Pivot: Reid Hoffman joins Kara Swisher and Scott Galloway to go deep on AI news
01:01:35 Share

Hi, it's Reid Hoffman, and today we're bringing something a little different to the Masters of Scale feed. Recently, I had the pleasure of joining my friends Kara Swisher and Scott Galloway on their podcast, Pivot, to talk about everything from the very real hype and fears surrounding DeepSeek to why I believe investing in AI is a good thing. Keep listening to hear the full episode, and if you enjoy it, be sure to subscribe to Pivot with Kara Swisher and Scott Galloway so you never miss an update.

Let me just say, Amanda loves your whole, this whole jam, Scott. She was like, Scott's on fire. I love it. She was vaguely attracted to you, I think. Vaguely is doing a lot of work there. Hi, everyone. This is Pivot from New York Magazine and the Vox Media Podcast Network. I'm Kara Swisher. And

And I'm Scott Galloway. Scott, we've got a very special guest today, someone I like very much in Silicon Valley, which is an unusual thing. Joining us today is Reid Hoffman. Good morning. So we're doing a new thing on Pivot. We're going to make people stay the whole show. And Reid is our very first guest to do this. We're having a really good friend of Pivot. It's not just a friend of Pivot. It's an extremely special friend of Pivot.

Reid is the co-founder of LinkedIn, obviously, the host of podcasts Masters of Scale and Possible. He's also the author of a new book, Super Agency, What Could Possibly Go Right With Our AI Future? I feel like he's positive about the future. Not on everything. Welcome, Reid, again. It's great to be here. I love being on this podcast with you guys. Good. So we're going to talk about a lot of stuff. And last time we talked was at your event right before the election happened.

Yes.

We're recording this on Thursday morning as the recovery operation is underway in the Potomac. There was a press conference a little while ago. Authorities confirmed there are no survivors. Twenty-eight bodies have been recovered so far. It was a flight coming in from Wichita, and the Blackhawk, they don't know. I'm not going to speak about it because I don't know anything about this.

I don't have any extra information. It's tragic and very, very sad. But the authorities here seem to be doing their best to figure out what happened and moving on. And we're not going to politicize it. We're not going to do everything that has already broken out online. Any reaction from either of you? You don't have to have any whatsoever. But Scott? Well, look, these things, obviously, it's a tragedy. The only thing...

There's such a spectacle that they attract a lot of media. And I believe about a thousand people a week in the U.S. or about 800 people a week die in automobile accidents. But they're not nearly as dramatic or as much as a spectacle. And, you know, when I hear this, it's a tragedy. But at the same time, when I look at the data, it's just sort of incredible news.

that it doesn't happen more often. I mean, what you said at the very top of this, it's the worst disaster in over 15 years because it doesn't happen very often. So I'm not saying it doesn't warrant scrutiny. I'm not saying it's not a tragedy. I do think it's just a miracle, though, that this form of transportation is as safe as it is. So anyways, I look at it

almost as a glass half full, if you will. Yeah, except for the people who died, of course. Go ahead. 100%. Well, and plus one, Scott's comments, obviously hearts go out to the families and all the people. It's a huge tragedy and loss. On the other hand, there's this tendency to try to blame the FAA. And you look at the fact is that it's much safer to be flying than it is to be driving.

And so, you know, there's a credit to how the whole, you know, aeronautics and air transport system works. And there's going to be, I think, overly much of a...

witch hunt on the FAA camp where actually, in fact, I think, you know, the system should also be acknowledged. I heard the air traffic controller was a lesbian. Your thoughts, Kara? Oh, anyway, I'm sure it was DEI. If they start with that, I'm not going to have any of it. It's ridiculous. There aren't enough air traffic controllers. And obviously, there's been a lot of uncertainty around the federal workforce right now, but it has nothing to do with this particular track.

But that said, it's a wonderful system that we have. And at the same time, just for anyone, I live in D.C., flying into DCA is terrifying. I find it terrifying. There's so much air traffic going on. You see helicopters, helicopters.

There's a weird twist that you have to do because of all the federal buildings, including the Washington Monument and the White House. And so whenever you're coming in, I find it always makes me nervous to come in because it's such a highly trafficked area with military and everyone else. But we'll see what happened here because you're right, it never happens. But let's move on. And we're so sorry for the families of all the people that were killed.

We're going to talk all about the deep-sea craziness in a sec. But first, I want you to start this. Why are you so positive about our AI future? I know you've talked about things that could go wrong and stuff. Your position on AI is somewhere between doomers, gloomers, and zoomers. You call yourself a bloomer, accelerating toward a bright future but managing risk. Can you just explain yourself so we can put you in the – where we put you on the map here? Well,

Well, I think in both creating immense value for humanity and also navigating risks, I think our future will have a much stronger tool set. So in the positive category, thinking about the fact that you can have a medical assistant that's better than today's average GP is

on every smartphone running for under $5 an hour for anyone who has access to a smartphone for doing that. A tutor on every subject for every age group. And then, of course, the fact that it's kind of the cognitive industrial revolution of increasing productivity in a lot of different vectors. And I think all of that is extremely positive. Now, that doesn't mean there aren't risks to navigate and some questions, you know, to navigate in kind of good ways. But

That's why I think I'm fundamentally an optimist and fundamentally also an accelerationist in this direction. But one of the things that a lot of people are talking about is cost to zero for lots of things. I mean, Marc Andreessen, I've heard it from Vinod, this idea that everything will cost pennies. Sometimes you all go over the top. So that's why people question your, not credibility for you, but some people's credibility. Well, that's the difference between bloomers and zoomers, right? The whole notion of...

Everything created with technology is going to create an abundant Star Trek universe immediately. Abundance is their favorite word. Yes. You know, freely across this is, you know, kind of a kind of exponential, you know, an exponentialism hysteria that I think actually, you know, doesn't have particularly good thinking. I think the question about saying, though, that we can create so much better of a future with technology and with AI is very important and just kind of drive, navigate intelligently.

So, Reid, I'm genuine with this question because I think you're one of the brightest blue flame thinkers in the world of technology. I'm sort of blown away by deep seek. And I want to I want to posit a hypothesis with you and think there's any and see if there's any merit. And that is we were just talking about the airline disaster, the air crash. And I was thinking about the airline industry. I was thinking about PCs.

We can skirt along the surface of the atmosphere at point eight, the speed of sound. It's added remarkable valuable to the economy and to consumers, air travel, jet air travel. PCs have revolutionized the world. Yet neither of those industries were able to capture or any specific companies were able to capture a great deal of shareholder value. It was consumers and the general public that got captured most of the value.

And I'm wondering if deep seek is in fact a signal that in fact, AI may be one of those industries where there's not a small number of companies that are worth a couple trillion dollars, but there's so much competition and the barriers are so low that it might have unbelievable winners,

But those winners will be further dispersed into the general public and the economy. And we might not have a small number of winners as we did in social or search, that the big winners similar to the airline or the PC industry might be the public, but a lack of really big winners in tech. Your thoughts? Interesting thesis. I would tend to think it'll be more like kind of software, you know, internet dynamics where

But I actually don't think that's necessarily because there's only one or two. I mean, I think one of the things that the Internet brought about with it is, you know, previously when it was kind of like hardware dynamics and PCs were dominant, it was part of the thing where everyone was like Microsoft's going to be competing with Disney and with airlines and everything else because it's the primary software OS and the Internet opened it up to allow, you know, kind of Google and Amazon and so forth. So I think there's

going to be multiple. And I think we're more, call it seven big tech companies heading to 15. But I don't think it's going

I think that the same kind of dynamics where you have like a network effect for a social network or a LinkedIn or a, you know, kind of an enterprise integration or a, you know, kind of the way that they kind of AdWords and buying the search traffic, you know, kind of works for, you know, kind of Google. I think those dynamics probably will still be present today.

But that doesn't mean that, you know, like I'm a venture capitalist investor at Greylock. I invest in a number of different startups. And I think there will be a like just literally a field of interesting companies. And I think that the public will benefit a lot from that, just kind of like the way they benefit a lot from, you know, Wikipedia, the Internet, you know, free communications and a bunch of other things. But I don't think it's necessarily kind of interesting.

you know, kind of completely broadly dispersed. You're not going to see it. So why don't you just go into, because one of the, some of our listeners are like, we didn't say enough about DeepSeek. Well, honestly, we don't know. Scott and I don't know, like you, for example. So it caused a frenzy in tech and markets. And so we brought, we brought Reid in to explain listeners, because we're idiots, fine. So DeepSeek caused this frenzy in tech and the markets. I'd love your, your

take on what we're seeing and what excites you about it and what worries you. And I will note that OpenAI, where you were an early investor, you're no longer on the board, and Microsoft, where you are on the board now, are investigating whether OpenAI's data was stolen to build DeepSeek's model. They were also relying on open source stuff like Lama from Meta, et cetera. So talk a little bit about, give your take on this and make it a hot take because that's who the person you are.

Yeah, happy to do it. By the way, one of the benefits of me being able to speculate is I have no internal information from either Microsoft or OpenAI, so I can speak entirely as a, you know, outside commentator just looking at this. So DeepSeek released a, you know, highly competent model from, you know, kind of China and, you know, kind of personalized.

part of the reason I took the market by storm is the thesis was that it was created for a lot less money and a lot less compute. And what I think is there's certainly some parts of the story that are incorrect. The thing that we're trying to figure out is which parts of it are incorrect. And I'm

I would speculate with some vigor that they actually had some version of access to larger models in helping training. Because this is actually something we all knew already last year, year before, is that large models will help train small models. And so that means when you train a small model effectively,

but you need a large model in order to do it, that's actually not disproving the need for these scale systems because when you have the better and better large scale models, you'll be able to train also really, really good. So they were riding off of your rails, in other words. Exactly. Right. And so I would hazard certain,

strongly that there's something like that in the background. Now, it could be that they had some kind of access to chat GBT. Certainly, some of the data and evidence suggests that in terms of the way that it answers and does certain things. It could be that they actually had access to a compute cluster of size because the so-called training run really makes sense. And I've cross-checked this across multiple groups.

you know, like outside groups saying, hey, you know, what is it? What makes sense here? And they're like, yeah, for the final training run on a serious compute cluster, that could be the dollars that was spent on this in order to make it happen. It doesn't include talent. It doesn't include all these other things. Right. So as they were saying, it was five to six million. No one thinks that because, again, the talent that they hired, they put in place. It was a lot of younger people, correct? That's their story. Right.

that it wasn't highly paid. Yeah, exactly. And I think that the, I think it's nearly certain that it's dependent upon the large scale compute, the larger models in some way. The only question is we don't know in what ways and how. And I think that's one of the things that everyone's investigating. And so I think the kind of market frenzy on, oh my God, AI can be here without large scale compute. And of course, by the way, AI can be here. We invest in startups that train small models and so forth.

But the large models still bring certain critical elements to the table in terms of ability to change small models, ability to get to performance. Like if you say, well, hey, moving from 10,000 GPUs to 100,000 GPUs, we only get a 20% better coder, medical assistant, legal assistant, tutor. Like, well, in a wide variety of those areas.

That really matters. That increase in cost when you amortize it across people accessing it across the entire unit, the billions of people that could use it, that's actually completely worth it and makes total economic sense. The cheapness of it. What excites you about it and why did it have such a market impact from your perspective? You saw Nvidia got crashed, it came back up, but this was the idea because the economic underpinnings of this,

which everyone's worried about this amount of money, $50 billion, $80 billion, et cetera. Microsoft is 80, I think. Yeah, by the way, I think that all of these questions are in the classic kind of like, you know, short-termism versus long-termism. Because if you're saying, hey, I'm spending, I'm doing, you know, kind of building this kind of CapEx thing of $50 to $80 billion, you know,

A, I can train much better intelligence, but B, I'm also these are data centers in terms of serving intelligence through various apps to the world.

You know, you could say, well, the payoff is longer than I'd like as a public market. I think the payoff to be three years and maybe it'll be five years or seven years. You know, that's the kind of range you're talking about. So I find that the general discussion on the, you know, X tens of billions of dollars to be short sighted. Generally, me as a private citizen, me as a venture investor, obviously making zero comment as a, you know, kind of Microsoft board member.

But the and so I think that it's an extremely important area to be investing in. And I'm actually glad that that we as a industry are doing this. I mean, one of the things that I've kind of thought about over December was that I want artificial intelligence not just to be amplification intelligence. I want it to be American intelligence in terms of the thing. They're always bringing the Xi and Xi or me kind of argument. Go ahead. Yeah. Yeah. And I think.

By the way, part of the criticism I used to get last year and the year before when I was saying, hey, look, we are game on with China, was like, oh, you're just trying to get the excuse that we shouldn't be interfering with you and slowing you down. And so it was like, no, no, I see the competition coming. And one of, I think, the huge virtues of DeepSeek is kind of like, yes, there it is.

Right. That that is serious and real competition. And and I think that's that's that's the world we're in. Yeah. Who would have thought that China would engage in IP theft to create a cheaper product? We've never seen that before. I love that AI is is taking the job of AI. Yeah, that's the big joke. That's what Jon Stewart said. Yes. So another another thesis, most consumer markets, if you think of this as not only B2B, but B2C.

They bifurcate. It becomes Walmart or Tiffany. It becomes Android, which is essentially free and it's ad supported or it's iOS. Isn't this potentially just the first sort of shot across the bow where this market is going to become similar to every other consumer market where we're going to have Walmart and Tiffany? And this is kind of the first entry into the Walmart 80%.

Kind of Old Navy. Old Navy is 80% of GAAP for 50% of the price, and that hits a large market, but a lot of people want GAAP or Banana Republic, that there's a market for both of these, if you will, both respective positionings. I think yes. And actually, in fact, when we kind of talk about

What is our agenda in the future? Most people tend to think, you know, there's going to be like, you know, be Lord of the Rings. There's one agent to rule them all. And actually, in fact, I think there's going to be, you know, in a sense, more agents than people because every person is going to have multiple agents. Now, there may be a limited number of code bases. Call it, you know, 1,000, you know, kind of powering all the kind of different agents. Could be 10,000.

But I think we're going to actually live in a, in a very rich Asian environment. And that's going to have a, it's not just going to be the kind of the Walmart or Tiffany's. I think it's going to be, you know, this,

This agent's actually particularly good at travel. This agent's particularly good at, you know, kind of interpersonal discussion. This agent's particularly good for, you know, people like Scott and this other agent's particularly good for people like Kara. Right. So an agent that writes better dick jokes. Go ahead. Go ahead. Yes, exactly. No, no. But the point is like, you know, this one will be sardonic and a little sarcastic. And Scott's like, yeah, that's that's my agent.

And this agent is going to be fiery and say, no, we got to hold you to a higher standard. And that'll be the care agent. And then, you know, then you'll have the nuanced agent of the one for Reed saying, well, on one hand and on the other hand. Yeah, yeah. That would be your agent, wouldn't it be? So let me, speaking of that, what he was talking about, you just raised 24, $25 million in funding for AI drug discovery startup, focusing on cancer research. This is something that, you know, Sam did this at the White House event, for example, he,

And your book is called Super Agency. So this is drug discovery. Super agency is what you're writing about. Of these many things, where do you think

the quicker is and the more important for the economy. And explain what super agency is, what you were just saying, correct? Yes. Well, super agency is what happens, and we've seen this from everything from, you know, the fire and agriculture to the printing press to, you know, the car and electricity, is what happens when millions of people all get access to the same superpowers

of increasing their agency at the same time. And we collectively, you know, get super agency. Like, for example, when I get a car, I don't just get a place to be able to, you know, get a broader geographic mobility. My friends can come visit me. A doctor can come do a home visit. You know, things can be delivered into kind of the local neighborhood. And so that's kind of the super agency. And the thesis is just like everything else with

you know, kind of, you know, the internet and the mobile phone, that we will all get super agency through AI because the cognitive superpowers that you get, that I get, that Scott gets,

we will get even more superpowers because we're all getting them at the same time. And that is, I think, kind of like it's William Gibson line, the future is already here, but unevenly distributed. That's already present. There's already a whole bunch of different things you can do with AI that most people don't realize is a huge amplifier in both how they navigate their personal life and how they navigate their

their kind of future life. Now, Manus is kind of my, you know, kind of thinking about like, what are the things that these AI things can greatly do to massively improve

kind of human life and human condition that are kind of in a different direction that most people are heading. So Manus, for example, is not building an agent. It's kind of saying, actually, in fact, the biology of our human bodies is a complex language. And part of what's going on with cancer is kind of misfires in that language. And we can take these great...

amplification that you get with AI and take the best of AI and take it with the best of science. And then you can tackle a problem that, you know, five-year-olds get cancer sometimes. It's not just, you know, it's the entire, everybody in the human race at all ages, our system is naturally trying to regulate cancer and all of us are at risk for it. And it's something that we can potentially solve.

with AI. Right. But that's not agency. That's the other things it's going to do, like climate and everything else. That's your focus. And so you're sort of, where is most of your money going? Agents, these agents that you just spoke of, or is it, you know, or is it these other things? Or is it just like electricity? We don't know where it's going to be applied. It can be applied, you know, to everything from light bulbs to, you know, electrocuting people, right? You just don't know what the

application is? Well, I think there's a broad range. And as an investor and as a kind of a theorist and thinker, it's anything that could make a massive difference for the human condition.

So I'm both very pro-Agent, Agent Universe, but I'm also pro all the other applications, whether it's kind of climate change, cancer. And I think that's part of the reason why this technology, like for example, electricity does an entire range of things. It's not just powering light bulbs. It's your heating and now cars and the whole range. Like we can't live in anything like even approximately close

to our modern condition without electricity. So, and I think that's the kind of technology that we're- The comparative, the comparative. Yes. So any amount of money spent is good money spent from your perspective. Well, yes. It doesn't mean that some of it won't be seriously wasted by a bad approach. And by the way, that's classic, as you know, for the venture industry. So there will be, you know, hundreds of foolish investments. But the overall industry-

will create a massively good surplus for humanity and society. In any case, we've got a lot to get to today. We're going to take a quick break, then we'll talk through some of this week's big headlines with Reid.

Scott and Reid, we're back. We've got a lot to get to today. Let's get into some headlines now. Meta will pay Trump $25 million settlement for shutting down his accounts after January 6th. Most of the payment will go towards Trump's presidential library again, much as Disney's payment of $15 million is going to do that. X, who also kicked Trump off the platform after January 6th, says it's negotiating its own settlement with Trump.

Apparently, the Trump people said Mark wasn't going to get into the tent unless he did this. You don't seem to be paying anybody millions of dollars to get in the tent. So what do you think, Reid? So obviously, I think it's— That was a long fly that you just had there. Well, look, I obviously think that the notion of—

of, you know, kind of this sort of payoff is, I think, is, to put it charitably, suboptimal. And I think that the, you know, the question of, like, the fact is when people are removed from services for violations of terms of service, they're removed for violations of terms of service.

And I think that's a perfectly good thing. And I, you know, myself, am a massive advocate for, you know, kind of the rule of law and kind of how these kind of contracts work. But I understand expediency in navigation. Okay. What does that mean? What does that mean?

It's a vig. It feels like a vig to me. That's my... Sounds like a mob move to me. Yeah. You know, I wouldn't want a nice company you got there. I wouldn't want anything to happen to it. Yeah, yeah. Well, let's hope that we see very little of that in the coming years, although obviously we have deep worries in the other direction. Couldn't he keep doing this, suing people and getting these things? Well, I guess the question will be is...

It's kind of like, how much do people kind of respond to this sort of excess pressure on these kinds of things? And I think that that, frankly, you know, we shouldn't want it as a society. And, you know, I think probably there will at some point be a bridge too far on it. And seeing what that

that bridge looks like, I think is still something where we're looking for that bridge too far is. Well, there's two dimensions to this. The first is from a pure shareholder standpoint, it probably makes sense when the president's coming after you to say, and you make $20 million a year in operating profit to say, yeah, just make it go away. Just give them $25 million to the presidential library and make it go away.

The problem is, and I wouldn't expect, based on pattern behavior, Mark Zuckerberg to think anything about this, is that this has real societal implications. And that is, despite the fact Bob Iger made $45 million last year, I would argue he's becoming more and more impoverished in terms of his citizenship.

And that is when a media company says things that are a fraction of the misinformation, slander, disparaging statements that the president has made himself, that happens every day online, and they agree to set precedent by bending a knee and bowing to this intimidation, it sends a chill across the entire fucking nation.

I'm on Morning Joe, and I call the president an insurrectionist and a rapist, and Mika stops the show to clarify he was found guilty of sexual abuse. Liable.

Or excuse me, sexual libel. He was found liable of sexual abuse. Is that the right terminology? Yes, I just made, call me Mika Brzezinski, but go ahead. But look at what we're doing, all right? And by the way, the judge then went on in his sentencing to say the street term for that is rape. So what do you have? You have a group of people, this is straight out of the fascist handbook, intimidate anyone who says anything negative about you.

And if these companies, in my opinion, had more fidelity to American values and the very important role media plays in checking power, they wouldn't be bending the fucking knee like this. So is it a practical thing to do? Yes. You can't argue with that. Has it sent a chill?

across the entire country in what is an incredible double standard that the critics of President Trump are now feel that they're being held to. Look at the shit he has said about people that's been incorrect.

So, you know, yeah, Bob and Mark, I expected this from Mark. I was disappointed to see it from Bob. Where are the men? Where are the Americans who are going to stand up and say, no, if you're not guilty, if we're not, if the entire ecosystem has never been held liable for things much more slanderous or disparaging than these statements, all right.

I'll see you in court. In this case, it wasn't statements. He broke the rules of Facebook and they kicked him off. That's all. It was not as Reid was allowed to do. Yeah, exactly. So when we when we talked in October, I asked if you were concerned about Trump and retribution and your own well-being. And you said, yes. Talk about now. How are you feeling? Because obviously you were one of the more prominent supporters of Kamala Harris, obviously a big Democratic donor. You do give to Republicans also. Right.

for which for some reason you get endless shit. I'm not sure why. I think that's probably fine to do that. How do you feel now? Do you feel in that crosshairs that we had discussed?

Well, it's unclear. I'm hopeful that a bunch of my friends who are around the administration say, look, that's just kind of all rhetoric and, you know, isn't actually going to be... That's what they say to you. Yes, that's what they say to me. And so that obviously is hopeful. But obviously it's one of the things that I think, you know, both me personally and we as Americans need to watch carefully because we do...

to continue to be, you know, the home of the brave and land of the free. And I think that that's very important to be, you know, resolutely against, you know, kind of abuses of state power and

for individual interests. And so, you know, like I said, I went and talked to a whole bunch of people who are, you know, kind of in, you know, and around the current administration said, look, you know, I have this worry. They said, look, we've talked to a bunch of people. They say, this is not something we're going to do. I was like, okay, then,

You know, let's wait and see. Are these the same people that said he wasn't going to let out everybody at January 6th? He's only going to let out—he wasn't going to let out the criminals, but then he did. That's the same people. It wasn't the same people as those. Yeah. He's done that several times. He's done that several times. There's worries. And, like, for example, letting out the people who assaulted police officers—

is, you know, as per what Scott was saying earlier, is a terrible signal. It's basically saying, hey, you know, like if you're doing violence in a cause that I am supportive of, you know, I have this thing of a presidential pardon. And, you know, that's obviously...

you know, frankly, you know, terrifying, concerning. - Right, but you didn't want a presidential pardon, correct? - No, no, of course not. - Yeah, you didn't do anything. - But here's the problem. If I'm a thoughtful guy who's high profile with a lot of business interests with a family, I would be inclined, I'm not just gonna speak for you, Reid, I'd be inclined to keep a very low profile over the next 12 months.

Whereas the people on the right are emboldened to be aggressive and pollute and flood the zone with misinformation and bullshit and attacks. And the people on the other side of the aisle feel like, well, maybe I should just keep quiet for a while because they are removing security details of people, which is nothing but repackage violence, right?

When you took out the head of the Iranian security forces and you ordered that strike as a general and the president, for whatever reason, doesn't like you, is removing your security detail, that is repackaged violence against that person. So what we have is a group. This is the road to fascism.

Keep them quiet. Put a chill across people. I'm going to keep a low profile. Maybe I'm just not going to be quite as aggressive. Let me be clear. One of the roads to fascism, and I want you both to respond to this, is littered with calls or accusations that people are overreacting. Call me overreacting. Let me just say, Amanda loves this whole jam, Scott. She was like, Scott's on fire. I love it.

She was vaguely attracted to you, I think. Vaguely is doing a lot of work there. Vaguely is doing a lot of work there. She was. She wanted me to tell you. I think you're right. But that's I think it's, you know, people that are very high profile, I agree, aren't certainly aren't backing down and don't seem to be disappearing. He's right here. Right. And he's talking about it. So I think the question is when Mark does these deals, when he just doesn't have to and it doesn't help them from a shareholder point of view. I really think that's nonsense.

It's really, but again, we expected Mark to do this. Sorry, I know, Reid, you were, Reid was early at Facebook and has been, had been a mentor to Mark. I don't think he's listening to someone like you anymore. But that said, you and Bill Gates had been and probably are no longer, I would guess, but I don't know. Look, the thing, I think that the American people should pay a lot of attention to the removing the security detail from the,

a person who spent their entire life serving the American people, putting himself in harm's way, kind of helping secure the safety of America, both locally and Americans abroad.

and saying, hey, for petty reasons, I am putting that person directly in the harm's way of violence. And I think that is a, you know, an unpatriotic, you know, personal thing that I think is extremely important that everyone should pay attention to. And I think it's exactly that, is I think that, you know, as Scott's saying, on the things that are like, that is the kind of thing that is deeply un-American, like I think people need to speak up about. And

And so, you know.

This move has Elon Musk's fingerprints all over it, down the subject line of the email sent to those employees of Fork in the Road, which was an email Elon sent to Twitter staff in 2022. I have been stopped by so many government employees saying, what should I do? I said, do not leave because they won't pay you because Elon still has severance issues with those people that were supposed to be paid at Twitter. It's one of these ploys to get people out the door. There's also federal employees have more...

private employees have rights in California, New York especially, but here they have more. What do you two make of this plan? It's unclear whether Trump can even offer this buyout package without budget authorization, if he has the money for it. They're hoping that 5% to 10% of federal employees accept the offer, which could mean hundreds of thousands of people. You know, you've been involved in companies that

do things like this, and obviously you end up paying them unlike what Elon did at Twitter. It looks like he went around everybody. It looks like he went around everybody. This plan was sort of foisted upon people without even Trump officials knowing it was happening. That was the story in the Washington Post today. So, thoughts? Yeah, so it is a technique that when deployed with honor is actually, in fact, something that has some strength in the private ecosystem.

because it's kind of in the private ecosystem, the way it works is if you're not committed to the future of this, then now is a good time to exit because, again,

And it works in the private ecosystem in part because, you know, part of it is not only am I committed to the mission of this, but there's also an economic reason to be keeping going. You believe in the stock options, the bonus plan, all these kind of things for that. Now, I worry in the public circumstance that applied here that a lot of the folks who go, OK, you know, I could get jobs in 10 different places, you know, I could get jobs in 10 different places.

I'm going to go do one of those and do this. And this will be, you know, kind of depriving the American people of some really great talent. So the good people will leave, in other words, which is always a worry when you're doing this, right? When you're doing one of these broad layoffs. Yeah. And that's, I think, unlike in the private ecosystem where you actually have a, and we have a rich incentive plan for staying too, right?

Here, it's like, you know, here's an incentive plan for leaving versus an incentive plan for staying. So the best people leave. And in the case of private companies, they try to retain or they specifically target certain employees. Correct. This is too broad in that regard. So the best people will go with the experience and the worst people will stay. Correct. Or the less good people will stay. Yeah. There's a worry about a selection effect. Right. Did you notice the echoes of what Elon had done here?

Yeah, of course. Yeah. And? Well, I mean, look, I think we will see many things that are the parallel to, you know, how Elon thinks an organization should be turned around from Twitter to the federal government.

The fact that you're looking now where, you know, what was the leaked emails? Like, hey, we're not doing very well, you know, is whatever X years on it is for Twitter. You know, it's like, well, you know, should be learning from that. And that's within the standard understood commercial ecosystem. Yeah. Yeah.

Just so you know, today, you're not going to talk about financial results, but Tesla's results were terrible, even though the stock is going up because they aren't selling as many cars. In any case, Scott? You guys have said it. I don't like buyouts because the people with the most options, the most talented people are the ones that exercise them. I don't think there's anything wrong with the thesis that there's too many federal employees and we need to

trim it and they should be subject to the same standards and insecurities and anxieties and, you know, work week that private employees are subject to. I don't have a problem with that. I think it should be based on what departments are least efficient or performance. But these

Your most talented people leave. The people with the most options leave. It is a degradation, I think, in the quality of the workforce. It's not the way to go about this. Is there a good way to do this, Reid? Elon's using all his playbook, right? What is the playbook that would work here to do what Scott was talking about, if you had to think of it off the top of your head? Well, I mean, the parallel on the commercial side is that you actually also have incentives for keeping the good people to stay.

whether they're stock, bonuses, other kinds of things as ways of doing it. So it's not just the stick, it's the carrot for going along in your long-term commitment to the organization and its mission.

So, you know, this would probably be harder and maybe not within the, you know, kind of presidential remit, but, you know, bonuses like you could imagine, like, you know, the kind of thing you'd say is, hey, if you figure out how to cut 10 percent of your budget, we'll give you a one percent bonus on these kind of things would be the kind of thing that I would do.

you know, potentially look at. But I think it's all, of course, very tricky and difficult. And difficult in a public environment versus a private one, right? Exactly. Because what is results? Because sometimes results, people being less poor, like what are the, it's not a stock. You can see it in the stock market versus something else. Exactly. Right, which is difficult. Is the concept of cutting people a good idea from your perspective? Fundamentally, yes. Just because it's part of what makes

What are the things we should learn from a bunch of the things that we learned from the commercial side? Part of it is refactoring organizations is essential for keeping them healthy.

And so that refactoring is extremely important. I'm quite certain if one looked through the cut of the federal government, one would find like, and I'm sure I'm using this as a hypothetical example. It's like, you know, you've got maybe still a weather balloon department, you know, and it's like, well, we should have one of those. We've got satellites, a bunch of others left. I'm sure we need one of those.

You know, that kind of thing. And so I think that refactoring is good. And I think getting to that refactoring, figuring out how to do it in a good, intelligent ways is one of the things that I'm I think is like we could be hopeful for. Maybe some good will come of this. Right. So a blunt force to it is a way to get it started. Look, I think one of the things we have to pay attention to is I think the default will be massively increase the deficit.

And I think that the problem should be as we should not be. We should be trying to reduce the deficit. I mean, if you look at the actual budget and you want to get a lot of money out of what we're spending, we're spending on debt service is one of the massive line items.

And so increasing that is just, you know, mortgaging our children's future more. I mean, it's already somewhat mortgaged. Let's try to pay off the mortgage versus add to it. Right. In doing this. Well, we'll see what happens. Anyway, one of the other things that that.

Besides doing this, Elon's doing is, and this is an area you obviously came together in with PayPal when he had x.com, but they're partnering with Visa to allow real-time payments through its upcoming X Money service. CEO Linda Iaccarino announced the deal saying it would enable instant funding to an X wallet via Visa Direct. It will also allow users to make peer-to-peer payments online.

Look, they were going to do this. They announced this. But I'm just curious where you would – where this is going, these ideas of the everything app. And, you know, this was a concept that – and not just X has. It's been going around forever, right? I mean, everyone talks about this. But where is the payment space right now? Away from this, I would never –

trust anything Linda Iaccarino's working on with my money, but that's different. Where is it right now, the payments space? It seems like Apple dominates this whole space right now, and PayPal. Well, payments is an area that at least has scale effect, if not sometimes network effects. And so

And one of the things is you have to get to a certain scale before you even have a viable payments network. And so getting that scale really matters. I think that we want to see a bunch of good innovation on this. It is a very kind of common pattern to kind of go, oh, the way I'm going to more verticalize and extend my service is by adding in kind of payments and banking. And we see it in a number of different contexts. I mean, Google has one.

There's a stack of these. And, you know, given my own background, you know, kind of doing the kind of payment space, it's actually one of the areas that I kind of look at because, you

it adds to the flexibility. One of the things I really loved about what we did at PayPal was trying to make every individual able to be a merchant. Now that's now much more true in the whole world, which wasn't true before, which allows individual entrepreneurship inside. I think it's a good area to improve. Lots and lots of people are doing it. Yeah. A platform built on rage, porn, and crypto scams handling your money, what could go wrong?

Look, the FDIC or banks have the FDIC. Twitter has dog memes. I just our financial institution is based on trust that you hit a button that they have figured out a way to get it there safely, that along the way it doesn't get sequestered, that money is not know your customer. There's all sorts of things. FDIC insurance.

I mean, there are so many protocols. Twitter, to me, does not reek of security or safety when it comes to people's money. It might actually, the underlying technology and its ubiquity, it might be a good use case for it. But the entire financial system is a case study in trust. And I don't feel this organization has created a great deal of trust.

All right. So where is the most innovative thing you're seeing in funding right now? I mean, obviously, again, I don't think I've touched, I've dealt with money for a year now. I don't even, I found a 20 in my wallet and I was like, oh, look at that. I use Apple almost entirely for all payments, everywhere I go, obviously, because I have an iPhone. But what is the most innovative thing you're seeing in this area? Well, I haven't, since I've been so focused on

on AI stuff. I haven't actually been looking at this particular area closely. Obviously, there's a whole bunch of things that are being developed within the crypto arena to try to have kind of ledgers and identity and all the rest of that. And the promise of that is to try to create

Something that has, you know, to elaborate on, I think Scott's excellent comments is, you know, how do we have trustless trust, which is a trust in the system that doesn't require trust in a centralized authority, you know, whose, you know, ability to hold that trust may be limited.

And so I think that's one area that I think continues to burb along. Now, none of it has any of the in-depth traction that Apple or other – because the key thing with payments tends to be ease of use and integration into your environment. And so –

You know, I think that part of what Stripe's doing that I think is great is making it very easy to incorporate, you know, kind of Delaware corporations from anywhere or from many areas in the world and then provide a kind of an economic powering structure underneath that to power entrepreneurship.

is, I think, like a really good thing. But like I said, I've been so focused on AI that in this area, someone may call me after this and say, this is really good. And I go, oh, yeah, that's the thing I should have said, but I just didn't know at the time. All right. Scott and Reid, let's go on a quick break. When we come back, we'll talk about RFK Jr.'s contentious confirmation hearing.

Scott and Reid, we're back as we tape this. HHS Secretary nominee Robert F. Kennedy Jr. is in the hot seat for the second day of his confirmation hearing. Things got a little heated on day one with Kennedy rejecting claims that he's anti-vaccine, addressing abortion flip-flops and struggling with the question about Medicare and Medicaid. He's also asked some of his previous controversial comments. Let's listen to exchange with Democratic Senator Michael Bennett, who's a favorite of Scott's and mine.

Did you say that COVID-19 was a genetically engineered bioweapon that targets black and white people but spared Ashkenazi Jews and Chinese people? I didn't say it was deliberately targeted. I just quoted an NIH-funded, an NIH-funded,

published study. Did you say that it targets black and white people but spared Ashkenazi Jews? I quoted a study, Your Honor. I quoted an NIH study that showed that certain races- I'll take that as a yes. I have to move on. I have to move on. Kennedy's own family is also expressing concerns. JFK's daughter, Carolyn Kennedy, sent a letter to senators ahead of the hearing. She called her cousin a predator and accused him of exploiting their family's tragic history. Let's listen.

Bobby continues to grandstand off my father's assassination and that of his own father. It's incomprehensible to me that someone who is willing to exploit their own painful family tragedies for publicity would be put in charge of America's life and death situations. Unlike Bobby, I try not to speak for my father, but I am certain that he and my uncle Bobby, who gave their lives in public service to our country,

and my Uncle Teddy who devoted his long-sent a career to the cause of improving healthcare would be disgusted.

Okay, that was the nice part, just so you know for everybody, which wasn't very nice. You know, all these things seem to be going through Pete Hegseth and everybody else. Any of them you'd like to comment on? The Kennedy one is particularly, it looks like he will probably get through. Same thing with Tulsi Gabbard and et cetera, et cetera, on down the line. What is happening here? I do believe a president should have pretty wide berth in terms of bringing in their own people and...

For all of the weirdness and incompetence parade of some of these nominees, I think this is the most dangerous. You have an individual here who is not only anti-vaccine. If you were to list the greatest innovations in history, I think most thoughtful scientists from both sides of the political spectrum are

vaccines would be near the top of the list. And the fact that we have now politicized it and have an individual with no science background spewing this information who seems to be committed to reducing or creating skepticism around vaccines, and then something that came out yesterday from Senator Warren was that he is being paid to find people to sue Gardasil,

an HPV vaccine that so far has shown to reduce cervical cancer in women by 90%. I don't know if either of you have known anyone with cervical cancer. I have. My God, we have something that can prevent nine out of 10 times this vicious, awful disease. And we have the head-on

of HHS being paid to try and discourage and financially damage that miracle. This guy has no business at HHS. He is probably the most dangerous person

of the nominees in terms of what it could mean long-term without attribution, where in 10 years we wake up and go, "Oh, cervical cancer is back," and reverse engineer it to this individual who is blatantly, repeatedly anti-vaccine. I think this is awful. I made my sons get it immediately when it happened. Reid, how are you looking at these? Obviously, you're making these AI investments in cancer research. This guy will be right there.

at the head of all these organizations that are very related to some of these things. So, look, I think the unfortunate prediction for RFK's probable confirmation is it will probably be measured in thousands of American lives lost.

I think that the question around the fact that he's saying, I'm just following science, no, he's not. It's not only is he anti-vaccine in many, many statements over decades.

Like, he is essentially anti-science. Like, I've heard of meetings where he is meeting with scientists where the scientists try to tell him, he's like, no, no, you're supposed to listen to me. It's like, no, no, the whole point is to listen to good scientists. And I think this will be, you know, if this confirmation goes through, I think the, you know, the senators who vote on it

should track how many thousands of American lives they're willing to spend in this calculus. Because if he implements any of the things that he has articulated and talked about, I think the cost is going to be measured in lives. Have you been surprised by how many tech people have been attracted to him? I'm not going to leave out Nicole Shanahan. She was not a tech person. She's a rich person who married

So I think...

that one of the, you know, the strengths and weaknesses tend to go together. And a lot of tech people tend to think they go, well, I've met him and it was a perfectly reasonable conversation. And so therefore, you know, like, oh, that's just all media stuff. It's like, it's media stuff that he said that he did in other rooms that you weren't in. Yeah.

Right. But there tends to be this really strong, you know, kind of like, no, no, I met him and I hear my own judgment. Yeah, I got the cut of his jib. Actually, Gates did this the other day with Trump, which was disappointing, but go ahead. And I think one of the things that's very important is to have a little bit of, like, look, just how a person talks to you is not necessarily how they actually, in fact, operate in the world. And you have to apply that. And so I've had a number of tech people try to say, okay, he's great. I'm like, no.

Just look, like, for example, in an alternative like hiring context, I prefer references to interviewing. I prefer to have both. But reference checking is much better. And it's like, I've seen the references on RFK. Like, that's part of the reason why it's like, okay, this appointment will be measured in RFK.

thousands of American lives lost. Well, let's leave it at that then. All right, Scott and Reid, one more quick break. We'll be back for predictions. Okay, Scott and Reid, let's hear a prediction. I'm going to go first, actually. I'll be at the White House.

with Tracy Flick, the new press secretary. The White House is welcoming podcasters to the briefing room. Carolyn is running it right now. She has decided that new media, which I agree with, including independent journalists, podcasters, and influencers, will be able to ask questions at briefings. I welcome this.

The Biden administration should have done it, should have had different and not just the traditional people. So I'm super excited to go to the White House and I will be there this year. That's my prediction that Carolyn is going to let me in, Tracy slash whatever the heck she is. Thank you. Any comment? I love that prediction. And by the way, I think it's a good thing to begin to broaden the kind of set of different folks in the briefing room. I just think it's critical if you think like I

I am – this is the office of the president for all Americans to have a diversity of access to it. I mean, so, for example, you have a, you know, kind of a, you know, a Democratic president. You still have Fox News, obviously, in the room. You need to have that breadth of perspective to be representative to the American people. So I love that prediction. Yeah, I know she wants me there. She wants me on that wall. Whoa.

First off, along your prediction, I immediately read that and I sent Jim Acosta an email saying, I need you to start a podcast. I want you back at those press briefings. Yeah. This is the CNN person, for people who don't know, who left when they tried to put him on at midnight. He started his own. He has started a bunch of stuff on his own, and he's a great guy. We love him. He's a great journalist.

So I'm fascinated by what I think is kind of a tectonic shift with DeepSeeker, just the notion that there might be a reasonable facsimile of the best AI models at a much lower price. I think that this, regardless of what we find out, it was probably exaggerated, there is now, I think, a much greater likelihood we might be able to see cheap and cheerful kind of old Navy-like models.

So my prediction is there's an entire layer of companies and pharmaceuticals in the consumer sector that had allocated $100, $200, $500 million to

For CapEx, for the cost or OpEx, for the cost of AI that they were going to have to spend to rent these models to develop new drugs, come up with a better itinerary for your Airbnb stay, that all of a sudden are going to decide, wow, we're going to be able to get all of the great taste of AI with a lot less calories, a lot less OpEx.

And you're going to see a bunch of companies that were setting aside a ton of reserves because they thought this was going to be a lot more expensive than it's going to be recognize, take those reserves back and it'll juice their bottom line. So I'm in the midst of trying to identify this. I don't know if it's Airbnb. I don't know if it's Johnson & Johnson or GlaxoSmithKline, but I think there's an entire layer of companies that

They just got great news that their capex on AI could be 30, 50, 80 percent less than they'd originally anticipated. Well, I think that actually we will see a variety of those. I don't think it's necessarily because of like the deep sea canoes. I think that we were actually already heading towards that. And I think that the part of the agentic future, this is commenting on Scott's.

Part of the agentic future will be actually, in fact, compositions of smaller models that deliver really effective services across a wide range for individuals, for organizations, and so on. So I agree with his prediction, but not necessarily just as a deep-seek response. Okay. Reid, your prediction?

So, you know, I was thinking about this a little bit, and I think that one thing that would be an interesting thing that people wouldn't expect out of this is I think that the AI agents will trend towards creating agents that will be actually massively positive for human mental health. And I think that the actual natural market dynamics will reach in that direction because

As people interact with these agents, and I think I've already seen this from what, you know, my startup Inflection did with Pi that, you know, has helped a bunch of other agents following this to be high EQ, kindness, and other kinds of things. And I think those will actually, in fact, help people feel more heard, more seen, a dynamic of interaction that is more who we aspire to be and will actually help naturally through people just choosing the agents they want to interact with.

to an increase in kind of mental health and well-being. Well, unless, you know, as you know, I interviewed the mother of the character AI, there's the lawsuit there. They can take that turn where they shouldn't be around young people, for example, or they shouldn't, you're talking about adults here, correct?

I'm talking about adults, although I think we also can and I think we should be much more careful about how we engage with children in various ways. That's saying it's one of the important things for the technology industry to get much better at. But I tend to think that the natural dynamics of we will prefer the agents that have this characteristic. Now, I think part of the thing that it can be is

you know, having even agents that are trained to be very good in mental health will also be good. So that they will, so that there are good agents. My only issue is the potential for abuse is so massive and you're seeing it. The lack of care, you know what I mean? Like, why didn't that kid, why didn't that company alert it when the kid said suicide? Like, for stuff like that, like, what, where was, why was that never there? And that's

That's the problem. And their argument is now that it's free speech. It was going to be Section 230, but now it's these agents. Like in that case, should bots have free speech? Like that's the kind of stuff we're going to have to deal with as a society. Is it free speech when it's bot generated? I think at the moment it's definitely not free speech when it's bot generated. Yeah. Well, that's one of the issues we're going to talk about because it will be just off the fly. I prefer kind agents to mean ones, but I suspect –

This is why I love Reed. He's such a sunny character. But I expect really mean agents because people like that, like abusive agents and that kind of stuff.

It may be that they prefer agents to be abusive of other people, but I think most people prefer them to be kind to them. I don't know. I'm going to, as usual, take a darker turn on the thing. We'll see which our future brings us. Elsewhere in the Scott and Kara universe, this week on Profiting Markets, Scott spoke with Robert Armstrong, U.S. financial commentator for the Financial Times. Let's listen.

In the world we lived in last Friday, having a great AI model behind your applications either involved building your own or going to ask OpenAI, can I run my application on top of your brilliantly good AI model? Now maybe this is great for Google, right? Maybe this is great for Microsoft, who were shoveling money on the assumption that they had to build it themselves at great expense.

Interesting. Interesting. Okay, Scott and Reid, that's the show. Thank you for joining us, Reid. And again, your new book is Super Agency, What Could Possibly Go Right With Our AI Future. Thank you, Reid, again. Scott, read us out. Today's show is produced by Laura Naiman, Zoe Marcus, and Taylor Griffin. Ernie and her Todd engineered this episode. Thanks also to Drew Burrows, Moussa Vario, and Dan Chalon. Nishat Kerouan is Fox Media's executive producer of audio. Make sure you subscribe to the show.

wherever you listen to podcasts. Thanks for listening to Pivot from New York Magazine and Vox Media. You can subscribe to the magazine at nymag.com slash pod. We'll be back next week for another breakdown of all things tech and business. Have a great weekend.