Skip to content

Podcast

Debt Spiral or NEW Golden Age? Super Bowl Insider Trading, Booming Token Budgets, Ferrari's New EV

All-In with Chamath, Jason, Sacks & Friedberg

Source ↗ ← All highlights
  • AI Amplifies Work, Not Replaces It
    • AI tools intensify work: employees using AI take on broader tasks and work more hours while feeling more productive.
    • Early adopters who can structure work with AI agents gain outsized leverage and workplace advantage. Transcript: Jason Calacanis It was a big week for AI. New study published on Monday, February 9th in the HBR Harvard Business Review, suggesting that AI tools intensify work, but do not reduce it. Two UC Berkeley researchers spent eight months embedded at a 200 person tech company. So this is one company’s experience. What they found, employees who use AI worked at a faster pace, took a broader scope of tasks, and extended work into more hours of the day. Workers reported feeling more productive, but they also felt a little more stress and burnout. Saks, your hot take here, your quick take on this study. Obviously, it’s just one company, but it does track, I think, some of my experiences. All right. David Sacks Well, a few points here. Number one, as you may recall on the prediction show for this year, my most contrarian belief is that AI would increase demand for knowledge workers, not put them out of business. And I think you see in this UC Berkeley study, the reason why that might be the case is because the employees who use these tools, like you said, they work faster, they took on a broader Scope of tasks, they actually ended up working more hours in the day. So they did more work, not less, and even more effort rather than less, not because they were required to, but just because they were more motivated. And I think they were more motivated because their work was getting up-leveled, right? They’re kind of able to offload more menial tasks to AI and it made their work more purposeful and meaningful. So I think we’re kind of moving from what some people, I think maybe Jensen has called task-based jobs to purpose-based jobs. And I think a key skill of employees is going to be the ability to structure work for themselves and their AI agents. And the employees who can do that are going to be far more productive than those who can’t. That kind of brings me to point number two, which is that I think there’s a tremendous opportunity this year for employees who are early adopters of these tools or so-called AI natives To demonstrate their value to their employers. They’re going to be able to get a lot more done. They’re going to appear to have superpowers. (Time 0:00:20)
  • Adopt AI Bottom-Up Fast
    • Bring consumer AI tools into your workplace from the bottom up rather than wait for slow top-down RFP projects.
    • Learn to build, manage, and educate agents to become dramatically more productive than peers. Transcript: David Sacks They’re going to appear to have superpowers. They’re going to be the people in meetings who can take an assignment that would have taken days before and get it done in two hours. Whether it’s a presentation or a spreadsheet, people are going to be shocked at how quickly they can get these things done because they’re going to be facile at working with AI. So I think there’s a big opportunity there. And there was an article that went viral this week by Matt Schumer called Something Big is Happening, where he talked about this career opportunity that’s going to be available to kind Of AI early adopters. And I think that brings me to my third point, which is I think that you’re going to see massive enterprise adoption of AI, not just chatbots, but agents this year. But I think it’s going to be driven by the bottom up. It’s going to be driven by these early adopter employees coming in to their workplaces, bringing in these kind of consumerized AI tools, start using them at work, as opposed to top-down Initiatives. I think there’s a lot of top-down company transformation initiatives that are happening in large enterprises where the CEO has tasked a team with figuring out how to use AI, how to transform Their business with AI. Those initiatives are going to take months. They’re going to be studying what tools they should use. They’re going to be doing RFPs. And I think it’s ultimately going to be very slow. And while those things are trudging along, I think there’s going to be these early adopter employees who just make the transformation a fait accompli by, again, bringing these tools Into the workplace from the bottom up. (Time 0:02:34)
  • On-Prem Returns For Confidentiality
    • On-premise infrastructure may return because enterprises want control over confidential data used by AI agents.
    • Using public AI endpoints risks leaking proprietary prompts, responses, and agent traces to model providers. Transcript: Chamath Palihapitiya What are your thoughts, Jamal? I think there are two open questions that I find really interesting right now. The first question is, I tweeted it this morning, but is on-prem the new cloud? Is weird to think that that could even be possible. But we’ve spent since 2008 migrating everything to cloud because there were these economies of scale. And it created better margin and lower OPEX and lower CAPEX because you could essentially share infrastructure with other companies. And that’s how AWS and GCP have built such gargantuan businesses. The counterpoint to that, though, is that in the AI revolution, companies, I suspect, will be fighting for their lives. And I think it’s very much unclear whether it makes sense for a company to allow the natural leakage of their edge and their confidential and proprietary information out into the wild Versus the control that they would get if they ran on-prem. That’s a really important question. What do I mean by all that? Once you use these tools, it is very difficult for a company to be able to control how their data is used subsequently thereafter. Meaning, if I gave you, Jason, a PDF of some really important strategy document or a PowerPoint deck or a really critical model, and you’re interrogating it with one of these models, If you’re just using ChatGPT, the mainline instance of it, you’re leaking all of that prompt and response metadata back to ChatGPT, back to Gemini, back to Cloud. And there’s nothing a company can do about that. If you’re using a set of agents to act on all that information, all those agent traces are going back to these model builders. That may or may not be a problem for some, but I suspect it is a deep problem for others and they just haven’t uncovered it yet. When they realize that that is a problem, the enterprise will have to decide, do I just give up and keep running all of this stuff in the cloud in a shared experience or do I bear the incremental Cost of running this stuff in a more coordinated manner that I control on-prem? And that would be a crazy shift just to completely go back to where we were 20 and 30 years ago. That’s a non-so thing that may happen. So that’s number one. And then number two, I also tweeted this, there was this really interesting ruling around what happens inside these cloud environments, which was a judge saying there is no attorney-client Privilege and confirming that once you start to use those tools, all of that stuff is complete public domain material. If you put these two things together, it creates a very interesting set of questions for enterprises. You will need AI to survive. But if you use the tools as they exist today at a public endpoint, you will give up all control, all security, all confidentiality that you today have, and the ability to follow through And control what your employees do with it. The only solution is to have the pendulum swing all the way back and have private provision networks, which increases cost. But then if you save a bunch of money because of AI, maybe it all balances out. That is the big question that I’m wrestling with right now. (Time 0:06:31)
  • Firm Uses Replicant Agents At Scale
    • Jason describes deploying multiple AI replicants with Notion, Slack, and Google accounts to automate firm tasks.
    • He reports 20% of an investment team’s work now handled by agents and an Ultron meta-agent supervising them. Transcript: Jason Calacanis Am. Have now seen that every week, 5% to 10% of the work we’re doing inside of our venture firm is being moved over to OpenClaw. We call them replicants. You can think of them as personas. So we now have three or four of these. We give them a Notion account, a Slack account, and we give them a Google Docs account. They have their own email. And I think all of this technology was here all along. It was really, or maybe for the last six months, let’s say, really good models out there. But no company would give the keys to the kingdom to allow these agents to actually act on your behalf. Why? Because they don’t want to be responsible if it ships your Bitcoin keys or your passwords to somebody else so in order to use these you have to trust them and if you trust them and then you Are monitoring them the results are unbelievable we uh have also to your point jamath fired up max studios we have kimmy on them we are moving all of the work onto these And then they’ll Use Kimmy for most of their easy jobs, which is free. Then they will use Claude 4.6 Opus to orchestrate things. We also, now that we have four of them, Friedberg, we’ve created OpenClaw Ultron, which is one meta replicant that is managing the other four. And it checks their work, it talks to them all day long about what they’re doing, and then summarizes it. And we’re building skills into each one of these. (Time 0:11:24)
  • Set Token Budgets For AI Agents
    • Give agents limited, monitored token budgets and monitor costs because API spend can scale rapidly per agent.
    • Balance token allocations so AI-driven productivity doesn’t bankrupt your org. Transcript: Jason Calacanis A lot because he’s got Slack bot, Claude’s got co-work, but none of them have the keys to the kingdom. So what I’m doing is I’m upgrading to the enterprise version of Slack, Chamath. You’re, I think, probably your number two investment in your career. What an amazing investment that was. It’s number four. Number four. Okay, listen, keep grinding. Top five investment for you i’m upgrading to the highest level and i’m ingesting every single slack message and then i’m upgrading and giving the api key for every single email in our Organization to ultron they will know everything going on in the organization it is mind-blowing how fast this is going and then finally just a plug i’m investing in 10 startups in open Claw space, 125k each to come to the accelerator. If you’re doing work on this open claw at launch.co, email me what you’re doing, because we want to invest in at least 10 or 20 of these companies right now. Chamath Palihapitiya This is the 100% focus on our firm. It is insane. When do you guys think enterprises have a huge freak out around all of this and say, wow, we’re leaking all of our most important information out into the wild? But Sachs, to your point, the industrious person trying to get ahead, (Time 0:13:44)
  • Prediction Markets Scale, Invite Insider Risk
    • Prediction markets hit scale at the Super Bowl, raising fresh concerns about insider information and market manipulation.
    • Large players with private edges can dominate liquidity and extract profits from casual bettors. Transcript: Jason Calacanis About prediction markets, gentlemen. They hit critical mass this past weekend at the Super Bowl. More than a billion bet on Cal sheet, 700 million on Polymarket, almost $2 billion in wagering. The media has been obsessing a bit about market manipulation, insider trading, and all these issues that are totally valid to discuss around prediction markets, which are something New in the world, at least at this scale. Two specific examples from the halftime show. A day-old anonymous Polymarket account correctly predicted 17 out of 20 halftime show bets, including the special appearances by Lady Gaga, Ricky Martin, but it only profited 17k, A tiny amount. And then another account created less than 24 hours before the game correctly bet on bad bunny set list wall street journal this morning with an article titled israeli soldiers accused Of using polymarket to bet on strikes israel arrested several people including army reservists for allegedly using classified information to place bets on israeli military operations Quote the account in question raked in more than 150,000 in winnings before going dormant for six months. It resumed trading last month, betting on when Israel would strike Iran. Polymarket data shows. The name of the account, Rico Suave 666. Suave. Rico Suave. Rico Suave. The name of the account, Rico Suave 666. I think that’s also the alias that you were using in Vegas for a little while there at your hotel. Rico Suave 666. The platforms are regulated, of course, by the CFTC. (Time 0:19:20)
  • Prediction Markets Echo Pre-Reg FD Dynamics
    • Insider information in prediction markets mirrors pre-Reg FD equity markets where asymmetry created massive returns.
    • Regulators face a hard choice: police insider bets or accept markets that favor sharp, informed traders. Transcript: David Friedberg Think the question is, is it really insider trading? If you and I were making a side bet, and I knew something about you and I had some edge or some advantage and I made a bet with you, is that fair? Should the government have a role in regulating that? This kind of goes back to securities regulation that everything needs to be registered. And then there’s this concept of insider information. It’s a real challenge and a real question on keeping the open platform of opportunity for trading on anything while also trying to mitigate the risk of what people call insider information In these trades. There’s a good chart that I think we talked about in our group chat that shows the distribution of accounts. There’s a few accounts that have a huge amount of money and make almost all the profits and then a lot of accounts that have a very little amount of money, and they get burned through very Quickly. They actually don’t have an edge. So the accounts that have a lot of money, they generally only trade in things where they have an edge, where they make markets, they actually have an arbitrage or yeah, sharps, and they Eat up all of the capital. So if you’re a marketplace like this, you probably also want to be thoughtful about the fact that over time, you could burn and churn through all of your customers, all of the users on The platform, if they’re constantly going to be making trades where they simply don’t have an edge, and all the capital, all the liquidity is coming from the accounts that do have an Edge and effectively trade off of inside information. So just be that these things end up eating themselves up. Jason Calacanis Shabbat, man, we had in trade, I’m sure you remember that. And I don’t know if that was in the early 2000s. This idea has been out there, but it has clicked right now for some reason. Chamath Palihapitiya What are your thoughts, broadly speaking, on the value of these platforms to society? Let’s define some terms first. So in betting, there are two kinds of people. There are the sharps who know what’s actually going to happen with a better edge. And then there are the squares, which is everybody else. And they are grist for the mill. And in a traditional market, like a sports betting market, there have been edge cases where you try to throw a game or throw a fight or shave points, and the sharps are involved in that. But it’s increasingly harder and harder to do because the sports leagues analytically are studying these things so closely to make sure that that never happens. But what you get are people with a smarter sense of what’s going to happen and people with less of a smart sense of what’s going to happen. The thing with prediction markets is it’s not just that. There will be those things, but then there are going to be these fundamental markets that are purely about inside information. And the question is, what can a regulatory body or a society do about that? And I think the answer is not much. And the reason is, is that if you try to regulate this, it looks like a securities market. And I think the problem there is that these things are too fluid and too dynamic and too ephemeral for them to be legislated like a security. And so why are these things happening? It’s because there’s too many of these prediction markets that can be manipulated this way. Somebody knows something that somebody else doesn’t know, and there’s no way to arbitrate that. This used to exist in the securities market, too. And this is where now I’m going to get a lot of people really upset with me. In 2000, we introduced the law called Reg FD. And what was the point of Reg FD? It was basically that if you’re a CFO, you cannot talk to an individual stock manager and tell him something that you then don’t tell everybody else, essentially inside information. That used to be not illegal. I won’t say that it was legal. I would just say that used to be not illegal. You call your buddy, he says, hey, how you doing? He goes, man, Corder was a blockbuster. You would go and buy the stock. And starting in the 2000s, it became illegal. And there used to be these networks of information arbitrage that took advantage of this. Now, this is an example of Warren Buffett’s returns pre and post Reg FD. Now, what do you see? His returns were double the market returns when this kind of information sharing was legal. And the minute that he became illegal, and you had to basically act on the same edge as everybody else, his returns went to the market return. He generated zero alpha. In fact, he probably on the margins lost a little bit. So this is the single best investor in the world. This is what happens when you have information symmetry. So it’s just meant to explain that markets thrive when there’s asymmetry. Billions and billions of dollars will be made in asymmetry. (Time 0:22:17)
  • Higher Rates Can Trigger A Debt Spiral
    • CBO projects unsustainable fiscal paths with deficits and rising debt; higher interest rates could create a self-reinforcing interest-on-debt spiral.
    • Small rate moves materially increase annual interest expense and accelerate debt growth. Transcript: Jason Calacanis The new CBO report is out. Freeberg, you said we are in a debt-death spiral. The congressional budget offer released its long-term budget forecast on Wednesday, February 11th. Here are the numbers. 2026 deficit is $1.9 trillion. That’s nearly 6% of GDP, much higher than the 3% GDP target we heard from Scott Bessett on this podcast. Social security, we talked about that before Freeberg. Trust runs out in 2032, one year earlier than previously expected. That’s obviously going to trigger all kinds of discussions around austerity measures that folks will not like. The debt will now grow from $31 trillion today to $56 trillion in 2036. So it is not stopping, folks. We are looking at an average of $2.5 trillion per year from 2026 to 2036. Also, currently, we’re at 120% debt to GDP. House Committee on Budget expects it to be 135%, so slightly up in 2036. For comparison, Japan is 237. Singapore, 176. Venezuela, 164. The Greeks, 154. UK, 94. 20 years ago, our debt to GDP was but 60%. Here’s a direct quote from the report. The fiscal trajectory is not sustainable. Okay, Dr. Doom. What do you think, Freiburg? This is your story, your chance to shine. David Friedberg Well, there’s no outlook to 3% deficit to GDP. There he is. And if you look at the assumptions, one of the key assumptions is that the short-term interest rate, which is largely how a lot of the debt is getting refinanced, is modeled to be around 3.1%. But if rates climb closer to 5%, as I mentioned in the past, just using the current debt levels, it adds another $650 billion a year of interest expense, which takes interest expense Almost up to $2 trillion a year, just paying the interest on the past debt. And because we’re running a deficit, that new interest expense increases the debt every year. So the debt goes up and up and up just by adding interest on past debt. And so this becomes the death spiral that we’ve kind of highlighted many times. (Time 0:32:44)
  • Growth, Not Only Austerity, Solves Debt
    • Economic outlook hinges on growth assumptions: modest CBO growth forecasts make debt look dire while AI-driven CapEx could materially raise GDP.
    • Strong productivity-led growth is the primary practical escape from fiscal strain. Transcript: David Sacks All agree about the problem of federal spending and the deficit and the debt, and we’re all concerned about that. With respect to the CBO study, however, I’ll just note that one of the key assumptions here is that CBO projects that real GDP will only grow by 2.2% this year in 2026. That’s a very low assumption, given that we grew by over 4% in Q3 last year, and the preliminary number for Q4 was over 5%. And I think all of our predictions for GDP growth this year when we did our predictions episode was 5% plus. So 2.2% is a pretty low number. And then they predict that it’s going to slow to 1.8% after 2026. So again, these are very meager, anemic growth assumptions. And if you believe that all of this CapEx that’s being invested in AI infrastructure is going to have a payoff, then the growth rates could be a lot higher. And that ultimately, I think, is the way to get out of the debt spiral is we need strong growth. Without that, we’re not going to get out of this problem. So look, I think that if you believe in growth, then the situation is not quite as dire. What would I do? Well, I mean, if I could wave a magic wand, the two key charts you want to look at are federal net outlays as a percent of GDP. This is from Fred, right? And then you want to look at federal receipts, which is tax receipts as a percent of GDP. And you just don’t want those lines to be more than call it 3% apart. I think that’s what Secretary Besson said is try to reduce federal deficits to 3% of GDP. Historically, tax receipts have bounced around 17% and the federal net outlays have bounced around 20%. So if you get back to that, we’d be in pretty good shape. And we were before COVID, our federal net outlays between spending as a percentage of GDP was around 20%. But then with COVID, it bounced all the way up to 30% in 2020 because of both the function of all the stimulus, but then also the fact that the economy shrank because of COVID. And we’ve never quite gotten back to that magic 20% number. Right now, it’s trending around 23%. So we’re doing a lot better than we did under COVID, but it’s still just a few percent higher. I mean, if it was up to me, I would just freeze federal spending until the economy grew to the point where federal spending as a percent of GDP is 20%. And then you could let federal spending continue to grow as the economy grows. And we’re not even talking about cuts here. We’re not even talking about shrinking the size of the government. We’re just talking about limiting the rate of growth until the overall size of the economy can catch up with it. (Time 0:38:00)
  • Cap Spending Growth Until GDP Catches Up
    • Freeze federal spending growth until spending as a percent of GDP returns to historical norms around 20%.
    • Let spending grow again only once GDP expansion restores the balance between receipts and outlays. Transcript: David Sacks And we were before COVID, our federal net outlays between spending as a percentage of GDP was around 20%. But then with COVID, it bounced all the way up to 30% in 2020 because of both the function of all the stimulus, but then also the fact that the economy shrank because of COVID. And we’ve never quite gotten back to that magic 20% number. Right now, it’s trending around 23%. So we’re doing a lot better than we did under COVID, but it’s still just a few percent higher. I mean, if it was up to me, I would just freeze federal spending until the economy grew to the point where federal spending as a percent of GDP is 20%. And then you could let federal spending continue to grow as the economy grows. And we’re not even talking about cuts here. We’re not even talking about shrinking the size of the government. (Time 0:39:51)
  • Hedge With Real Durable Assets
    • Hedge against currency and fiscal risk by owning durable real assets like gold and productive businesses.
    • Prepare for prolonged high debt-to-GDP regimes and global fiscal synchronization rather than panic over absolute levels. Transcript: Jason Calacanis Obviously, great thing that we’re shrinking the size of the government those people becoming more productive going into the private sector that’s a big win we all agree 10 great job In the first year hey maybe five percent the next two or three years would be even better but the debt continues to be a problem uh are you worried do you think there’s a solution here what Chamath Palihapitiya Would you do if you were running the show i think you have to take a broader historical context to this. Does debt to GDP matter? It depends on many things, but mostly I would say it doesn’t matter. And it’s very easy for people to get agitated about that. Now, there are things that matter when you print too much money, which is the value of the dollar, the value of exports, the cost of imports, and how to actually protect your earnings And your wealth. That’s a different question. This is a historical look back from about 300 years of debt to GDP of the largest functioning economies in the world. Now, what do you see? What you see is the trend where you, you know, if you smooth it out for wars, which by the way, has this weird effect of first escalating the debt to GDP, but then severely impacting it in A positive way. The Napoleonic War, the Fanko-Prussian War, World War II, these things all had positive effects on bringing debt to GDP once the war was over. But the general trend since 1700 to now is up and to the right. And the key observation is that it moves in unison, that these things are relative problems. So if the entire world moves in unison like this, there is an argument to be made, which is that you could end up at 300, 250, 200% of debt to GDP. But if everybody is there, nothing really changes that much. The real question is if one country is able to decouple itself and its economic output is so meaningfully different than everybody else’s. So my first take on this whole debt to GDP thing is I think you have to look at it together as a group. But for these other reasons, for earnings, for inflation, for all of those very practical reasons that impact your daily lived life. And what do we know there? We know that President Trump was elected on a massive mandate to secure the border on one hand, but to look at waste, fraud, and abuse on the other. And on that side, what did he do? He drafted the most important and prolific private businessman in the history of the world to be his tip of the spear. And what happened? They identified hundreds of billions of dollars, but when it came down to it and Congress had to act to solidify these cuts, they haven’t done much of anything. Is a way of saying that if the most conservative Congress in the history of the United States has not done much to solidify these cuts that were identified by the White House and Doge, Then as Friedberg said, it’ll only get worse if there’s ever a democratic house and democratic control. So what do we have to do? I think we have to just acknowledge that if debt to GDP continually moves in unison, the music isn’t up for a very long time. That’s just an observation. I’m not saying it’s right or wrong. It’s just the observation. But you got to find ways of hedging and owning real durable assets. Because the underlying currency that is used in these economies, even on a relative basis, will fluctuate wildly and just fall off of a cliff, which will mean that it will erode the value That you have created for yourself and your family. That I think is the most important takeaway from all of this, which is, we probably see things like gold do much, much better over time, because people will be afraid about the durability Of their dollar denominated resources. But it will also be true for all these other denominated resources. But I think debt to GDP, quite honestly, if I had to be a betting man, will trend into the 2345 600 on a relative basis for all countries. (Time 0:42:00)
  • Hosts React To Ferrari EV Leak
    • The hosts reacted to leaked Ferrari EV renders and interior, praising tactile controls but criticizing the exterior design.
    • David Sacks liked the interior’s mix of screens and buttons while Jason disliked the projected exterior. Transcript: Jason Calacanis With his partner mark newsome who also designed the iconic ford o21 concept car. We’re involved in this. Wait, what is that? This is like, if you’re a car nerd, this was like this incredibly innovative moment in design that never happened, that Ford did. It looks very similar to an Apple product. Here’s the key for the new Ferrari. David Sacks It looks like an animated character in cars. It does. Jason Calacanis You have this beautiful square glass key like an iPhone. You put it in and the yellow Ferrari yellow drains out and goes into the shifter. That was one nuance that people thought was very beautiful. The screen looks very Mac-inspired, except unlike Tesla, which is no buttons and removing buttons, they’re adding buttons here and making the buttons very tactile. All the sports car enthusiasts love tactile memory-based buttons that you can just have fun with and flip and feel like you’re a fighter pilot. Finally, turning the car on is like starting up a jet. You have a launch button, you twist and press, and it makes the whole car turn Ferrari orange or red. And yeah, that’s the inside. Saks, you buying one? You like it? David Sacks I saw everyone just, you know, all over this design. And I thought it was a little bit unfair in the sense that I actually overall liked the interior. I thought I found a compromise between, you call it the all-glass cockpit of a Tesla versus a totally analog old Ferrari interior. Like you said, it had a combination of screens, but then also buttons. And they made a point of showing that the buttons were not only nicely tactile, but they also made pleasing sounds and that kind of stuff. It seemed very heavy duty. So I thought the interior actually was pretty good. Again, nice balance between kind of the interior of a race car, the simplicity of that iPad screen, but also having enough sort of buttons that you develop muscle memory around where All the controls are. You don’t have to go hunting for them through a menu. I thought the mist here on the inside. I thought it was on the outside. I hate the look of the outside of this car. Jason Calacanis It looks to me like- By the way, just to be clear, that look is what people are projecting. It’s not the final version. (Time 1:03:58)