High Signal Podcasts Evidence ledger
Method
Browse
← Back to evidence

Evidence receipt / preference

Published · transcript-backed

Lex Fridman: preference

3 Sept 2024 Lex Fridman Podcast #442 – Donald Trump Interview

“I think generally governments should stay out of censoring or even pressuring social media platforms, and I think arresting a CEO of a tech company for the things said on the platform he built is just nuts.”

— Lex Fridman

Source trail

Everything needed to verify it.

Speaker
Lex Fridman
Attribution
Verified speaker
Claim type
preference
Recorded
3 Sept 2024
Publisher
Lex Fridman Podcast

Transcript context

…I still think I suck at, but let’s say it’s a work in progress. You also have to make the scheduling work and set up the entirety of the environment in a way that is conducive to such a conversation. This is hard, really hard with political and business leaders. They are usually super busy and in some cases super nervous because, well, they’ve been screwed over so many times with clickbait got you journalism, so to convince them and their team to talk for two, three, four, five hours is hard. And I do think a good conversation requires that kind of duration. And I’ve been thinking a lot about why. I don’t think it’s just about needing the actual time of three hours to cover all the content. I think the longer form with a hypothetical skilled conversationalist, relaxes things and allows people to go on tangents and to banter about the details because I- … agents and to banter about the details, because I think it’s in the details that the beautiful complexity of the person is brought to light. Anyway, I look forward to talking to more world leaders and doing a better job every time as I said. I would love to do interviews with Kamala Harris and some other political figures on the left and right, including Tim Walz, AOC, Bernie, Barack Obama, Bill and Hillary. And on the right, J.D. Vance, Vivek, George W. and so on. And on the topic of politics, let me say, as an immigrant, I love this country, the United States of America. I do believe it is the greatest nation on earth, and I’m grateful for the people on the left and the right who step into the arena of politics to fight for this country that I do believe they all love as well. I have reached out to Kamala Harris, but not many of the others. I probably should do a better job with that, but I’ve been doing most of this myself, all the reach out, scheduling, research prep, recording and so on. And on top of that, I very much have been suffering from imposter syndrome with a voice in my head constantly pointing out when I’m doing a shitty job. Plus a few folks graciously remind me on the internet, the very same sentiment of this aforementioned voice. All of this, while I have the option of just hiding away at MIT, programming robots and doing some cool AI research with a few grad students, or maybe joining an AI company or maybe starting my own, all these options make me truly happy. But like I said, on most days I barely know what I’m doing, so who knows what the future holds. Most importantly, I’m forever grateful for all of you for your patience and your support throughout this rollercoaster of the life I’ve been on. I love you all. ng, so who knows what the future holds. Most importantly, I’m forever grateful for all of you for your patience and your support throughout this rollercoaster of the life I’ve been on. I love you all. Okay, now let me go on to some of the questions that people had. I was asked by a few people to comment on Pavel Durov’s arrest and on X being banned in Brazil. Let me first briefly comment on the Durov arrest. Basic facts, Pavel Durov is CEO of Telegram, which is a messenger app that has end-to-end encryption mode. It’s not on by default, and most people don’t use the end-to-end encryption, but some do. Pavel was arrested in France on a long list of charges related to “criminal activity” carried out on the Telegram platform, and for “providing unlicensed cryptology services.” I think Telegram is indeed used for criminal activity by a small minority of its users, for example, by terrorist groups to communicate. And I think we all agree that terrorism is bad. But here’s the problem. As the old saying goes, one man’s terrorist is another man’s freedom fighter. And there are many cases in which the world unilaterally agrees who the terrorists are, but there are other cases when governments, especially authoritarian inclined governments, tend to propagandize and just call whoever’s in the opposition, whoever opposes them, terrorists. There is some room for nuance here, but, to me at this time, it seems to obviously be a power grab by government wanting to have backdoor access into every platform so they can have censorship power against the opposition. I think generally governments should stay out of censoring or even pressuring social media platforms, and I think arresting a CEO of a tech company for the things said on the platform he built is just nuts. It has a chilling effect on him, on people working at Telegram and on people working at every social media company, and also people thinking of launching a new social media company. Same as the case of X being banned in Brazil. It’s, I think, a power grab by Alexandre de Moraes, a Supreme Court justice in Brazil. He ordered X to block certain accounts that are spreading “misinformation.” Elon and X denied the request, then de Moraes threatened to arrest X representatives in Brazil, and in response to that X pulled the representatives out of Brazil obviously to protect them. And now X, having no representatives in Brazil, apparently violates the law. Based on this de Moraes banned X in Brazil. Once again, it’s an authoritarian figure seeking censorship power over the channels of communication. I understand that this is complicated because there are evil people in the world and part of the role of government is to protect us from those evil people. But as Benjamin Franklin said, “Those who can give up essential liberty to obtain a little temporary safety deserve neither liberty nor safety.” It’s a trade-off, but I think in many places in the world, many governments have leaned too far away at this time from liberty. to obtain a little temporary safety deserve neither liberty nor safety.” It’s a trade-off, but I think in many places in the world, many governments have leaned too far away at this time from liberty. Okay, next up I got a question on AI, which I emotionally connected with. I’ll condense it as follows. “Hello, Lex. I’m a programmer and I have a deep fear of slipping into irrelevance because I am worried that AI will soon exceed my programming skills.” Let me first say that I relate to your fear. It’s scary to have a thing that gives you a career and gives you meaning to be taken away. For me, programming is a passion, and if not for this podcast, it would probably at least in part be my profession so I get an uncomfortable feeling every time, Claude, the LLM I use for coding at this time just writes a lot of excellent approximately correct code. I think you can make a good case that it already exceeds the skill of many programmers, at least in the same way that the collective intelligence of stack overflow exceeds the skill of many individual programmers, but in many ways it still does not. But I think eventually more and more the task, the professional programming will be one of writing natural language prompts. I think the right thing to do, and what I’m at least doing is to ride the wave of the ever improving code generating LLMs and keep transforming myself into a big picture designer versus low-level tinkerer. What I’m doing and what I recommend you do is continually switch to whatever state-of-the-art tool is for generating code. For me, currently I recently switched from VS Code to Cursor, and before that it was Emacs to VS Code switch. Cursor is this editor that’s based on VS Code that leans heavily on LLMs and integrates the co-generation really nicely into the editing process. It makes it super easy to continually use the LLMs. What I would advise and what I’m trying to do myself is to learn how to use it and to master its co-generation capabilities. I, personally, try to now allocate a significant amount of time to designing with natural language first versus writing code from scratch, so using my understanding of programming to edit the code that’s generated by the LLM versus writing it from scratch and then using the LLM to generate small parts of the code. I see it as a skill that I should develop in parallel to my programming skill. I think this applies to many other careers too. Don’t compete with AI for your job, learn to use the AI to do that job better. But yes, it is scary in some deep human level, the threat of being replaced. But at least I think we’ll be okay.…

Stored transcript either side of the excerpt. The highlighted words are the published quote; the surrounding text is unedited source, never generated.

Search evidence