company, the generative AI coding company, Magic, which we've been talking about, Eric Schmidt, landing a $320 million investment from Schmidt, Atlassian, and others. What is it about Magic that's so interesting for you?
Well, Magic is focused on an area that a few others are too, which is automated software engineering. And I think after teaching, software engineering is probably the job most transformed by ai you use it a lot don't you i use it constantly uh it it in fact i couldn't do what i do without it it's it it it makes me better in so many different ways uh than i actually am so yes but this is a bit beyond what i do i i use ai as a co-pilot where i'm driving this is actually using ai more holistically as an initiator, like fully end-to-end. And so I think it's kind of interesting. And Eric Schmidt, whatever we think of his philandering, he's a very astute observer of innovation.
And actually, I tend to think that it's on the coding front that AI has the most potential to change the world as we as we uh went to broadcast today as an interesting piece in the new yorker by one of the best science fiction writers around especially on ai ted chiang uh who wrote a piece the weekend essay this week you didn't get into your newsletter why ai isn't going to make art i tend to agree with him on that front i think it commodifies a certain low-end cheap art but So the real opportunity, I think, for AI may be in coding rather than in art. But maybe we can come back to that next week. Finally, Keith, X of the week from your favorite Xer. Not everybody's favorite Xer. A certain Elon Musk on the Californian case, SB 1047. What is Musk saying about 1047?
Well, I put it alongside his other tweet, which was for free speech. But this one is supporting the California, let's say, modified legislation because he has been on the record saying that AI can be dangerous.
But it does show that Musk is also capable of hypocrisy.
Words and timings
ButitdoesshowthatMuskisalsocapableofhypocrisy.
Speaker 2
Oh, that's it. That's that's news. Keith, are you still half in love with him? What does he have to do to make you hostile to him? It seems like, uh, whatever he does, you still rather like him.
Well, I think you can, um, you can, uh, have more than one opinion about a person and both can be true. And my opinion of him is he's one of the most important innovators the human race has produced. And at the same time, I think he's like a teenage child in a tantrum often. And both things can be true.
And what about the way he, I think one thing that pisses some people off, probably including myself, is the way he seems to use X now just as his own platform in the same way as Trump uses truth social.
Well, he always did that, but now he's the owner. I don't think his behavior's changed. It's just that our perception changes because he's the proprietor. But he is a risk taker, right? I mean, he's prepared to open up his reputation by saying what he really thinks, which in some ways, is admirable, even when you disagree with him.
Well, in Brazil, it seems to have got And to have his company shut down. It'll be interesting to see what will happen. We are entering September, Keith. August supposedly is a time of no news. Lots of news this week. And we will see you again next week where I'm sure there will be much more technology news and this growing storm between conflict between governments and tech companies. So have a great week, Keith, and I'll see you again next week. See you, Andrew.