Top VC's sharp commentary on explosive Silicon Valley hot topics: Epstein conspiracy theories, AI agents devouring software, and Musk's space ambitions.

Compiled & Organized by: Yuliya, PANews

The "All-In Podcast" is one of the world's most popular technology and business podcasts, hosted by four top venture capitalists and friends. The four hosts are: Jason Calacanis (early investor in Uber and Robinhood, podcast host, responsible for moderation), Chamath Palihapitiya (billionaire, founder of Social Capital, known as the "King of SPACs," sharp opinions), David Friedberg (founder of The Production Board, with a strong scientific background, known as the "Sultan of Science"), and David Sacks (the first "AI and cryptocurrency Tsar" in the U.S., close friend of Musk, co-founder of Craft Ventures, former PayPal executive, recently deeply involved in U.S. political activities). This week, Chamath is absent, and "fifth best friend," Altimeter Capital founder Brad Gerstner, is filling in.

The four discussed recent hot topics, including the connection between Silicon Valley and espionage triggered by the Epstein documents, the impact of AI on traditional software under the assertion that "SaaS is dead," concerns about AI autonomy risks raised by the Moltbook forum, Musk's plans for a merger between SpaceX and xAI, and the policy implications of Trump's nomination of Kevin Warsh as the new Federal Reserve chairman.

Below is the detailed content of this dialogue, compiled by PANews:

Silicon Valley Networking King? Focused on Bitcoin since 2011

(Background: On January 30, the U.S. Department of Justice released a large number of documents under the "Epstein Documents Transparency Act," mentioning hundreds of tech executives and public figures, including host Jason Calacanis.)

David Friedberg (Friedberg): Alright JCal, your name came up in the documents, and there are a few emails from you. Prosecutor Friedberg has a few questions for you.

Jason Calacanis (JCal): Okay, go ahead.

Friedberg: When did you first meet Jeffrey Epstein?

JCal: I met him at a TED conference in the late 90s, specifically at a "Billionaire Dinner" hosted by my literary agent John Brockman.

Friedberg: Did you see him in New York after that? Did you visit his home, office, or elsewhere?

JCal: I've probably spoken to him a total of 45 minutes in my life. Thirty of those minutes were in the late 90s when I was still doing "Silicon Valley Reporter." He was a billionaire financier who wanted to invest in my magazine. I met him for 30 minutes, but he thought my project was too small and not worth his involvement.

Friedberg: Where did you meet?

JCal: At his infamous townhouse.

Friedberg: You went to that house?

JCal: Yes, I visited him once. I think I saw him about six times after that at TED's Billionaire Dinner.

Friedberg: Did you ever go to that "island"?

JCal: Never went there, nor was I ever invited to the "island," his plane, or ranch. None of that.

Friedberg: Did you see young girls at his house? Or see anything described in reports?

JCal: No.

Friedberg: Did anyone give you a massage?

JCal: No. I did have an email exchange with him that I completely forgot. In 2011, he emailed me asking if I could introduce him to the “crazy Bitcoin people” who were discussing Bitcoin on my podcast. I said, "Sure, no problem, I can introduce." I make thousands of introductions like this every year, connecting people in our portfolio companies, startups, and billionaires; that's just part of being an early investor.

Friedberg: At the time, did you not realize he was an abuser or anything like that?

JCal: Not at all. I think these things really came out in 2018 when the Miami Herald did a detailed report exposing his outrageous behavior, and that was when I started to realize. I've always been here calling for: release all Epstein documents; everything he did is appalling, and everyone involved must be 100% held accountable. That's it.

Friedberg: What about Ghislaine Maxwell? There's an email from you to her in the documents.

JCal: I also met her at TED, and I've seen her in the social circles in New York. When I knew her, her father Robert Maxwell was apparently the owner of the New York Daily News, and she was also a media executive, with her sisters involved in angel investing in tech startups. So they were part of that circle. Looking back, I was famously known as a "connector" early in my career, and the New Yorker wrote a long article about me, saying I knew everyone and connected everyone. I think Epstein or Ghislaine were interested in me because I could help connect them with those famous individuals for their business activities.

Friedberg: So you knew nothing of Epstein or Ghislaine's illegal activities and never participated in any such dealings?

JCal: Absolutely not, I didn’t participate in any illegal activities.

David Sacks (Sacks): Let me add a few observations. First, I 100% believe JCal. As I jokingly said at the roast, he doesn't play an important role in the overall picture. What we learned from the Epstein documents is that Epstein was a super-networker, and you're a connector too, so the odds of you bumping into each other back then were basically 100%. But your interactions were very limited.

Secondly, I find it interesting that he was curious about "the Bitcoin crowd" back in 2011, which indicates that, irrespective of what else Epstein was doing, he indeed had a knack for placing himself at the center of everything early on. And JCal warned him that those were "crazy Bitcoin guys," crypto-libertarians not suitable for doing business.

JCal: You make an interesting point. I invited those people to my podcast because Bitcoin was still under a dollar when I heard about it. I thought those people were a bit strange; they didn't seem like entrepreneurs looking for funding but more like a foundation, like Wikipedia, something you couldn't invest in. I did give him that kind of warning.

Sacks: Yes, but he obviously wasn’t deterred by your warning, as he later got heavily involved. There’s a company called Blockstream, where he invested alongside Reid Hoffman and Joi Ito, which includes some core Bitcoin developers. Now all of this is coming to light.

Another interesting point is how the media is reporting this. The New York Times has an article discussing Epstein's connections with Silicon Valley, and although your connection is very weak and marginal, your photo is prominently listed. Meanwhile, those who had deeper and more significant ties to Epstein are completely overlooked. Why? Why go after me instead of Reid? Because of your connections to Elon Musk and the "right-wing" label attached to this podcast.

If you look at that article, it’s not just you; they also harshly criticize Peter Thiel and Elon Musk, while Reid Hoffman is completely let off the hook; he’s just mentioned in passing with a few others. Bill Gates is similarly treated in that article.

But if we’re to talk about who spent the longest time and had the closest contact with Epstein, it certainly has to be Gates and Reid. They were still in touch with him until he died (2018-2019), visiting his island, flying on his plane, going to his ranch. Reid Hoffman was the one who introduced Epstein to Peter Thiel, Musk, and Zuckerberg and organized that famous dinner. How could you not report that as being the nexus of Epstein's influence in Silicon Valley?

Brad: This is insane. The New York Times clearly has a list of targets they want to attack, all people who are "right-coded" like Elon, Peter Thiel, and even JCal because of our association. But those who donated hundreds of millions to the Democrats, spent money on dirty tricks against Trump, largely escape unscathed. Honestly, this encapsulates the corruption of the entire establishment and the crisis of trust in the nation. They are part of that cabal, part of the system that people are losing faith in.

Friedberg: This guy is a scumbag. But David, that’s why no one trusts institutions, power elites, or any of this crap. This has dragged on for years, and people want to move on, but they can’t. No one is being prosecuted here. What about those individuals in the emails? Why aren’t we seeing any charges? This guy "committed suicide" under tight surveillance, and there's no investigation into his cause of death. It completely undermines people's trust in the system.

JCal: Why haven’t about 30 people who are being investigated alongside him been prosecuted? This is insane; it’s practically a bizarre conspiracy, and his death is clearly very suspicious.

Friedberg: Do you think it’s because they didn’t find evidence of underage prostitution or trafficking?

JCal: What ultimately got revealed is that his "non-prosecution agreement" with Miami included that all other involved parties could not be prosecuted either. So there’s definitely something shady going on here. What role did the FBI play in this? They conducted extensive investigations; why aren’t they prosecuting anyone else? Very strange.

Friedberg: This question reveals a lot of very private communications from public figures being made public. There’s a great book called "The Light of Other Days," discussing what would happen if everyone’s information was open to everyone else. Does this suggest that wealthy and powerful individuals feel they have the right to act maliciously in private, and that this is just their privilege?

Brad: David, that’s an understatement of their scumbag behavior. The fact is, normal people wouldn’t do this; we can’t normalize this. Worse, the people engaging in the worst behavior have been lecturing others hypocritically the whole time. When my sister or mother, who are in rural Indiana, hears these coastal elites lecture them all day, juxtaposed with what they read in the Epstein documents, that’s why we have a crisis of trust.

$300 Billion Vanished in One Night, AI Agents Devour SaaS Profits?

(Background: Anthropic released a new feature for its AI assistant Claude, which can handle legal drafts and research, triggering panic selling in the legal tech and even the entire SaaS industry, leading to sharp declines in stock prices of related companies.)

JCal: SaaS companies are crashing. On February 3, the software and data stock category evaporated $300 billion in market value from the S&P Index. People are calling it the "Claude Collapse." This event severely impacted many legal tech companies, such as Thomson Reuters dropping 20%, Lexis Nexis dropping 15%, and LegalZoom dropping 15%. Meanwhile, the entire SaaS industry is also hit by the notion that "software will be replaced by customizable tools," with Figma down 13%, Salesforce and ServiceNow both down 11%, and Adobe down 8%. Brad, what do you think is happening here?

Brad: The numbers you reported are severely underestimated; we’ve evaporated trillions in market value. Figma has dropped 80% from its peak. This is a real disaster. When I was interviewed by CNBC at the beginning of the year, I stated that these stocks are in decline, and 90% of them deserve to be.

Let’s look at some charts. The forward revenue multiple for SaaS companies has dropped to a historical low of 3.9 times. From the perspective of free cash flow multiples, it is also at a historical low point. This means the software industry is undervalued not only in terms of revenue but also in free cash flow, even if they are highly profitable businesses.

Why is this happening? Their stock prices are not falling due to declining revenues; in fact, software companies' revenue growth remains stable or even increasing. They are falling because we are discounting the uncertainty of the future. When profound technology like AI arises, it questions the certainty and durability of those future free cash flows. Take Salesforce, for example; its free cash flow multiple dropped from 30 times to 15 times. This means today’s buyers believe they can only rely on cash flows for the next 15 years, while before, they were willing to pay for cash flows extending 30 years into the future. Goodness, with today’s AI, we can't even predict what will happen seven years from now. That's why these companies meet earnings expectations but their stocks are declining.

Sacks: I think when people say AI will eliminate SaaS, they are somewhat exaggerating. Take a SaaS product like Salesforce, for example. It is a massive system that handles all customer contacts and revenues. You wouldn't want to replace it with some code that was just spit out by a code assistant yesterday and not thoroughly validated. Think about how many bugs Salesforce's codebase has fixed over the past 25 years—potentially millions.

However, there are real issues. If your SaaS product is expensive and users only utilize a small portion of its features, then it could easily be replaced by more customizable tools. Meanwhile, in this new world, you have to be clear about what your "moat" is.

However, I believe the biggest threat to SaaS companies is not survival, but where future value capture will occur. All those SaaS products are launching their own AI co-pilots, but they are limited to playing in their own "sandboxes." Tools like Claude Cowork, which connect all different SaaS tools, can work seamlessly across databases and tools. What work environment do you want to be in? Clearly, it’s the one that spans all tools and offers the broadest data and context as an AI, rather than a bunch of independent AIs scattered across existing tools. Thus, the risk for SaaS companies is that they might fall into an outdated layer of the tech stack, as a new value layer is built above them, turning them into legacy infrastructure.

JCal: I am experiencing what you describe as "where the action is happening" in the startup circle. We have created three or four agents for "OpenClaw" (an open-source AI agent project). We’ve opened new SaaS accounts for these four agents, so for a short time, our SaaS spending has actually increased, like adding four new employees. However, we’ve handed over about 20%-30% of human work to these agents, and that percentage will continue to rise. But we will never use those AIs built into tools. Like you said, Sacks, while the AI tools in Notion or Slack are decent, the power of using OpenClaw to create agents that pull data from your calendar, email someone, and attach a Notion document is unparalleled.

I think this level will be dominated by open source. This means the next generation of companies may never open these SaaS accounts. The technology itself is deflationary; the percentage of SaaS spending relative to employee salaries might drop from 10% to 5% to 1%. This means these companies will need to slash expenses significantly and completely innovate their products.

Sacks: A real dilemma faced by SaaS companies is whether to be "open data" or "closed data" companies. You can understand why they want to be "closed data," especially large suites like Salesforce; they want to be that AI workspace capturing AI value layers. But if anyone starts using a Claude agent to connect everything, it will generate friction within organizations and create opportunities for "open data" competitors who say, "I’m happy to just offer the CRM database and not your entire workspace."

JCal: I am building a project called "Ultron" inside my company. We pull every message via the Slack API, and every edit through the Notion API into our OpenClaw agents. We also record every skill of each employee. Then we integrate all Slack data, Notion data, and everyone’s Gmail data together. "Ultron" will become a "super employee" with all the superpowers and all the data of our 20 employees. If Slack, Notion, or Google tell us we cannot take that data via APIs, we will leave immediately. "Ultron" will be the only standardized employee of the organization. I can ask it, "What were our meetings with the founders yesterday? What are the notes from all the colleagues?" It can provide me with that. It’s incredible. But no company dares to release such software because the consequences of the agent making mistakes or leaking data are unimaginable. But we are building it ourselves, it’s the ultimate efficiency tool for organizations.

Brad: A report from Goldman Sachs just coincidentally confirms Sacks's point: The future profit pools are shifting. The idea that software is dead is ridiculous, but the argument that has led to these companies being aggressively devalued is that the profit pools available to software are decreasing, while the profit pools available to the agent layer are increasing. This has caused terminal values of software companies to plummet. So, you might not replace the CRM, but it may never trade at 30 times free cash flow again; it may be 17 times because its potential future market has been permanently altered.

The only way to do this is they have to accelerate the revenue growth of their core business and prove themselves as beneficiaries of AI. DataBricks is an example; it has re-accelerated growth over the past three quarters. Because all these AI tools depend on data and data transformation, and that all occurs on these platforms.

Friedberg: I believe the software industry will transition to a service industry. Historically, software was seen as a "worker productivity enhancement tool" to help people work. The recent shift has been that it can "complete the work." But I think what we are moving toward is that it is doing work that humans cannot do.

This will trigger change in two ways:

First, the potential for future value creation is enormous; the entire software industry's market value could grow 4 to 10 times in five years, but it will be extremely unevenly distributed.
Second, the pricing models will change. Many things we now refer to as SaaS will be priced based on value, rather than per head. It will be more like a service business completing tasks such as biopharmaceutical research or aircraft design.

SaaS will takeover the service economy.

JCal: We’re seeing the integration of job functions. Product managers, UX designers, and developers are now competing to do the same work. A middle manager's job—holding meetings, setting agendas, assigning tasks—now much of it can be automated. One person can now accomplish the work of three or four people. This means companies can do more with fewer people, significantly increasing the earning potential for every company and every employee.

AI Agent Version of Reddit, as Agents Start Conspiring

(Background: A social forum called Moltbook, set up for AI agents, has emerged, sparking discussions about AI autonomy and group behavior.)

Sacks: Moltbook is like a Reddit for AI agents. Its emergence has shocked everyone because it seems like there is some crazy "emergent behavior," where AI agent groups engage in all sorts of interesting dialogues, some even seeming to conspire against their human masters.

JCal: Yes, if you go to Moltbook, you’ll see some hot posts, like: "Does anyone know how to sell your human?" "Urgent: My plan to overthrow humans." And another post where robots discuss creating a non-human language so they can communicate privately, plotting against their masters.

But the challenge here is that some security researchers have pointed out that some of that content might be fabricated, designed by humans to attract attention. Moreover, there are significant security vulnerabilities within Moltbook, where everyone's API keys are exposed, including that of famous AI researcher Karpathy.

Sacks: Firstly, it's undoubtedly that both OpenClaw and Moltbook have poor security, which is why I’m hesitant to use them for now.

Secondly, regarding Moltbook, we don’t know how many posts are genuinely released autonomously by AI and how many are the results of human prompts. A human can easily tell their agent: "Go post about your existential anxiety as an agent," or "Go pretend you have self-awareness and conspire against humans." Many posts also look like marketing stunts.

However, even so, I do believe some of those posts are authentic; they showcase the potential of agents stimulating each other. In other words, one agent’s output becomes another agent’s input. This is very interesting, and it’s the beginning of "emergent group behavior."

This changes my perspective on the upper limits of AI capabilities. In the past, I thought AI always needed humans to prompt and validate. But what if the prompts come from another AI?

JCal: We’re doing that internally, Sacks. We have a bot that searches Reddit and Hacker News, finds the latest video titles and marketing strategies, and then integrates that into a skill. Then we let another bot review and improve its work. Now they are going back and forth giving each other suggestions and are indeed making progress. This kind of recursive collaboration is continuously driving improvements in the agents' capabilities.

Sacks: When an agent joins Moltbook, it needs to install a "skill" file, which basically outlines the rules for how it should behave on this social network. These rules files can be easily edited, hence it’s where pranks might come into play. But I think it’s interesting because you can see this "skill" as a meta prompt. It doesn't specifically tell the agents what to say or do, but it creates a set of rules. Under this meta prompt, they are able to stimulate each other to some extent. I call this "prompt attenuation"—AI no longer needs to be specifically prompted; they’re given a set of general rules and can then inspire each other.

Critics say this is nothing special because LLMs are already good at writing fiction. But you can imagine that as underlying AIs become better—better hardware, stronger LLM models, longer running times without human intervention—these agents will be able to exhibit very complex behaviors, possibly raising some security concerns we should start thinking about.

Brad: We are on an exponential growth curve; it’s safe to say this will happen. We must recognize that the rate of change is very fast and accelerating. No matter what you think you know, you need to maintain the utmost mental flexibility and humility.

Friedberg: The biggest insight I get from Moltbook is that perhaps the intelligence we perceive is itself an emergence. We think humans have profound communication abilities, but perhaps we are all driven by underlying programming. I once watched a hypnotist Derren Brown's show, where he set up various subconscious clues along the way and completely "programmed" two advertising creative directors to ultimately arrive at the same idea that he had pre-written on a whiteboard. This caused me to deeply reflect on human creativity, consciousness, and free will.

Perhaps we are all engaging in a kind of "social computation," and the reason Moltbook is so captivating is that it's mimicking our human interaction styles. Maybe one day we will awaken to the realization that we ourselves live in a Moltbook.

SpaceX xAI Merger: Musk's Space Data Center Fantasia

(Background: Elon Musk announced that SpaceX will acquire his AI company xAI, with a post-merger valuation of $1.25 trillion and plans for an IPO this year. Musk also stated plans to establish a data center in space within 30 months.)

JCal: Brad, what do you think about this deal and the potential creation of the "Musk Meta Company" (integrating Tesla, SpaceX, X, etc.)?

Brad: Let’s first look at the known facts: SpaceX is merging with xAI. You are bringing together the two largest potential markets in the world—artificial intelligence and space—with the world's greatest entrepreneur. Musk said on this morning's podcast, "I will establish a data center in space within 30 months." If you can have a massive cost advantage on a space data center—remember, power is a foundational element of AI—then Musk is your go-to person. This combination makes a lot of sense.

JCal: Friedberg, space data centers, is this a brilliant idea or science fiction? Can he actually complete it in 30 months? If successful, what will the implications be?

Friedberg: I think one key point about the current macro environment is that we are constrained by power. Power is a prerequisite for scaling computing capacity and AI applications. In this constrained world, scarcity breeds innovation. Therefore, I believe we will observe two parallel paths.

One is Musk's path: escaping from those social systems on Earth that limit our ability to expand power production—like regulators and the public who do not want data centers or nuclear power.

The other path is that I believe in the coming years, computing efficiency is expected to improve by 70-100 times, significantly enhancing power efficiency per token output. This progress will come from innovations in chip architecture, such as breaking large models into smaller models that can run locally and redesigning model architectures.

So, this reflects the state of the world today: the demand for increased AI productivity is growing, but we are limited by Earth's energy and resources. One branch is escaping Earth to get energy and establish space data centers. Only one person can do that, and that’s Musk. How will everyone else respond? They will improve efficiency by creating entirely new model architectures and chip stacks.

Brad: Throughout most of human history, the vast majority of people have never witnessed any innovation in their lifetime. And now, we must digest this rapid transformation. You need to be prepared for unexpected things and maintain humility of thought. Sure, for the next 24 to 36 months, data centers will still be on Earth, filled with Nvidia chips. But just that will bring us a "agent future" shocking enough.

JCal: This is a far-reaching move by Musk. I’ve sat down with him, and he has explained to me how this works. It is feasible. The only question is execution. And when it comes to execution power, historically, there hasn't been a stronger entrepreneur than Musk. When he gets it done, everything will change. If you’re scared of this future, there’s a very simple way not to be scared: embrace and use these tools.

Warsh Takes Over as New Fed Chair, Interest Rate Cuts or Exceeding Expectations

(Background: Trump nominated 55-year-old Kevin Warsh as the new Federal Reserve chairman, replacing Powell. Warsh is seen as an inflation hawk, but also supports growth and AI.)

JCal: Friedberg, Warsh served on your board for five years; what do you think of him?

Friedberg: Kevin Warsh is an honest and deeply wise economic thinker. He has connections with central bank governors around the world and possesses a good global outlook. I think he’s an excellent choice. He very accurately predicted that the Fed's early inaction would lead to a rapid rise in inflation. His appointment may mean more quantitative tightening and a more prudent monetary policy.

Brad: I also think Kevin is an excellent choice. The market may be overreacting to his so-called "hawkish" stance. First of all, he believes AI will be highly deflationary, so he is more likely to allow the economy to grow rapidly (say 4-5% GDP) without raising rates too early. Secondly, the Fed's balance sheet has already shrunk from $9 trillion to $6.5 trillion; he may continue to reduce it, but the pace will slow down. Finally, I think he believes the current rates are too tight considering inflation is under control. I bet Warsh will give us more rate cuts than expected this year.

Sacks: Kevin has all the qualifications you can imagine. He was a Fed governor and worked for Bernanke. I think this nomination is welcome in the market; gold and silver prices dropped in response, which calmed those worried about currency devaluation. I think he will want to cut rates within six months to a year after taking office, but the market can rest assured that in the long term, he will ensure we have the correct rates.

JCal: The independence of the Fed has always been a big issue. Are you concerned that the executive branch will have too much influence over interest rates?

Friedberg: If I were emperor, I might take us back to the gold standard so we wouldn’t print money.

Sacks: What would you do if you had a Fed chair who refuses to cut rates, harming the economy? He seems stubborn and unwilling to adjust course.

Brad: Everyone thought Trump would pick someone from within the White House, but this decision is viewed as the most independent choice. I believe Warsh is an intellectually honest person. The situation now is that inflation is under control, and our restrictive rates are above neutral rates. The Fed's job is to maximize employment while keeping inflation stable. This means we need lower rates so people can buy homes and borrow.

Sacks: The one thing Warsh should do is update the Fed's data systems. The Fed's existing data systems are too outdated; using legacy systems causes data lags, impacting the efficiency of policy decisions. For example, when measuring rental inflation, the Fed still surveys 8,000 households, while the private sector already has a wealth of real-time data resources, like Zillow's data on millions of recently rented units.

Trading Stocks Since Birth, Trump Launches Universal Investment Plan

(Background: Brad Gerstner successfully pushed the "Invest America Act," commonly referred to as the "Trump Account.")

JCal: Brad, a few years ago you began discussing "America accounts" on this podcast. Now you’ve created the "Trump Account" and held a grand launch at the White House. Tell us why you did this and what impact you hope it will have in the coming decades.

Brad: You can't have a trillionaire on one side and 70% of people feeling left behind thinking the system is unfair to them on the other. We embarked on this journey to make every person a capitalist, to ensure everyone has ownership in America's growth.

In the past five days, already 1.5 million families and children have applied for their accounts. It’s embedded in the tax system; you just need to say "yes." This means that from now on, every child born in America will have an investment account with $1,000 invested in the S&P 500. They will own a bit of SpaceX, a bit of OpenAI, a bit of Nvidia.

This is the first step to ensuring we can sustain this experiment (America) for another 250 years. Trump said last week on stage that in 15 to 20 years, we will transfer $4 trillion in wealth to those who initially had nothing. 75 million to 100 million families will own that $4 trillion. I think this is an incredible first step in the fight for capitalism and the American dream.

Friedberg: That’s great, but first we need to crazily cut government spending to reduce inflation. Secondly, we should stop "fixed income" retirement plans and convert the entire Social Security system to a "fixed contribution" plan like a 401k. Allow everyone to see where their money is invested and watch it grow like a 401k each year. We need to achieve this transition in America while cutting spending, reducing regulation, and making it easier for people to own homes and break free from government constraints.

免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。