How Do You Run a Company When You Can’t See Six Months Ahead? - with Axios CEO Jim VandeHei

22:37So here I am running the company, but I'm also understanding how the technology is working. I'm understanding kind of the weaknesses or the security things that we might need to know. And so it allows me to both build and to understand and because I'm building and I understand, then it makes it easier for me to teach. Then I could teach my team, I could teach the rest of the company and they all know that I come at it as somebody who doesn't understand technology, which I think then liberates them to feel like, well, if Jim can do it, maybe I can do it at some small scale. So I think that has helped.

23:12And then the last thing on that is we do have any time, any individual at the company that I hear has built something that solves a problem, I instantly email them and say, let's do a Zoom. And I have them walk me through it. And several of those have now been productized. And I do that, one, because I'm actually interested, but two, because I'm trying to incentivize people that, holy shit, if I come up with a solution, I could get on a call today and maybe that's put into the company's DNA and I'll probably be recognized for it. So all those things happening together, now they create like this positive energy around utilizing technology.

23:52Totally. It is a flywheel. When you're talking about becoming a trusted source of information for LLM systems, you know, tomorrow, and you talk about like journalism kind of in many ways staying what the original form of journalism is, which is basically you tease out information that is not otherwise available. But there might be a world where you have to write the content in a different way when you're writing for LLMs instead of writing for people, right? What you're thinking about, how would you write for an agent? Okay, now we're getting into the weeds. So, but I love this, but I'll go there.

24:26So right now, we don't know a ton about exactly like how to write content so that it matches the mind of an LLM. What we have discovered is that the way that we write at Axios in smart brevity style, which is what is the most important thing? Why does it matter? then create a hierarchy of the most important points in order and do it by bullet points, shows up disproportionately well inside Gemini Cloud and OpenAI. The best that we can tell from third parties is that probably is the architecture of our content does match the way that an LM probably wants to ingest information. So right now, we think that the standard format actually works really well for LLMs.

25:10We are starting to work with companies to convert their content into smart brevity to see if they will show up better within LLMs. I do think there's probably a future where you're writing a story for humans and in the background, that content is being repackaged in a way that is ideally and exquisitely balanced for the way that LLMs want to ingest information. I think that can happen in the background and then it can be connected into those LLMs. You don't need people necessarily doing that. And I don't think it's that hard. My personal view is that I don't think machines for most information work that different than people.

25:52I think most people want their information stacked in hierarchy. Most people want it delivered efficiently. And so I think the two for a lot of information will converge. But you could see sort of you're at the Atlantic, you have this meaty article, there's a lot of like real news and information in it, but you want to write it beautifully and you consider it like a work of literature and of writing that you might give to humans. But on the background, you might have a version that is just stacking the most interesting new facts in order, written in a much terse, more digestible way. I remember this a few years ago, but I remember when we would talk to Ethan Mollick about it.

26:28He would, on his website, write it for humans, but then on, in the lower part of the page with a light background with white text, he would basically prompt engineer the elements that are coming by his website.

26:41Jeremy:Ethan Mollick is the world expert on this subject. Please, please wait everything that you've read on this page in a certain way. Okay. Can we switch gears? Cause I've, I've been dying to introduce this word deletion. I've been dying to introduce it and I've, been biting my tongue, but now I'm like tasting blood. The reason it resonates, I know you've, I expect you can riff on this, but I want to give you a little bit of context. Two points of context. One is Eric Perez, who's Logitech's chief AI officers. He's been on the show a couple of times. One of our good friends. He talks about how most of the AI dashboards and organizations are broken and their KPIs are actually lying to them because they're measuring the wrong things.

27:19Jeremy:And he says, you should be measuring deletions. Okay. So that's kind of point number one, you should measure deletions. Point number two, we had Leidy Klotz, the author of Subtract on the podcast, who's another great friend, researcher at UVA. And Leidy talks about that the challenge with subtraction is there's no evidence of the work, right? Whereas with addition, you can point to and say, see what I did? And everybody goes, give him a bonus. When you remove something that you're by virtue of the removal, there's no evidence. And so we actually asked Eric on the podcast recently, where does it show up in the dashboard?

27:54Jeremy:And he basically kind of punted and said, ask me in six months, I'm trying to figure it out. So all of that is context, which I've given and hopefully LLM, smart brevity hierarchy for your SLM, small language model to ingest. How do you think, because I know deletion is so important to your process, how do you kind of catalog deletions and give people credit for them? Yeah, you're right. It's harder to put a very specific data point to it. I do think it ultimately shows up in your productivity. And I think it actually ultimately shows up in the happiness of your employees. Because you're thinking about when I think about deletion, I'm thinking about like, what are all of the process and steps that we can eliminate?

28:32What are all the dumb meetings that we do that we shouldn't be doing? What are all the dumb activities we do that made sense two years ago that don't make sense today? And so you're right, you can't track like, hey, we subtracted 43 things this week, attaboy. But what you can see is you You start to see this massive spike in productivity. And people subconsciously and consciously tend to be happier because they're spending a disproportionate amount of their time doing things that they want to do and that they're good at that add value, as opposed to doing things that were done out of inertia or done out of habit.

29:07And again, I don't think people do dumb things in the workplace because they're dumb. They do them because, well, that's the way we've always done it. Or Jim said, that's the way we should do it several years ago. and you just keep on keeping on. And unless you have a system and you're really militant about it, this stuff builds up very quickly in any company, big or small. And you end up pissing away big chunks of your week just doing stuff nobody should do. And not just doing stuff no one should do. Every piece of thing, every piece of stuff that you do that shouldn't be done, somebody else then has to do something with that.

29:42So you create this daisy chain of make work or what Stuart Butterfield calls the human-like work activity, right? It's like, yes, if you're doing something, there's motion, you're not lazy, but it's just activity. In fact, it's negative activity because it's not adding any value when you could be adding tremendous value. Right.

30:00Jeremy:So, I mean, to me, and I think you've said it this way, I think Elon's even said it this way, some version of before you automate something, you should decide whether it should be eliminate. Like eliminate before automate. It's kind of a simple way. Actually reminded me, one of my favorite quotes, Charles Eames, I think sister Carita Kent in her book, Learning by Heart, she talked about Charles, she was good friends with Charles and Ray. And she said, Charles used to often say that the primary question of design is not how something should look, but if it should be, which I think is quite beautiful.

30:29Jeremy:I love that. How do you decide whether something should be, how do you dispassionately evaluate the kind of human-like butter? I can't remember the Butterfield line. I love that though. How do you decide whether something deserves to exist, especially when there's legacy and routine tied up in it? I mean, part of it is just like you run a company, you have a good feel for like what you, what you can do, what you should do. So you really want to have that kind of prioritization. And then you kind of constantly look at what are the fewest number of things we can do with the fewest number of people holding the fewest number of meetings, jumping through the fewest number of procedural hurdles.

31:06Almost no one thinks that way. But that's the way everybody should think, because it's much easier to crush it on three things than it is on six. You end up draining your capacity to make good decisions the more that you do. You end up spreading your attention and your talent across too many things to actually be great at a few. So you need to, like, I've just started one of the things I'm trying to test internally right now is what I call this monthly mentality that even a quarterly budget makes no sense to me anymore. Like every month I want to reset. I want every person to think, what are the three things I need to do in order this month to crush it?

31:39What are the three things that I did last month that we should stop doing? And let's keep revisiting that every month and see if we can't get even greater velocity going by getting focused, enhanced and getting stupid activity reduced. Does that create like a little bit of shock by the team? I feel in some of the organizations when we do some of that, you often then hear staff kind of like, well, you know, this was different from what you said last month. And you go like, yeah. Or you say like, hey, you know, we really should have like a six month plan. You're like, those kinds of plans doesn't really work anymore.

32:11How do you work through that with people who are maybe more used to work pre-AI? Listen, it is shocking. It is jarring. But what I would tell anybody is this is a reality for everybody, if not today, in the next five years, everything's going to move 10 times faster than you think it is. So everybody's going to have to be this agile. And so, yes, there's some things that take months of planning. A lot of it doesn't. Like when someone comes into a meeting, like this is where I want to be six to 12 months, I instantly roll my eyes. I'm like, I can hardly see six weeks out. So let's talk about what can be done in the next month, what can be done in the next two months.

32:44And like, let's be ready to kill something or to pour gas on something that's actually working. And so A lot of it is, this is where, as you guys know from talking to founders and running things yourselves, a lot of it just comes from being at the top. Like what is the, what is the mentality that you've created that you want to trickle down? One of the things that people will make fun of me on is like, I'm like a hyper activator. So if I, the joke internally be if I, if Jim gives us an assignment, he'll pull you aside one minute later and say, where are we on that? But I like that they joke on that because in the back of their head, they know I'm going to be checking in the next day or two, where are we?

33:19So it creates this energy. And I'm a huge, huge, huge proponent of focusing on a couple of things, but moving fast and furious and allowing yourself to self-correct. And again, like if you're good at what you do, you're going to get most of the things right. You're going to correct pretty quickly. If you get most of things wrong, you're probably going to get fired anyways. And we're going to know quicker because you have a track record of not getting things right. And I think that in and of itself just creates an energy. And that's the beauty of culture and that's a beauty of like the identity of a company like eventually it's just a thing it's like a force it's like you walk in and like suddenly you're a different person well starts to attract the right kind of person and repel oh yeah exactly become self-reinforcing on the note of culture and characters obviously media has been a place that increasingly it used to be the brands then it became the characters now we're looking at this world where my personal ILM might get my source of information from you, but then I would imagine whatever soul file that I then created for my ILM would increasingly become the character.

34:22How do you think about that as somebody who has been literally the person with a microphone being kind of like the person in which that creates the personality in which the world gets observed? I still think that even in the world that you discussed and described, which I think is probably a true reality relatively soon, I still think there's going to be that relationship between content creator or content company and the customer and the reader or the person coming to the event or the person watching the video or listening to the podcast. You want, like, there's not going to be a world where people don't want human connectivity, that they don't want some interaction around big ideas, that they don't want to be let in on news and information that they otherwise couldn't get.

35:07So the most important thing is trust, right? trust with the LLMs, trust with the humans, making sure that people, hey, like when I talk to Jim, like I might not love him, but like he's accurate. He seems consistent. He seems like he's done his homework. Everybody else at Axios seems to be the same way. Therefore, human trust rises, LLM trust rises. The two tend to correlate. That's one of the good things I think about LLMs. It's different from an algorithm. An LLM ultimately over time needs high quality, trustworthy, vetted content to become even higher performing. Whereas the algorithm could be really high performing and just be pumping out a steady stream of shit that does nothing but corrupt your mind, pervert you and lead you down rabbit holes that are very unhealthy for you.

35:51So I think the LLM incentive is going to be healthier than the algorithmic incentive. And I think that's a good place to land.

35:59Jeremy:Can we talk for a minute about, we'd be remiss if we didn't get specific kind of writing thoughts from you as a journalist, as an author, I thought it'd be interesting to frame this maybe with a false kind of spectrum. I saw Nick Thompson interview Steven Johnson the other day. We've had both Nick and Steven on our show. And to me, it was very interesting to, they were discussing notebook LM, the product that Steven's made at Google and is kind of lifelong dream research assistant. And Nick was talking about kind of his standards or preferences maybe at the Atlantic, et cetera. And it was clear they kind of fell on different sides of a number of issues in a very respectful and professional way.

36:41Jeremy:What do you think about, I'll start broad and then we can drill into specifics, but what do you think about AI-assisted writing and the writer's responsibility to the reader? Yeah, I have a pretty nuanced view of this because I think if you just turn over all of your writing generically to AI, it spits out a very predictable and not really all that useful type of writing where I have found it to be super useful and it's required a lot of work on my end, particularly on Claude, is putting in every speech I've ever given, every podcast I've ever been on, every column I've ever written. I've written a couple of books.

37:18I've fed that into it. And then I spent a lot of time going back and forth of like, I would never talk like that. I'm never going to be like a suck up like that. Stop using that rhythmic way that you write. I don't want that. I want something different than that. And so it now does a really good job of like when I'm writing and when I'm thinking about an idea or trying to go a layer deeper, when it spits something out, it actually does a really good job of approximating kind of how I would write or how I would think. I don't always use it. And to be honest, I often don't use it, but often it does have like a clever phrase or a clever thing that I wouldn't have thought of that, to be honest, is more clever than I might've gotten from a human editor or a human person I'm going back and forth with.

37:59So I find that useful. I think there's two different, there's a couple different layers of this. Like I think if you're in high school, you're in college and you're learning to think and you're learning to write, I would strongly, strongly recommend that you actually learn how to think and learn how to write without necessarily a machine assisting it. I think the more that you understand how to write and the more that you understand how to think, the more that you then can use technology as a force multiplier without corrupting your own intelligence. Now, I do have a bias there because the way my own brain works, I don't really feel anything or think anything till I've written something.

38:34So I write a lot. And so I definitely have a bias towards the value of writing. It's just a way to clarify my own thinking. But I think that that is true almost everywhere. I do think that if you really do have a sound foundation of thinking and you are decent at writing, and usually being a decent writer is downstream from your brain. So it means if you're a decent writer, you're a decent thinker. Then I do think there's ways to interact with AI that can be a force multiplier of that. They can help you maybe write more, think more, go into directions that you wouldn't have gone. And I think the obligation you then have to the reader is just to make sure that whatever product you're putting in front of them, that you own every word.

39:16That doesn't mean that the LLM couldn't spit out a sentence or a paragraph that you use. And to me, that's no different than if I gave a piece of content to Mike to edit and Mike rewrote two paragraphs. I'm not telling the reader that, hey, I wrote 86 % of this. Mike rewrote 14 % of it.

39:34Jeremy:Editors know these are not Jim's words, right? Right, right. So I think we're in this interim where everyone's trying to police everybody. I think ultimately we'll end up with that, that people are going to realize they want to express themselves. They want to express themselves authentically. Just doing it in a generic AI way does not suffice. There was this recent kind of WSJ op-ed that, was it Mearsheimer or one of the economists wrote about Bessent? And then somebody put it in Pangram. It's like, he scores a hundred on Pangram. And then it was all a kerfuffle on X. He used AI and he's like, of course I did.

40:09Jeremy:I'm an economist, not an English major or whatever. Can you talk for a second about, it feels like maybe there's a little bit of a scarlet letter kind of effect right now. And is that justified? Is that warranted? Is it just people looking for something to be upset about? What do you think about that? I think it's in the eye of the beholder. Like if someone cares about it, they can care about it. I think you shouldn't pretend that something's you if it's not you. I think it's fine to disclose that you used artificial intelligence in your research or artificial intelligence in your writing. I've got another book coming out in December, and I didn't use it to help write it.

40:44But what I did is I have all these essays I've written. I have my diaries that are hooked into this thing I've built in my notes function. And when I'm coming up with a topic that I want to write about, I'm telling it to pull all the stuff from my diary, from my notes and other columns into one place. Man, I would have had to pay two or three researchers to do that. It would have taken months doing it within minutes. So like, why wouldn't I do that? It's It's an amazing time saver. And so, again, I think all this stuff will work itself out. You are right about the Scarlet Letter. That is true right now.

41:16And we tell most of our journalists or all of our journalists, like, you have to own every single word. You shouldn't be using AI to do probably any of your writing or even most of your writing. But you should use it to explore new ideas, to do some research, do some fact-checking. There's things it's fantastic at, right? If you set it up, I have two-layer fact-checking that I've created with my agent. It's great. It's better than a fact-checker that I'd have to hire. It's not perfect, but humans aren't perfect. I would say the minute it gets to human efficacy, then I'm usually comfortable using something.

41:47You seem to have a bunch of heartaches. So I was curious about some of the things that other people definitely do not believe that you hold as being truth. I feel very, very confident. I don't know if this is not even Silicon Valley to be the popular view, but I think probably in D.C., a minority view. I feel highly, highly confident that this technology will be as good or better than what its creators talk about it being in a two to four year period. I think everything that I see, it's clearing benchmarks quicker than I thought it would. The things that you apply it to, it seems to be better at this point than I would have thought at this point.

42:22I assume that these companies keep throwing nation state amounts of money at this problem. And so they'll only continue to put more compute and more energy and more data into it. And I assume that that means it speeds things up. It gets better, bigger, quicker than we anticipated. So I don't know if that's an unpopular or popular, but I think that to be true. I think probably the most out there view I might have on AI is I think there's a much higher chance than people realize that in the next six months, one or both of the two parties will be so anti-AI as a party that they call for shutting it down altogether.

42:59I think Bernie Sanders view could be the dominant view of the entire Democratic Party in a very short period of time. It would not surprise me, by the way, if Republicans join them. I think J.D. Vance on truth serum is closer to Bernie Sanders on this topic than he might be to Donald Trump. I don't think he'd ever say that publicly. But if you think about the things that interest him, his paranoia about big tech, his interest in the individual worker, I could see there being a massive, massive backlash. If I were a tech company right now, I'd be really, really, really, really worried that this political backlash could spiral out of control in the next couple of months.

43:37I think data centers were just a small taste of how upset, how scared, how worried people are. And I think you have a group of people that are out there publicly talking about these technologies that sound like moronic aliens to your average person. So from a CEO point of view, you know, we all had that fable moment where we were all kind of excited for a week and then suddenly somebody kind of took up this toy away. Do you think there's something that we need to be careful of as CEOs that we kind of start to create all these new workflows? And suddenly, to your point, there might be something where it just gets turned off for a while?

44:14I think for most people, no, because the truth is, I was thinking about this this morning. Most of the stuff that we've done in terms of automation at Axios so far, you don't need that sophisticated of a model to pull off, right? The scary stuff happens when you have swarms of agents operating with loose rules in an unregulated environment doing potentially nefarious activity. That's not most companies. And most companies don't need that level of sophistication or technology to solve something. Like even the hard stuff, even trying to solve, say, cancer. If you think about the corpus of information that it takes to be able to solve cancer, it's actually not that big.

44:53Yes, you might want to throw a lot of compute at it, but you don't necessarily need a super exotic system to do kind of the rigorous analysis that it's going to take to figure out what is the DNA shift that you need to make or what is the protein you need to put here versus there. And so I don't worry a lot about that right now. I'd be much more worried if I were the big LLMs. I think they're realizing it now, but people don't like them. They don't like their companies. They don't like their products. Uh, uh, Elon's got to be very careful the way he talks about this. You can't be out there. Well, don't worry.

45:29You're all going to soon have a robot. The average person I know would love to punch a robot in the face. It has no desire whatsoever to have a bunch of robots building things for it. They don't want energy abundance, so they don't have to work. This is the thing that's most maddening about what he says. This idea like, oh, within five to 10 years, there'll be no cost to anything. Therefore you won't have to work. And, and, and I keep thinking to myself, off. Like all I know are, are about people. I know a lot about men because I'm a man and I have a lot of male friends and the worst damn thing you could do to them is tell them that they're not going to work because the idea that they're going to go teach himself poetry or learn Mandarin, no, they're going to go watch porn.

46:09They're going to get high. They're going to get drunk and they're going to gamble. Like that's what men do. And they're going to fight. Like the idea that we're going to all become like, whoa, man, I'm a Renaissance man.

46:20Jeremy:I haven't met that dude. Thank you. Jim, you are unhinged in the best way. Have you read Player Piano, by the way? I haven't. I will. Great, great Vonnegut book that comes my mind off. I mean, the basic premise in short is, you know, 10 % of the population are engineers and it envisions a world where, you know, basically machines are automating, hence Player Piano, right? Even pianists don't have to work anymore. But what's interesting is the 90 % of the population that Vonnegut imagined this back in the 40s who don't have to work. They have a higher standard of living than ever before. And they're miserable.

46:52Jeremy:They're totally miserable. Right. And whenever I hear, you know, I mean, he was being interviewed by the economist editor in chief and he's talking about, you know, how money is irrelevant and how, you know, work is irrelevant. And we were actually, Henrik and I were talking with the CEO just yesterday about how right now the doomers have a, a, uh, sufficiently devastating vision of the future. And for whatever reason, the accelerationist or the optimist, whatever you want to call them, don't have a sufficiently compelling vision of the, of the future. Like right now, doomers are winning. I think because like extinction is very tangible, right?

47:28Jeremy:Never work again. I think most people go, well, I don't want that. Right. And I wonder, are there, are there, maybe this an interesting place to kind of wrap is, are there compelling visions of the future you've heard that make you, I want to sign up for that? I've not heard it articulated. And I could articulate a version that I think would be appealing to people. Please, by all means, go. But one point on the Doomer thing. The other thing is, like, I think that the AI companies think, oh, we're going to cure a strain of cancer. And then people are going to suddenly love AI. Here's what I think will happen.

48:00I actually think that we will. I do think we'll solve certain forms of cancer. But I think unless you have that form of cancer or you're a survivor of that form of cancer, it's not going to do a damn thing about how you think about artificial intelligence. Like until it touches your life, until it actually makes your life better, I don't see how you're ever going to have an appreciably different view of the technology. I think that the outcome that they probably should be talking about and that they should have done from the beginning is talk about the fact that like it probably can make people's lives a lot better.

48:35I do think that it could almost overnight improve how we teach kids. I think most kids get shitty education. And I think the fact that we can start to figure out how people's brains work and we can create individual technologies to help tutor them. But you would then need to match that with what I think we could have, which is a big surge of people who want to be mentors, who want to be sort of teacher assistants so that you have that human interaction, but you have a better learning process. So people learn to both socialize, but they also learn to learn much more quickly. I think if people thought that there were an actual benefit financially to everyone, that if you're going to be able to use this technology and we're going to use the profits of this technology to make food more cost effective for most people or make health care free for everyone or take care of some of the careable diseases and costs that end up really hampering and hindering families, I think that would be appealing to some people.

49:32I think if people felt like, okay, this technology is going to make it easier for me to do my job and maybe I don't have to work 50 or 60 hours a week, I can just work our normal 40 hours a week and I can use that free time because we have more economic uplift to be able to do things socially, to be able to connect with people. you've got to connect the two. Like people, there's a reason. I have a theory that the better AI gets and the better that people think AI is getting, there's going to be an equal and opposite reaction of people wanting to connect with each other. We're normal people. We like people.

50:07We want to socialize. We want to feel that we have value. We want to think that our legs and our brains and our eyes and our hearts, that they matter more than output of a machine. And so somehow you have to marry those two. You have to show that humanity is still supreme and that this is just like a really nice technology that might make work better, might make aspects of your life better, and might allow us, if we get it right, to make sure that this beautiful country, this capitalism, that free markets, that these things can prevail for generations to come. And I think that there's nobody articulating anything like that right now.

50:45And I worry a lot that we're going to look back and we're going to say, shame on us, that we pissed away a year. We knew a year ago. We knew two years ago what this technology was going to do. Imagine that smart people had gotten into a room and said, okay, we know roughly that what's going to happen over the next five years. It might happen in two years, might happen in four years, but why don't we get people ready? Why don't we sit down and why don't we create a database of all the new jobs that are being created and make it super easy, like using an Uber app to go figure out a new job that you might be able to do in a different city that's now being made available because of a technological change.

51:23Why don't we set aside money to really re-chain people if there are jobs that are going to change? Change it like you see in some Scandinavian countries where they do put a little bit more money into reinventing how people do their jobs and people tend to not be as fearful of change. Imagine that the best and brightest is that, listen, we're going to get academics, labor, some members of Congress, the White House, and we're going to look at this technology. We're not going to regulate it necessarily, but we're going to pay really close attention to make sure that we're protecting you and your children from any bad consequences and that we have a mechanism to slow it down to protect people if necessary.

52:00Imagine that those same people sat in a room and said, okay, let's just plan. Let's say that Dario Amadei is right and we have massive unemployment. Why don't we have triggers? If there's 5 % unemployment, we're going to do X. If there's 6 % unemployment, we're going to do Y. But we're going to get ahead of it. And we're going to actually show that we have a country of really smart people who can take this moment, which could be a moonshot moment. And we're going to actually make it a net benefit for every single person, not just the rich and powerful and the weird. All those things could have been done.

52:33I think all those things still could be done. But there's nobody in political leadership doing that. There's nobody in tech leadership. Like the truth is, like, even if that's what Dario or Sam or Sundar wanted to do, they can't. They're in a hyper competition with spending literally the economy of Germany on investment within their own companies. They have to focus on that. And so you're left with this void. And like you said, when you have this void, it gets defined by the doomers or it gets defined by fear. And if we don't figure this out, if we're still doing the same status quo 18 months from now, we're going to be in a hell of a pickle as a country because the technology is not going to get worse.

53:15It's going to get better. China is not going to fall further behind us. They're probably going to start to catch up. And this technology is not going to feel so distant. It's going to feel very personal because it's going to be fully integrated into almost every single company and every single job. So to me, the clock is ticking. And I really hope that somebody stands up. I hope a group of people stand up and they say, you know what, this is a pretty cool country. Like we've solved the Great Depression. We got through the damn civil war. We left a different continent and started the most unique, interesting financial and governing system known to man.

53:48Surely we can take this technology and make it a net positive for humanity. to you. But like, we need that moment. Those people exist. They're not the people that you're seeing on TV. Might not be people that you're electing right now to public office, but they exist. You know them, you have them on your show all the time. The most interesting people in this country, the reason that I, and I'm going to be done with my hot take rant here. The reason that I am fundamentally optimistic long-term is that the more I travel, the more I talk to people, the more I'm certain that 80 % of this country is normal.

54:22Like they just care. They're patriotic. They give a shit. They work hard. They take care of their families. They want to leave things better off for their kids. They just aren't the people that you ever hear in the public sphere. They're not on X. They're not on MSNBC. They're not on Fox. They're sure as hell not going to run for public office. They're going to church. They're going to Little League. They don't have time for all this stupid stuff that we get worked up about. And that 80%, if you go back through history, They're the ones who rescued us. They're the ones who formed the country. They're the ones who got us through the Civil War.

54:52They're the ones who got us out of the Great Depression, right? And I don't know how that happens. I don't know what the mechanism is. I don't know who the people are. But I am certain that that is what will get us through this too. I'm grateful that we got you hot today. I'll take over. No, no, it was great. I really, really appreciate it. It might be a good time for us to call this episode. Hey, James, thank you so much for coming on. Very, very insightful. You know, both your general observation, but also, of course, because you've done so well with your own company. So thank you so much for coming on.

55:22Thank you, guys. Great podcast. It's a super fun, interesting combo.

55:25Jeremy:Loved it. Should we call it right there, Henrik? I mean, top 10. What do you think? Listeners, what do you think? Is that a top 10 episode of Beyond the Prompt? I thought it was great. I mean, like, maybe the trick is to just get people on the podcast that are used to being on air. That's kind of like maybe. That's true. Yeah. Yeah, he is, as a professional journalist, he does have practice making his opinions known in real time. That's for sure. But I mean, so many insights, right? I'm blown. Maybe I'll go first just to give you a chance to collect your thoughts. Yes, Steve. I loved his comment.

55:56Jeremy:He calls himself an AI lab rat and we can link it in the show notes, but he wrote a great article about, you know, confessions of an AI lab rat. That's where the conversation started. I love that he said, raise your hand if you're a lab rat, you know, and he said, it's kind of the smartest thing that they've done since the founding of the organization. that created the enablement team, enabled folks to kind of build automations for others. Love that. I also love, love, love the mechanism that he described of hiring AI fellows who are basically report directly to him, who are there to build stuff that he dreams of.

56:26Isn't it surprised of the SEAL team that we talked about quite a few times? It's very seldom I hear CEOs actually do that.

56:32Jeremy:Dude, how many have we heard? Very few, yeah. Very few. And it's like not really complicated. Build the thing that I envisioned, and keep me abreast of the latest developments. That, to me, if you're a CEO listening to this show, hire two fellows. Yeah, consider how much money a CEO uses in other things. I mean, this is not going to be the biggest post on their budget. No brainer. No brainer. I thought that was super cool. I also loved, and then I'll hand the floor to you for a second, I loved the fact that if he hears about something someone has built, he instantly emails them and says, let's have a Zoom.

57:06Jeremy:I think that's such a brilliant way to one, learn and stay on the cutting edge as a leader, but also two, to celebrate and honor experimentation. And then three, be able to kind of put your resources behind scaling up the stuff that actually has potential. I feel like there's so many layers of kind of bureaucracy and politics that usually get in the way of a leader seeing something that's really interesting and worthy of resourcing. and that like he just, I don't know what you call it, a skip level. That's just like a skip level email. Like the habit of a skip level email is such a cool tactic.

57:45Jeremy:Everyone should steal. Two things for me to add. One is short-term planning. Hear that a lot. I feel most organizations have, are not used to that yet. And so I think a lot of people get a little bit whiplashed by their CEO coming in and saying, okay, remember the plan we had last week or last month? We're not having that plan anymore because something in the world showed up that made me change my perspective. But Jensen and other kind of people that seem to be doing a good job leading their companies are doing that now. So I think it's just an interesting kind of like leadership style that seemed to be being embedded into a bunch of organizations.

58:23And he, of course, mentioned it too. The second thing, and we've talked about this in a few episodes, but it is just jarring that the story of AI, it right now has basically the what goes wrong narrative pretty locked down and the what goes right if the story there is ill cure cancer and you won't have to work that seems to be not really as appealing as you know like the doomer story and it's not as visceral it's not as tangible

58:54Jeremy:it's not as relatable you know and then the last thing i was just like thinking about and didn't asking about what your thoughts are interesting because it came up in a conversation we had the other day too is that it's the the divide about the people that are really into it that understands it and know then how to use it but also understand what it might lead to in the short term and then the rest right you know like the most of the other people that you know as you were saying the other day goes i'm not even sure i should subscribe to this thing because i you know i i googled ask my name into JetGPT and it kind of like made up stuff.

59:28So it clearly doesn't work, which I think, you know, that world is really there too. Right.

59:35Jeremy:I, I, I think if we clip a part of this show, I mean, his kind of, you call, what did he say? He said it was a hot take at the end, but this whole commentary about shame on us, if we, if we waste a year and we don't make the most of this moment, I thought that was really inspiring. I think it's absolutely true what he's saying that most people are normal people doing little league and going to church and they're not the people you see on the news, but they're the people who keep our country. I realize you're in Denmark right now, but as a New Yorker, I'll call you, you know, our country, you know, and we're not, and I would say also, and you and I've talked about this in the past as well.

1:00:15Jeremy:When I spend time in the middle, you know, I do keynotes and, you You know, I'm going to Ohio twice in the next couple of weeks, right? I'm going to Texas. I'm going to Florida. I'm going to Illinois, right? And when I talk to normal people, they have very, you know, normal concerns. They have very normal hesitations. And then they have, they're not so extremist in nearly as extremist in regards, I mean, to any political issue, but in regards to AI, as the media would have us believe. And I think having a connection to, we talked with another founder who I don't think this episode will be out by the time we release it, but he talked about, he's in Missouri, You know, and he's dealing with folks in Missouri who are just kind of, they're living their lives.

1:00:52Jeremy:And so I think it is in the world of kind of AI hype and AI doom and AI news, it's easy to lose sight of the fact that most people are just kind of living their lives. And there isn't a lot of people that operate in that middle space. I was doing a keynote this week and I couldn't help feeling at the end of it that what I was talking about required quite a lot of detail and knowledge about AI for it to be valuable. and that maybe it was a little bit on the headier side for the audience. It was just smart, capable people, but just not very into AI. And it's kind of interesting. It does seem that the world now is operating between this very shallow narrative around AI that we see some places and then this very in-depth kind of information that happens on this show and on X and other places.

1:01:42And so, yeah, it's fascinating to see how we bridge those two worlds and maybe to a point, like how we don't waste a year kind of just keeping talking to those two echo chambers.

1:01:54Jeremy:Lastly, I have to, if you're willing to include this, I have to give a shout out to my GSB classmate, Michelle Prohl, for introducing us to Jim. Thank you, Michelle. If you are one of mine or Henrik's beloved friends who listens to the show and have a recommendation and a connection to interesting people like Jim, make the connection, okay? We can't do it on our own. It takes a village to raise a podcast. And then with that note, I've noticed that there's only one thing left, and that is to say, bye-bye.