The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch

20VC: NVIDIA Leadership Turmoil, Anthropic's Record IPO, and AI Agent Disruption

1h 13m Listen to the original Episode page
NVIDIAAnthropicVenture CapitalAI AgentsSaaS DisruptionIPOWorkforce AutomationSpaceX

Summary

Speakers analyze the fallout from Jensen Huang's departure from NVIDIA, debating the legal and reputational risks that may delay his return ahead of Anthropic's historic $1.4 trillion IPO. They highlight Anthropic's extreme operational efficiency, noting it reached $100 billion in revenue with fewer than 6,000 employees, a shift that is redefining venture capital metrics toward $20 billion+ outcomes. The discussion covers the consolidation of venture capital, with deal volume down 21% but capital deployed up 2.5x, and examines late-stage investments in Kalshi, Mercor, and Fractile. Finally, the hosts explore how AI agents are displacing traditional SaaS for small businesses, citing Meta's Muse and the workforce reductions at Jason Calacanis's company as evidence of a broader industry transition.

Episode sections

  1. 0:00 – 14:00

    NVIDIA Leadership Crisis and Jensen Huang's Departure

    Speakers discuss the fallout from Jensen Huang's departure from NVIDIA, citing personal allegations and export control compliance issues as key factors. They debate the likelihood of his return, noting the board's long-standing relationship with him versus the legal risks posed by DOJ and Commerce Department investigations into chip sales to China via third countries.

  2. 14:00 – 19:00

    Anthropic's Record-Breaking IPO and Market Implications

    The podcast highlights Anthropic's $1.4 trillion IPO, described as 7x to 10x oversubscribed. Speakers analyze the company's extreme efficiency, noting it reached $100 billion in revenue with fewer than 6,000 employees, compared to tens of thousands at Google and Meta. This efficiency is framed as a major shift in the venture landscape, raising the bar for future returns.

  3. 19:00 – 23:00

    Venture Capital Dynamics: The $20 Billion Club

    Speakers observe that venture firms are now graded on the number of $20 billion+ outcomes rather than unicorns. Data from Carta shows deal volume is down 21% while capital deployed is up 2.5x. The discussion notes that mega-funds are raising quickly in Q1 2026, potentially fueling rapid markups and compressing future returns.

  4. 23:00 – 27:50

    Investment Committee: Kalshi, Mercor, and Fractile

    The hosts review three late-stage deals: Kalshi ($65B valuation), Mercor ($20B valuation), and Fractile ($7.5B valuation). Jason Lemkin expresses discomfort with Kalshi's regulatory risk and Fractile's lack of revenue, while favoring Mercor's data positioning. Harry Stebbings argues for Kalshi's political entrenchment and Fractile's strategic orders from Anthropic and NVIDIA.

  5. 27:50 – 32:51

    NVIDIA Deal Spree Under Threat: AI Startups Lose Their Safety Net

    Speakers discuss how NVIDIA's leadership uncertainty weakens the 'liquidation preference' safety net for AI startups, as Jensen Huang's potential departure reduces the likelihood of NVIDIA buying out struggling companies like Poolside or Hugging Face. This shift increases regulatory and financial risks for venture deals previously backed by NVIDIA's implicit guarantees.

  6. 32:51 – 41:45

    HubSpot's Founders Return as Customer Growth Slumps

    The return of HubSpot founders to the executive team is analyzed against a backdrop of declining net new customer acquisition, with growth rates falling from 10,000 to under 5,000 per quarter. The discussion highlights the difficulty of reversing SaaS growth trajectories and compares HubSpot's situation to Intercom's founder return, noting that while founder energy helps, it may not be enough to restore high growth.

  7. 41:45 – 45:25

    Small Businesses Shun SaaS: AI Agents Win the Budget

    A significant shift in SMB software purchasing behavior is identified, where small businesses are delaying or skipping traditional SaaS purchases like HubSpot or Monday.com in favor of AI agents that automate front-office tasks. This trend suggests that AI agents are replacing the need for dedicated CRM or workflow software for many SMBs, who prefer low-cost, high-ROI automation over complex software stacks.

  8. 45:25 – 54:07

    Microsoft Outage & FAA Shutdown Expose Fragile Infrastructure

    Recent outages at Microsoft and the FAA are attributed to classic IT configuration errors and legacy infrastructure issues rather than AI-related cyberattacks. These events highlight the increasing fragility of critical societal systems, which are vulnerable to cascading failures and simple load-based attacks, raising concerns about the competence and security of companies managing mission-critical services.

  9. 54:07 – 54:48

    OpenAI Buys Odessia: The End of Vertical Personal Assistants?

    The segment concludes with a discussion on OpenAI's acquisition of Odessia, a travel-focused personal assistant. The speakers question whether this signals the end of verticalized personal assistants, suggesting a potential industry shift toward horizontal AI solutions that handle multiple domains rather than specialized vertical apps.

  10. 54:48 – 58:31

    Meta's Muse as a Small Business Agent

    Jason Calacanis describes using Meta's Muse as a primary horizontal agent for his business, running 20-30 workflows including outbound sales, content editing, and deal screening. He notes that while the moat is shallow, the high switching cost of established workflows creates lock-in, potentially displacing traditional SaaS tools like marketing automation software.

  11. 58:31 – 1:04:28

    Competitive Dynamics: Muse vs. Perplexity's InStinct

    Calacanis reveals he spends ~$1,000/month on Anthropic's Claude Pro to power Muse, whereas Meta's Muse is free or low-cost. He expresses concern that Muse's superior capabilities and distribution could undermine Perplexity's InStinct, though he acknowledges InStinct may retain users if it offers unique capabilities like headless browser interactions.

  12. 1:04:28 – 1:06:42

    Workforce Displacement and Agent Management

    The speakers discuss the shift from humans using agents to agents replacing humans, noting that companies will stop backfilling roles rather than firing staff. Calacanis states his team has reduced from 20 humans to 3 humans and 21 agents, requiring significant time to manage the agents.

  13. 1:06:42 – 1:08:47

    SpaceX Valuation and Compute Focus

    Rory Currie argues that SpaceX's market cap is driven by its compute business (Starlink) rather than its rocket launches. He suggests buying SpaceX stock is a bet on future compute demand and pricing power, specifically referencing potential payments from Anthropic for compute resources.

  14. 1:08:47 – 1:12:57

    Jensen Huang and Anthropic IPO Speculation

    Harry Stebbings predicts NVIDIA CEO Jensen Huang will return to work within three days, arguing that the news cycle pressure surrounding the Anthropic IPO makes a quick return necessary to maintain momentum. He contrasts this with Sam Altman's quicker return.

Assets discussed

AssetKindStanceContext
NVIDIANVDAequityneutralI think the liquidation preference has been weakened by the NVIDIA drama.
MicrosoftMSFTequityneutralI feel we've got to talk about Microsoft going down for two days.
OpenAIprivate companyneutralIt's OpenAI buying a company called Odesia.
Anthropicprivate companyneutralIt is the best software I've used since since like Anthropic 3.x got good.

“Stance” is how the speakers talked about the asset, not a recommendation.

Claims worth verifying

Jensen Huang was let go from NVIDIA, and the stock lost up to $600 billion in value in a single morning.

Harry Stebbings ·0:45

How to verify: Check NVIDIA stock price history for the week of the release date (Nov 2026) and review NVIDIA press releases or news reports regarding Jensen Huang's departure and the board's statement. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Anthropic's IPO is valued at $1.4 trillion and is 7x to 10x oversubscribed.

Unidentified speaker (speaker_1) ·0:21

How to verify: Monitor financial news outlets (e.g., Bloomberg, Reuters) for the official pricing and subscription details of Anthropic's S-1 filing and IPO launch. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Anthropic will have less than 6,000 employees at $100 billion in revenue, compared to Google's 81,000 and Meta's 71,000 at similar revenue levels.

Unidentified speaker (speaker_1) ·14:30

How to verify: Verify employee counts via Anthropic's S-1 filing (SEC EDGAR), Google's and Meta's most recent 10-K filings, and compare against their reported revenue figures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Anthropic spent approximately $10 billion to train its last model.

Unidentified speaker (speaker_0) ·12:47

How to verify: Check Anthropic's S-1 filing (SEC EDGAR) for disclosed training costs or compute expenditures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Venture deal volume is down 21% while dollars invested are up 2.5x in Q3 2026, according to Carta data.

Unidentified speaker (speaker_1) ·19:29

How to verify: Access Carta's Q3 2026 Market Report or Venture Capital Index data to verify the percentage changes in deal count and capital deployed. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Fractile has a $250 million order from Anthropic.

Harry Stebbings ·25:57

How to verify: Check Fractile's press releases or SEC filings (if public) or news reports confirming the order size and customer. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Anthropic's S-1 discloses that 70-75% of their compute budget is allocated to specific activities.

Unidentified speaker (speaker_0) ·27:30

How to verify: Review Anthropic's S-1 filing (SEC EDGAR) for the section detailing compute costs and budget allocation. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

NVIDIA would have bought Hugging Face for around $10 billion, but the deal is now at risk due to regulatory and leadership changes.

Unidentified speaker (speaker_1) ·30:04

How to verify: Check SEC filings for any Hugging Face acquisition announcements or regulatory filings related to NVIDIA; monitor news for official statements on the deal status. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Poolside sold because they could not raise the additional $10-20 billion needed to sustain their business model.

Unidentified speaker (speaker_1) ·32:22

How to verify: Review Poolside's press releases or funding announcements for details on their capital needs and sale rationale; check Crunchbase or PitchBook for deal terms. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

HubSpot's net new customer acquisition dropped from 10,000 per quarter to under 5,000 per quarter.

Unidentified speaker (speaker_1) ·33:31

How to verify: Check HubSpot's quarterly earnings reports or investor presentations for customer acquisition metrics; verify against company guidance. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Monday.com's net new customer accounts fell by 80%, from over 1,200 to under 200.

Unidentified speaker (speaker_1) ·33:50

How to verify: Review Monday.com's quarterly earnings reports or investor presentations for customer acquisition data; verify against company guidance. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Intercom's founder return led to operational fixes but only $100 million in agentic revenue before acquisition by Salesforce.

Unidentified speaker (speaker_1) ·35:50

How to verify: Check Salesforce's acquisition announcement and subsequent earnings reports for Intercom's revenue contribution; verify Intercom's pre-acquisition financials. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Salesforce has not disclosed its customer count for four years.

Unidentified speaker (speaker_1) ·40:22

How to verify: Review Salesforce's quarterly earnings reports and investor presentations for the last four years to confirm the absence of customer count disclosures. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Microsoft's outage was caused by a network configuration issue, not a cyberattack.

Unidentified speaker (speaker_0) ·46:16

How to verify: Check Microsoft's official incident reports or press releases regarding the outage; verify with cybersecurity firms' analysis of the incident. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

The FAA outage was caused by a router issue involving a 1998 Linksys router.

Unidentified speaker (speaker_1) ·47:57

How to verify: Review FAA incident reports or news articles detailing the cause of the outage; verify with technical analyses from IT security firms. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

OpenAI acquired Odessia, a travel-focused personal assistant founded by Francis Davidson of Sonder.

Harry Stebbings ·54:27

How to verify: Check OpenAI's official announcements or press releases regarding the acquisition; verify Odessia's founder and business model through company profiles or news articles. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

NVIDIA's leadership uncertainty has weakened the 'liquidation preference' safety net for AI startups.

Unidentified speaker (speaker_1) ·29:25

How to verify: Analyze venture capital term sheets for AI startups to assess the prevalence and enforceability of liquidation preferences; monitor news for NVIDIA's M&A activity and leadership changes. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Speaker 1 uses Meta's Muse for 20-30 workflows, including outbound, competitive review, content editing, and deal screening.

Unidentified speaker (speaker_1) ·55:41

How to verify: Check Meta's official documentation or press releases regarding Muse's capabilities and use cases; verify via public case studies or interviews with Jason Calacanis. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Speaker 1 spends $1,000/month on Anthropic's Claude Pro to power Muse workflows.

Unidentified speaker (UNKNOWN) ·58:35

How to verify: Verify via Speaker 1's public financial disclosures or social media posts detailing his AI tooling costs; cross-reference with Anthropic's pricing page for Claude Pro. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Meta's Muse is free or low-cost, while Perplexity's InStinct charges for usage.

Unidentified speaker (speaker_0) ·58:31

How to verify: Check Meta's official Muse pricing page and Perplexity's InStinct pricing page for current subscription models. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Speaker 1's team has reduced from 20 humans to 3 humans and 21 agents.

Unidentified speaker (speaker_1) ·1:05:37

How to verify: Verify via Speaker 1's public statements or company filings (if public) regarding headcount; check for public announcements about team restructuring at Scale AI or related entities. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

SpaceX's market cap is driven by compute (Starlink) rather than rocket launches.

Unidentified speaker (speaker_3) ·1:07:14

How to verify: Analyze SpaceX's S-1 filing or investor presentations for revenue breakdowns between launch services and Starlink; compare with market cap data from financial platforms. For recent events, verify 2026-08-19 through 2026-11-19. Use another period only when the speaker explicitly names it.

Primary source

Do your own research

  • Verify NVIDIA's S-1 filing and recent SEC disclosures regarding Jensen Huang's status and any ongoing DOJ or Commerce Department investigations into export compliance.
  • Check Anthropic's IPO prospectus for revenue figures, employee count, and the specific terms of the $1.4 trillion valuation to assess the 'efficiency' claims.
  • Review Carta's Q3 2026 Venture Market Report to confirm the 21% drop in deal volume and 2.5x increase in capital deployed.
  • Analyze HubSpot's latest earnings call transcript and quarterly reports to validate the decline in net new customer acquisition from 10,000 to under 5,000 per quarter.
  • Investigate Kalshi's regulatory filings and state-level gambling/commodity exchange licenses to assess the 'regulatory risk' cited for its $65B valuation.
  • Examine Meta's Muse documentation and pricing models to compare costs against Anthropic's Claude Pro ($1,000/month) and assess the 'workflow lock-in' claims.
  • Search for recent news on SpaceX's Starlink revenue breakdown to determine if the 'compute business' is indeed the primary driver of valuation separate from aerospace.
  • Look for counter-arguments from SaaS incumbents like Salesforce or HubSpot regarding the adoption rates and limitations of AI agents in enterprise environments.
Full transcriptexpand
0:00
Unidentified speaker (speaker_0): If I own 100% of the stock of this company, the person on the planet I would most want to run this company is Jensen.
0:06
Unidentified speaker (speaker_1): But the company's been maimed. The whole ecosystem's been maimed. I
0:08
Unidentified speaker (UNKNOWN): mean,
0:08
Unidentified speaker (speaker_0): the winner here, of course, is Anthropic.
0:10
Unidentified speaker (speaker_1): I think the liquidation preference has been weakened by the NVIDIA drama.
0:13
Unidentified speaker (speaker_0): You buy SpaceX the stock entirely on the basis of what you think Entropic will be paying you one year from now for that compute.
0:21
Unidentified speaker (speaker_1): LPs are going gonna to grade us all on how many 20 billion and up portfolio companies we
0:25
Unidentified speaker (UNKNOWN): have.
0:25
Harry Stebbings: This is 20 VC with. me, Harry Stebbings. We're joined in the hot seat today by Rory O'Driscoll, Jason Lemkin breaking down the biggest news in tech. Wow, what a week of news we have. The biggest IPO ever, dear Lord Dario coming out with anthropic $1.4 trillion, 7x oversubscribed. We have Jensen Huang. Will he
0:45
Harry Stebbings: come back? Will the board give in? We shall see Microsoft outage. outlook, teams, copilot down. Their users probably smiling. I'm gonna get in trouble for that one. And then SpaceX catches Starship. Incredible! What an innovation! Elon Musk stuns the
1:05
Harry Stebbings: world again. This and so much more in the show this week. But before we dive into the show today, Base 44 is one of those products I used, and I just thought straight away, man, I wish. I'd had this when I started 20 VC with $50 and no contacts in Venture. I wanted to see what it could do. So I built a guest research hub for the show. Upcoming guests,
1:25
Harry Stebbings: company notes, previous interviews, sponsor contacts, follow-ups, all in one amazing location. I described it in plain English. A few minutes later, I had the product open in front of me. Database, login, backend, and hosting already handled, and just all in one. It made it so easy. And that's what I loved. I didn't write a long spec. I didn't stitch five different tools together.
1:45
Harry Stebbings: I built the thing and I started using it. For any founder listening, that's the point. The faster you turn an idea into something real, the faster you launch. Apps, dashboards, internal systems, AI agents, whatever you need to build, base base44 44 gets you there. Go to base44.com forward slash 20 20vc. VC. That's base44.com slash
2:05
Harry Stebbings: 20vc. While base44 helps you build, AlphaSense helps you find answers really, really fast. Quick one for anyone doing diligence in venture right now. AlphaSense just launched Super Analyst, and it's one of the more genuinely useful AI products that I've seen come out of this space. It's an always-on always on agent that works like a power user inside
2:25
Harry Stebbings: AlphaSense, but on your behalf. So check this out it pulls from over 500 million premium sources, covering both public and private companies. That's filings, earning calls, broker research, expert transcripts, and it doesn't stop when you close your laptop, point it at a sector, a company, a thesis you're tracking, and it keeps working in the background.
2:45
Harry Stebbings: And here's the part that I really care about it remembers your process, it learns your templates, your firm's logic, and it applies that to every new task. So it compounds, it gets better the more you use it. Isn't that the value? And every single output links back to its source. So there's no black box, there's no trust me. And you can see exactly where something came from.
3:05
Harry Stebbings: So you have complete traceability. Pick the deal memo or the market map you've been putting off and hand it over to Super Superanalyst Analyst instead. Start your free trial at AlphaHyphen alpha-sense.com Sense.
3:15
Harry Stebbings: forward slash 20 VC. That's alpha-sense.com forward slash 20vc. While Alpha Sense finds insights, Intercom helps you serve customers. As AI agents become more common in customer experience, teams often end up juggling multiple silo tools for every job. Well, Finn Fin was built to change that. It's a
3:35
Harry Stebbings: single unified agent that works across your entire customer experience. from service to sales to success and beyond. Fin is the agent making perfect customer experiences possible for thousands of customers. It's powered by custom models, trained on years of real customer interactions, so it understands the nuance and complexity of customer service better
3:55
Harry Stebbings: than any other agent. That means faster resolutions, more consistent support, and just better experiences for every customer. It's also designed to be fully self-manageable so you can easily improve and adapt it as your business evolves. No third parties required. Leading companies like Gamma, Asana, DoorDash, and Crypto.com already use and love fin
4:15
Harry Stebbings: to deliver better customer experiences. So for a limited time, you can get $500 a month in Fin credits. For your first three months, learn more at fin.ai forward slash two zero vc.
4:26
Unidentified speaker (speaker_3): You have now arrived at your destination.
4:29
Harry Stebbings: Boys, we have more on this week than I think almost any other week. We've got the largest IPO in history, I believe. And then we also have one of the greatest founders potentially coming back. I mean, the news just doesn't stop. Janssen Huang, since we last spoke, falls from NVIDIA. Stock loses up to six 601 hundred and one
4:49
Harry Stebbings: billion in a morning.
4:51
Harry Stebbings: And then the biggest investors, Gavin Baker, Altimeter come out with letters, and like 5,000 employees, 200,000 people sign the petition. Let's just start with: do we think it's right that he was let go? What
5:07
Unidentified speaker (speaker_0): do you mean by right?
5:08
Harry Stebbings: If you were on the board, would you have let him go?
5:11
Unidentified speaker (speaker_0): If I were on the board with this guy who'd founded and built a company for 30 years, one of the things I've learned is decisions in haste are often wrong. Finding some way to say, take a leave of absence, we need to figure this out, get to the bottom, slow the whole thing down so you can make a better decision. I look at when I've made dumb decisions and I've made plenty. Usually it's you're reactive, you're
5:31
Unidentified speaker (speaker_0): quick, you don't have time to think and let it settle in. In the same situation as a board member, I'd probably have tried to find a way to make it less a interesting. irrevocable decision and get more time to think. Right? So I actually think that, hey, take a leave of absence, interim, CEO, whatever, and let's figure what's going on here. So that's the first comment. And then just two other comments. Just the second thing
5:51
Unidentified speaker (speaker_0): is this is a messy thread because you have the personal allegations both with a customer and with employee. Then separately you have this emerging dynamic of compliance with export controls and all. the dynamics around that and it kind of getting intermeshed with the administration. So one of the odds I've had here is you don't know what the board knew at
6:11
Unidentified speaker (speaker_0): any particular point in time. And you know, was it just as simple as, oh, I can't believe there's a second affair, I'm done, I've lost trust, or or were these other dynamics coming into play? I just don't know. But again, I go back to more time to figure out what's going on would have been useful.
6:24
Harry Stebbings: Can we just tamper my hopes and expectations? Do we think that he will come back?
6:29
Unidentified speaker (speaker_0): As of this morning, he just made that statement which basically says, if you want me back, I'd love to come back. Proving the one of the old rules, never resign in anger or even in shame. Take your time. So he's now sitting there going, hmm, maybe I shouldn't have quit. But yeah, so that's the statement now. But you asked a separate question, will he come back? I want to say something about the board. Four or five of
6:49
Unidentified speaker (speaker_0): these board members have been on for 30 years. They backed Jensen 30 years ago when he was in his early 30s. I mean, there's two VCs and then there's an audit chair and one other person. It's probably been the single most significant business relationship of all of those board members' life. And I've been on boards for 10 years. I've never been on a board for 30 years, and I've never been on a board for 30 years of a company
7:09
Unidentified speaker (speaker_0): that went from a startup to the most valuable company on the planet. But I know how much I like my CEOs that I've backed. for ten years, and how hard it would be for me to vote to replace them. I can only imagine how hard that must have been. I
7:21
Unidentified speaker (speaker_1): mean, Mark Stevens made four billion dollars off uh the you don't want to fire the guy, and he just sold a billion cash this year. You don't want to
7:27
Unidentified speaker (UNKNOWN): find the money. So you did
7:28
Unidentified speaker (speaker_0): remind me that while they technically did not find them, they also did I think again, you get a lot of leaks here. So I would just want to acknowledge one thing. When four or five people who are widely experienced, have made a ton of money, decide to go against someone that they've really, as I say, been the most important successful business relationship with the right, you kinda really go, there must be something that's really getting on the goat. So
7:47
Unidentified speaker (speaker_0): given that, it's hard to imagine this board saying a week later, you know, upon sober reflection we made a horrible decision. We're going to just vote you back in. That seems a big lift. That's the first comment. Then technically you can have a shareholder vote to replace the board. and then the new board can do. So it's technically possible. I think my guess is at the
8:07
Unidentified speaker (speaker_0): margin. The thing uh you don't know and I don't know is separate from all the relationship questions, these whole Department of Commerce, Department of Justice comments on NVIDIA's actions in terms of selling chips to Indonesia that may well end up in to Singapore, that may well end up in China. My guess is until that thing is run to ground, I mean the Department of the
8:26
Unidentified speaker (UNKNOWN): United States.
8:27
Unidentified speaker (speaker_0): Justice and the Department of Commerce issued a statement on this just in this week, which is interesting. So there's that pending issue, which wasn't a crystallizing issue for the resignation, obviously the relationships were, but it's kind of out there. It's hard to imagine a board saying, Oh, let's come back and fight this one. So my gut would be, in the absence of a board change, no. I mean, I could be wrong, Too long, but my gut would be no. And I
8:47
Unidentified speaker (speaker_0): could look like an idiot in two days.
8:49
Unidentified speaker (speaker_1): I'll tell you why I think there's a high chance he comes. back. And
8:51
Unidentified speaker (speaker_0): by this week?
8:52
Unidentified speaker (speaker_1): Maybe, maybe. Because and this week is fast, right? He shouldn't come back this week. It should take a couple of weeks to happen, but he may. I'll tell you why. And listen, we're we're all getting facts and these leaks really bother me. This is the most leaky ship I we've ever observed. It's horrific. It's it's disgusting. But I think what's changed since then is that I don't think there's gonna be a huge issue of of the spy who loved him, right?
9:12
Unidentified speaker (speaker_1): I think that's fading. Maybe not completely because. the facts are actually getting worse, but I think it's gonna fade away. I think we're gonna say he didn't know she was a spy, let's move on. And I think this issue, because it's already happened, it's already been banned. This issue of allowing data centers adjacent to China to use uh top of the line chips and and shipping and trans shipping, breaking open the servers and then
9:33
Unidentified speaker (speaker_1): driving the the chips off the border, this has already become a massive issue this week. Okay, massive issue. This will reverberate. for a long, long time. And it may be 20% of NVIDIA's total revenue if you count China. The problem is if Jensen's guilty, so is the CFO signed the financial statements. So is the head of supply and engagement, so is the CTO. So I think you gotta bring them back because otherwise you you gotta
9:53
Unidentified speaker (speaker_1): fire all of them. They're all guilty. They all set up these data centers.
9:56
Unidentified speaker (speaker_0): And it's funny you should say that, because actually this the the CFO actually being the CEO now is interesting because she she's gonna be on the hook. Is she gonna reaffirm the guidance she gave a month ago? Is she gonna say all the accounting that I booked the CFO is still good when I'm the CEO? And it it's very hard to walk it back and say, but it was it was me, but I've changed my mind, you know, gents and
10:16
Unidentified speaker (speaker_0): so yes, there's complexity there.
10:18
Unidentified speaker (speaker_1): How can she say it wasn't me when she signs the financials?
10:21
Unidentified speaker (UNKNOWN): No, I agree.
10:21
Unidentified speaker (speaker_1): And she it's her job more to know where every chip goes than his. He's the guy out eating eating chicken and schmoozing. Her job is to know where every GPU is. No, seriously, she can't say it wasn't me.
10:32
Unidentified speaker (speaker_0): Well it's interesting because the board has now decided it's their job too, because they've formed this board committee on importation, exportation of chips. Some basically some board level risk supervision, which exudes to me that they think that's the big tail risk, right? That that something happened deliberately or accidentally, you have some kind of breach of whatever you know, ambiguous litigation we have around
10:52
Unidentified speaker (speaker_0): these issues. So that's clearly being the dominant issue now. And just one other comment here, which is, Harry, you asked the question, will he I mean there's the two questions, will he come back? Should he come back? They make this clear statement. If I own a hundred percent of the stock of this company, in the absence of the US government trying to put my CEO in prison, the person on the planet I would most want to run this company is Jensen. And I'd want to
11:12
Unidentified speaker (speaker_0): make everything else go away. The question is can you, with the morass of being public? If he were to come
11:18
Unidentified speaker (UNKNOWN): back,
11:19
Harry Stebbings: what price bump would the stock get? It
11:21
Unidentified speaker (speaker_0): wouldn't go back to where it was.
11:23
Harry Stebbings: Really? No. Oh, I think so. But
11:26
Unidentified speaker (speaker_1): the company's been maimed. The whole ecosystem's been maimed. It will never be the same company.
11:30
Unidentified speaker (speaker_0): You've got two things. The first question would be: what percentage of your revenue really was coming in directly from China? And it's worth pointing out, like 10 years ago, back before it met. mattered, I think China was a 15% plus or minus of the business. So it's a good slug of the business goes away. So if 15% of your business goes away, do you go down by 15%? Maybe. I don't know. And then the second thing is, and this is a weird
11:50
Unidentified speaker (speaker_0): phenomenon, you you you see it. Is that he's been the master banker of the whole AI boom? And everyone's been, yeah, yeah, that's great, that's amazing. And then the way humans are, they're all peop people are awful. The minute he's out of the picture, then everyone is dumping on those data. deals and saying those deals are stupid and they're all gonna blow up. The question is if he comes back, can you kind of go back to doing those deals? I don't think so.
12:10
Unidentified speaker (speaker_0): I think you would not be able to use capital as aggressively as he has been, and that would have some impact at the margin on revenue.
12:18
Harry Stebbings: Boys, is it not fundamentally so bad for the venture ecosystem not having Jensen Jansen losing his aggression
12:25
Unidentified speaker (UNKNOWN): in direct deals
12:26
Harry Stebbings: in a Yeah.
12:27
Unidentified speaker (speaker_0): It sucks. I mean just start with the very basic comment. I mean the winner here, of course, is Anthropic. The biggest single thing that Jansen was pushing, the thing that was most significant for I think the ecosystem was with his post-out acquisition and hugging face, an open way US alternative was his big agenda. And it takes real money to build these models. I mean, you know, you look at the statement in
12:47
Unidentified speaker (speaker_0): the S1 S one that it took, I think, ten billion to train the last anthropic model. No one else has the dollars to build an open source product. So if he is not able to come back and that agenda doesn't get pursued, you don't end up with a U.S. open source open weight product of significance. Second comment is given the whole China dynamic, you probably end up with more pushback on using, as Jas
13:07
Unidentified speaker (speaker_0): Jason pointed out two weeks ago, Chinese open weight models. The winner is the closed weight provider in the U.S. who's now got pricing power. Apparently there's been leaks and rumors of senior anthropic people smiling happily about this because it is just good for them. They're not big NVIDIA customers, and OpenAI is, so their competitor is weakened. And the provider of the cheap
13:26
Unidentified speaker (speaker_0): alternative product has been weakened as well. This is a good day.
13:32
Unidentified speaker (speaker_1): But to Rory's point, I think what NVIDIA did was perfectly legal. It was okay to to set up data centers outside of the
13:38
Unidentified speaker (UNKNOWN): U.S.
13:39
Unidentified speaker (speaker_1): China that was used essentially remotely by Alibaba and friends. It was completely legal, and as of now it's not, right? Apparently it's no longer is. It was legal. But when you're anthropic and and you're feeling distilled and ripped off, that seems like the crappiest loophole on planet Earth. How can they get away with this? You're just stealing our IP, you're stealing our hard work by building this data
13:58
Unidentified speaker (speaker_1): center a couple a couple hundred miles away from our biggest competitors. It's uh it's epic. I don't know that Dario is gloating, but man, it's it's a boost.
14:07
Harry Stebbings: Speaking of a good day for anthropic, it's about to be their big day. The special day that we've all been waiting for. And again, we're slightly before what will be the news item of the day, so to speak. Seven X oversubscribed, some numbers reporting up to 10x oversubscribed for the IPO, up to a 1.4 to 1.5 trillion dollar price.
14:27
Harry Stebbings: largest IPO ever. How's this gonna go?
14:30
Unidentified speaker (speaker_1): Can I bring up my one big takeaway from it for what it's worth? It it has nothing to do with the 1.5 trillion we've talked about. The oversubscribed book is mildly interesting. Rory can give us our thoughts. I know we knew this, but I didn't realize it till I wrote it up this morning. At 100 billion in revenue, which is where Anthropic will be B by Q1 or whatever at run rate, right? Google had at 81,000 employees. At 100
14:50
Unidentified speaker (speaker_1): billion in revenue. Meta had 71,000. At 100 billion in revenue, Anthropic will have less than 6,000. Now we've talked atomically on the show, and many folks have, oh, that's no big deal. It's it's uh inference and AI costs, and we're always 10 billion instead of humans. That's why. This is like not a SaaS model. It's not the margins are different, blah, blah, blah, blah, blah. All true, but
15:10
Unidentified speaker (speaker_1): let's just think about the human implications of that. The next hundred billion dollar company, yes, it got there fast. faster than ever. Yes, it's the greatest venture return of all time, potentially that. But it got there with a twelfth the employees of Google and Meta. A twelfth. Not like 30% less. Like the average go-to-market team, according to iconic data, is 30% smaller with AI. Like this is all
15:29
Unidentified speaker (speaker_1): the Harveys and Schmarveys and Tarvees and Sierra's about 30% smaller. Some are a bit smaller. But a twelfth the size, I mean, uh to me, this says John. jobs are even worse than than what we've been discussing. A twelfth. They got to a hundred billion with a twelfth the employees. Who cares at the end of the day, you know, as long as it's fundable, whether it's GPUs or BPUs or
15:49
Unidentified speaker (speaker_1): EPUs, that that is mind-bogglingly different for tech and our economy to get there with a twelfth the humans. I I don't I I don't mean to be one of the folks that is anti AI, but it is worth reflecting on in the moment when when we have ha when the average employee is making forty one million and and Ashton Kutcher is making more than I mean, good God, six hundred million, I want to quit venture. But this employees thing to me
16:09
Unidentified speaker (speaker_1): for a moment is the most important thing in all of this, because we've already used the products.
16:13
Unidentified speaker (speaker_0): Yes, this is a way more compute leverage, human non leverage play. You're exactly right. Which is why the humans who are there make more money because there's less of them. So you can share roughly the same part pot among a smaller number of people. Yeah,
16:24
Unidentified speaker (speaker_1): it turned out not to be that crazy what they're paying, right? four hundred $400,000 three. for uh for an office manager. It's a
16:28
Unidentified speaker (UNKNOWN): lot of you want the small number of people you have to
16:30
Unidentified speaker (speaker_0): be really good. I'm not sure it's as disastrous for labor as you think. It's not like every company is going to be ninety percent more efficient. I think these are the guys who can leverage compute the best. And I think you want to get lost in the weeds here. I don't think most old companies, like I don't think an old school consulting company is going to end up with five dollars or seven dollars in inference for every dollar in labor. And
16:50
Unidentified speaker (speaker_0): we've discussed this a lot. You know, that ratio will vary between old school companies, software companies, and then companies like Anthropic, and they'll be on the leading edge of it. And so I think the the companies that are net additive will be compute intensive. So I don't think it'll be as grim for jobs as you think, Jason. But nonetheless, it is widely noteworthy. Aaron
17:08
Unidentified speaker (UNKNOWN): Powell
17:08
Harry Stebbings: Did you see the rumored forty percent layoffs of service now today?
17:11
Unidentified speaker (speaker_0): I couldn't figure out I just saw it, I couldn't figure out if it's a rumor rumor. You know, is it real?
17:15
Unidentified speaker (speaker_1): I think it's true. I think it's a big deal. I think it is more telling than you just to challenge you a little bit. And even segments that that appear to be very human-centric, they're going through a lot of changes. Front offices, back offices are being automated, food deliveries being automated. I mean, you know, 80,000 to 6,000. I think this is the biggest disruption of our lifetimes beyond
17:35
Unidentified speaker (speaker_1): AI itself, right? Which is bigger, but my God.
17:38
Harry Stebbings: Guys, does the anthropic IPO change venture and our world in significant ways? For me, yeah, I see LPs much more willing to redeploy cash that they're going to get from it. They're feeling flush in some cases. I see a lot of anthropic early employees be angels in UK based companies. Any other ways where you're like, this
17:58
Harry Stebbings: has a meaningful knock on effect? Well,
18:00
Unidentified speaker (speaker_0): of course it will. I mean, depending on where it ranks versus. SpaceX, it's the biggest or second biggest IPO by dollars raised. More importantly, it happened in five years, not twenty. I mean, Founders Fund invested in SpaceX in 07 or 08. It was founded in 02, so it's a good twenty-six plus years ago. This was founded five years ago and got to a trillion dollars. It just massively raises
18:20
Unidentified speaker (speaker_0): the bar of what is even vaguely possible. You know, the people who did it earlier, you know, all credit to Spark in particular. and then Menlo in the follow-on around, are gonna make out like bandits. And as Venke said in that great post of his, they're gonna be playing with the house money for a long time to come. And congratulate them. You know, well done, great decision, well executed, well written. They're gonna make out like bandits. And the rest of us are gonna go,
18:40
Unidentified speaker (speaker_0): wow, I'm working really hard for one-tenth or one-twentieth of what you're gonna make. And that's just the nature of an extreme power law business it can get in your head. Because you can work really hard and you'll be
18:50
Unidentified speaker (UNKNOWN): like,
18:50
Unidentified speaker (speaker_0): wow. why anyone else's standards we're doing great, but that's the single biggest win in history and one of the and the fastest. So yeah, it's gonna raise the bar of what's possible. I mean you can see it even in the market right now. I think what it does is LP willingness to deploy and then GP willingness to take bets. You know, I mean everyone's like that bet worked. We should do more of them.
19:09
Unidentified speaker (UNKNOWN): I
19:09
Unidentified speaker (speaker_1): think, you
19:09
Unidentified speaker (UNKNOWN): know,
19:09
Unidentified speaker (speaker_1): I was just looking over the weekend at Carter's Carta's Q3 data, which um, you know, since uh what's his name went to open router. doesn't get as much attention as it used to, Peter Peter, right? Since he was so great at the data, I didn't even see anyone talk about it. But I I looked at it over the weekend. And this is obvious, but to answer Harry's question, so Q3 2026, the most recent data they have, deals down twenty-one percent, dollars
19:29
Unidentified speaker (speaker_1): up two point five X. To answer Harry's question. So deals are gonna keep going down or plateau, and the dollars will keep going up. That is just a radically different world than when we met. and maybe even radically different than when this podcast started. And everything else, I mean, it's just a derivative of what you talked about. It's a derivative of tripling and quadrupling down on winners. It's a derivative of AI spend, but fewer rounds,
19:49
Unidentified speaker (speaker_1): 2.5x the cash. That's the world today, and that's the risk GPs are being paid to take and are are happy to take.
19:55
Harry Stebbings: Weird new world of venture.
19:56
Unidentified speaker (speaker_1): It's so different, right? I know you're going gonna to be graded Uh the LPs are going gonna to grade us all on how many 20 billion and up port portfolio companies we have. That's how they're gonna look at our portfolios. How many 20 billion and up? When I started, it was unicorns, right? I don't know what it was when Rory started, but you had like one unicorn in your portfolio. LPs would seek you out. Like not the best ones, okay? But folks come out of the
20:16
Unidentified speaker (speaker_1): woodwork. I'm doing a unicorn analysis on a Google search. I found you have two. Are you are you are you raising? Now it's gonna be 20. How many 20 billion and up startups do you have that ideally you invested at before a $20 billion valuation? That would help, right? I think that's that's how LPs are gonna grade GPs. How many $20 billion and up to New World?
20:34
Unidentified speaker (speaker_0): The funds who got in early on in Anthropic, they're gonna do amazing. They're probably gonna have eight or nine X funds on big dollars. The amazing thing is the median fund in 1996 had a 6x DPI because they were all just doing seed and a and then the companies went public. And it wasn't even a UN. unicorn. You put something in a 20 million pre, it went public at 500, you had a
20:54
Unidentified speaker (speaker_0): 20X, life was great, right? So the odd thing is the multiple isn't that different than the two bubbles ago, but obviously the dollars involved are huge. If there's only 10, $20 billion outcomes, then the earlier you go BA seed, the probability of you landing on one of those 10 or 20 outcomes logically just has to go down, especially
21:14
Unidentified speaker (speaker_0): with lots of people there, right? It's why you could argue that the structure of having early and growth is a very good structure for this environment, because it is wort noteworthy that anthropic, no one got into anthropic from the venture side before five, no one meaningful before 5 billion when Spark did it still pre-revenue. From the growth fund. So it's just interesting, if you're going to be graded on getting $20 billion
21:34
Unidentified speaker (speaker_0): outcomes, and Andreessen Andreason have been very clear on this, then you just got to do whatever it takes to get those $20 billion outcomes. And if you can get some of them at the A, that's freaking amazing. But if the only way you get some of them's at the C, well, that's fine too. Especially if the non-20 billion dollar outcomes can't get liquidity in the same fashion.
21:48
Harry Stebbings: I think one shift we're seeing, we mentioned playing with the house money, is all of the mega funds raised in Q126. And I thought, great, you know, we'll raise We're raising in 27. And I was like, great, we'll raise it in 27. We'll be out of that their cycle. Fine. Every single fund. is queuing up out the door for Jan 1st. We have to get the bag.
22:08
Harry Stebbings: Every mega platform is raising in Q1 for what will be the biggest funds that they raise. It's a 12-month deployment cycle before they raise from last time. How do we think about this?
22:19
Unidentified speaker (speaker_0): Generally, when you when you raise money quickly, you tend to not make as much money. This is that stage in the cycle. It's worth remembering, you know. the cleverest checks in Anthropic were written in early twenty three and early twenty-four, when the world was very different. I mean, if you go back to where you were then, it was much people were much more scared. I think in general, I think now people are going to
22:39
Unidentified speaker (speaker_0): put money out with high velocity. It's not clear to me you'll end up with anything like the same return. Because I think that's just always the case. So but yes, that's what's going to happen because people are going to people.
22:48
Unidentified speaker (speaker_1): I think to deploy it that quickly also inherently fuels the rapid markups that we're seeing. And listen, the the best AI companies grow so quickly. Does it justify going from 10 to 20 to 40 in six months? Sure, on a spreadsheet, of course it does, right? But maybe the pace of deployment also self-justifies it. And if it does, to Rory's point, it will compress
23:07
Unidentified speaker (speaker_1): returns. Maybe it's fine, right? Maybe 8x compressing to 4X or to 3X is fine. But I it seems to be potentially partial fueling that, right? It can't all be the revenue growth since last month.
23:18
Harry Stebbings: So we have investment committee time, Jason, in the show me
23:21
Unidentified speaker (UNKNOWN): the
23:21
Harry Stebbings: money vein. We have three deals on the table today. We have Calchi, the prediction marketplace that's taken the US market by storm, $65 billion price. Not cheap. We have McCaw raising at it $20 billion, but one of the leaders in the data space. And then we have Fractile,
23:41
Harry Stebbings: $720 million at a $7.5 billion price, producing inference chips that compete against the likes of Grok, Cerebrus, and others. Yeah. Where are we going? Well,
23:52
Unidentified speaker (speaker_1): look, I'm disappointed at what you've brought to IC this week. We're gonna do a deal because we're going back to market in January. We have no choice. We're gonna do a deal. But last week I love. loved. Last week we had zipline, right? A little expensive. but Rory and I love Zipline. Who else did we have last week? We had Foul, right? I mean, you can't argue with its success. And who was the third one we had last week? I think you picked the third one, Harry. It was a great, great
24:12
Unidentified speaker (speaker_1): group of three, right? I know we're only able to deploy a few hundred million into that.
24:17
Unidentified speaker (speaker_1): Yeah, Revolut. I mean, come on. I mean, Revolut is just a question of the scale of the outcome. That was a great, that was a that was a fun one. This week, I don't like this one. Cal Calchi, she, I don't like the regulatory risk at $65 billion. I I think this is huridic. We may make money, but I'm still uncomfortable with the lawsuits and the other issues. We'll see. I'd rather stay low. Fractile, no revenue, the 11,000th
24:37
Unidentified speaker (speaker_1): inference chip startup. I mean, when is it gonna end, boys? When is it gonna end at seven billion? Who the hell knows what this thing is? And Merako, listen, I I like Brandon, right? I mean, he he's he's an old soul at 20. And this is a great space because. you know, anthropic and everybody else doesn't want to do this work, but they're still doing the dirty work. And I just can't
24:56
Unidentified speaker (speaker_1): predict the future of dirty work. So Mercur is one of those ones where I'm jealous of the guys that did the last round. You know, I'm jealous of the guys that did it at 10 or 5 or 2. I'm gonna pick it because we've got to put money to work. It's November 17th, seventeenth, but I like last week a lot better.
25:11
Harry Stebbings: Can I explain to you then why you're the associate and I'm the GP at this firm? Because Cow
25:16
Unidentified speaker (UNKNOWN): She's the unwavering.
25:17
Harry Stebbings: Because Cow Shee's the unwavering market leader in what is a monstrous sports betting market in the US. You have states that are gonna get addicted to this revenue so much that they won't allow for it to be banned in a new administration. So that's a pretty good place to be. We can see a three 3X X here. McCaw, McCor, it's not
25:37
Harry Stebbings: impossible to see a ten $10 billion. span from the LLM providers on the data that McCaw sells. If you believe that everyone's going to own their own intelligence, they're going to buy more data for it. It's an even bigger market. That's another $10 billion unlocked. That's a $100 billion business. So at 20, it looks pretty good. I agree with you. And Fractile, dude, they've got a $250 $50
25:57
Harry Stebbings: million order from Anthropic. You've got NVIDIA that are being weakened. You have Cerebrus, which some say is in trouble, some say is booming. TBD, Andrew's amazing. I love him, and I buy him and Cerebrus, so hopefully it's great. But dude, we're gonna see the specialization of chips, and it's a seven point five billion dollar entry.
26:17
Harry Stebbings: Come on. Which side are you on, Roy?
26:19
Unidentified speaker (speaker_0): As you'd expect, in the late stage business, you have to assume that every company is good. Because if they're not good, they wouldn't be raising late stage money. You know, the universe is not totally idiotic. As I said, at the stage we invest, that does business risk up the wazoo at the series A and B, right? By the time you get to these companies, you should be done with business risk. You're like you're dealing with typically a market leader, a number one or number
26:39
Unidentified speaker (speaker_0): two, a big ass market. And then the only question is value relative to TAM. So it's not surprising all three companies are great because they've all priced recently at good prices. So you're gonna got drill drilled down one level from it's not enough to say they're amazing companies. Harry. You've got to in each case get to grips with value.
27:06
Unidentified speaker (speaker_0): So there's the fast inference public company that shipped the chip, that's you know in production, that's booked a mil billion in revenue this year, and they're at valued at twenty-eight. Would I buy that over fractile if I could do public or private? I don't know. I don't have a good enough sense on the fractile edge. You hear amazing things, but I just look at value relative to TAM and relative to comps, and I go, I don't
27:26
Unidentified speaker (UNKNOWN): know. I
27:26
Harry Stebbings: mean I mean fractile is not the value play. Let's just be clear. No,
27:29
Unidentified speaker (UNKNOWN): no, it's not the value play.
27:30
Unidentified speaker (speaker_0): No, yeah, yeah, yeah, yeah. I mean, Mercury, I think Jason said it right. You this is one of those markets where you continually go, oh, I don't know, I don't know, and then it just they still spend more money. I mean, the amazing thing if you read the Anthropic S1 is 75% of their compute, and then they they also I think somewhat disclose their training budget. It's freaking huge. 70%, 75% of their compute
27:50
Unidentified speaker (speaker_0): goes to making the next model, right? It's something like 10 billion. So they're each gonna spend a billion, billion something on training data. So it you're right, you you have bad news, you have five or ten customers, good news they each spend a billion dollars. It's good for Tom. time, I don't know what it means for value. And then calcium, reluctantly, the great thing there is sports gambling is wildly addictive. The predictive
28:10
Unidentified speaker (speaker_0): market stuff, they they are pushing more of that now. But that's one where that could be huge. I really would want to sit down and understand. I mean, you say the market is huge and it's very big, but you know, you're already at sixty-five billion. They're doing five billion in revenue, so they're growing like a weed. But I suppose my question there would be: what's a sober assessment of the real TAM? You know, how many bored
28:30
Unidentified speaker (speaker_0): 25-year-old dudes are they betting on sports? How big can it be, assuming full legalization? And then can you get prediction markets on top? So I'd probably do the work on that. Aaron
28:38
Unidentified speaker (UNKNOWN): Powell
28:39
Harry Stebbings: I think we're all aligned that we're doing McCor, actually. No,
28:42
Unidentified speaker (speaker_0): I think Jason said it right. It's like the it it was cute. It's the I'd always want to do the last round because you're like, oh wow, they pull it off again. But yes, great business. And they appear to be the winner. I mean, it's not really a winner-take-all business. You've got surge, you've got scale, but these guys appear to have pushed their way in in a very impressive fashion.
28:59
Harry Stebbings: They've done they've done amazingly well. Their proximity to the labs, I think, is really very, very cool to their success. And Brendan's amazing.
29:05
Unidentified speaker (speaker_1): I
29:05
Unidentified speaker (UNKNOWN): think
29:05
Unidentified speaker (speaker_1): though we just put in 200 million and move on to next week because is I I like these, but we're gonna have an even more banger group next week. So let's do 200 and move on. We'll own one percent. And you know, even if we don't make that much, putting 200 million to Mercury 20 billion, we'll learn. Even if we don't make that much, it'll be it'll it'll be good to learn. And you know what, you know what we missed on the NVIDIA thing and you asked about it relating to IC? You
29:24
Unidentified speaker (UNKNOWN): know
29:25
Unidentified speaker (speaker_1): what I really think? I think NVIDIA, this whole Jensen thing damaged a lot of things in venture. I think it damaged Harry's hundred percent confidence in the liquidation preference. I mean, what was NVIDIA gonna buy? Reflection? Is that who they're gonna buy for something billion? That deal's dead now. There's some chance Hugging Face won't close. Harry's running around doing these deals at IC, left and right, in
29:44
Unidentified speaker (speaker_1): 12 months, always telling us our liquid. I mean, I'm I'm I'm joking, but not. I I think the liquidation preference has been weakened by the NVIDIA drama. Because NVIDIA would always buy it. NVIDIA friends would always buy it, or backstop it, or come in and bail out Nscale, or do something, or buy out the team, buy out the reflection team, buy out the pool side team, buy out the hugging face team. I can't
30:04
Unidentified speaker (speaker_1): lose on these deals because of my liquidation preference. Well, with Jensen gone, maybe you can lose on these deals. I don't think the CFO is gonna buy all these kids and bail out their pool sides and tool sides. I don't think she's gonna do it. And I don't think when they hire some retired guy from wherever, from AMSL or wherever, he's gonna go buy all these kids for 17 billion. I don't think he's gonna buy Clem for 15 billion
30:24
Unidentified speaker (speaker_1): to win open source. I think of these liquidation preferences. are are gonna be gonna they're not gonna be worth the paper they're e-signed on, Harry.
30:31
Harry Stebbings: Dude, I agree, but the hugging phase acquisition, like perfect timing. Like
30:36
Unidentified speaker (speaker_1): assuming it closes. Assuming it closes. Close to perfect
30:40
Harry Stebbings: timing.
30:40
Unidentified speaker (speaker_1): Perfect timing would be closed. And
30:42
Unidentified speaker (speaker_0): the biggest risk is probably regulatory if we all get caught up in stuff. Okay.
30:46
Unidentified speaker (speaker_1): Yeah. I mean, Elon and Twitter proved it sometimes it can be hard to get out of a deal, but yet there's also many. outs. Do you really think the CFO wants to do hugging face right now? I
30:54
Unidentified speaker (UNKNOWN): mean,
30:54
Unidentified speaker (speaker_1): I don't know if it's the top of the list right now. I don't think it's the the the whole team's super excited about this right now. Aaron
30:59
Unidentified speaker (UNKNOWN): Powell
30:59
Unidentified speaker (speaker_0): I would say it though that the the costs and sp like it's small relative to market caps, all other things have been equal, they'd want to do it. But I think the issue is more if NVIDIA is going to be subsumed in regulatory morass. And there's such as a random comment on the regulatory side. One of the things that's scariest if you're NVIDIA is it would appear on the other side of the table in Washington, there are two different sides on the issue. There
31:19
Unidentified speaker (speaker_0): are the, you know, sell shit to China people, Lutnik and Commerce are doing that. And then you have the don't sell shit to China people, and it seems Marco Ruby and some of the other people in that. So one of the real complexities, you're not just dealing with a government that might be mad at you, you're dealing with a government that might be mad at itself. So you could easily see getting caught up in a bureaucratic swirl for the next two years. And if
31:39
Unidentified speaker (speaker_0): that happens, and then you've been asked to, you know, the department, whatever the the FTC's been asked to approve a, you know, some hugging face acquisition. No, maybe we should just kick this back. Don't want to piss off Marco Marco, don't want to piss off Hauer. You know, you're right. The regulatory risk here definitely went way up. I hope they got a good breakup fee, because they deserve they deserve to be not screwed over by this. If you
31:59
Unidentified speaker (speaker_0): don't take hugging face fancy out of the good guys, you've run out of good guys. Sorry,
32:02
Unidentified speaker (speaker_1): just briefly we could move on. But you could imagine twenty twenty-seven being a banger year in many ways. But if in Jensen isn't there and say Salesforce doesn't buy anybody because they're digesting these deals this year, the world would look very different if N Nvidia isn't doing deals and Salesforce aren't doing deals next year. Okay. It'd be a very different year than our than then you can't lose with our Lick Pref year of 2026.
32:22
Unidentified speaker (speaker_1): It'd be very different. Poolside could be bankrupt. They didn't have they said they didn't have the capital. That's why they sold, right? What was plan B?
32:28
Harry Stebbings: Oh don't be negative, Jason, please.
32:30
Unidentified speaker (UNKNOWN): No, no, I'm not. I mean,
32:32
Unidentified speaker (speaker_1): amazing outcome. But but what was great about poolside is it was a an element of transparency we don't usually see, right? Everyone's full of it, right? This is them saying, listen, we cannot afford to keep this business model going. We've we've done great job, guys, but we can't raise the 10 billion, 20 billion gap. Will thinking machines survive? I just don't know. You know, it could be a very different world.
32:51
Harry Stebbings: We have a friend in the show, Mr. Brian Halligan. Back in the hot seat, baby, HubSpot has, you know, had a challenging 18 months, I think everyone would agree. And they've, you know, categorized the refounding moment, Damash and Brian. It's been awesome to see them get back together. What's the take on this, Jason? I gotta start with you. Did they have to do this? Like, was it like
33:11
Unidentified speaker (speaker_1): Well, look, HubSpot basically for two years, look, it's been under pressure as if many have, right? Stock price fell 50% this year, but let's ignore the up and downs of the stock price. I think Rory will be with me on this one. For years, even when there are other, even during the the start of the SAS pocalypse and the 2024 dramas, they reliably closed 10,000 net new customers a quarter by like
33:31
Unidentified speaker (speaker_1): clockwork. Now those customers had different values at different sizes, 10,000, 10,000, 10,000. Two quarters ago, six 6,000. thousand. One quarter ago, 5,000. Next quarter, they're guiding below 5,000. Monday went from like, which is a peer went from like 1,200, 10 plus net new customers, 10 plus seat net new customers to under 200. Okay. Fell by
33:50
Unidentified speaker (speaker_1): 80%. So something had to change, and it had to change urgently. And the tough question with love is it will work. Okay. We can talk about why it will always work to have the founder energy in the back, the founder ability to make decisions others can't. The founder being more connected to product always works. But man, I'm not saying it's too late, but it's tough. That
34:10
Unidentified speaker (speaker_1): decline has already been in place. And I don't really care, you know, revenue up and down, stock price, who cares? But the thing is, net new customers, this is like when I started investing, this was my lighthouse metric, and we get distracted. It is the only thing that matters at scale. And people hid this for two years with endless price increasing. increases. Endless. If you have high and our decent NRR and
34:30
Unidentified speaker (speaker_1): high and price increases, you can mask 20% growth for several years. And people that don't look closely won't know. But if you don't if you're not attracting net new customers, your business is if it's not terminal, you're into deep value mode, right? You're into D P mode because the Is that there there are no new customers. So he's got to turn he and Darmesh have to turn this around almost instantly or
34:50
Unidentified speaker (speaker_1): otherwise you risk falling below the hard deck of net new customers. And I don't I don't envy them. So it will work. The question is, will it work enough? And everyone's in founder mode. Dev's back at Mongo, who was on the show, right? Back at workday, back at HubSpot, back at UiPath. It's all working. The question is, is it enough? And the last thing I'll say is every
35:10
Unidentified speaker (speaker_1): VC in Yahoo points to Finn and Intercom. Rory was a large investor. I was a time. tiny investor since the early days. I'm not sure Intercom worked. And I say this with love. I don't mean it literally, but let me finish my thought. It definitely worked, okay? Owen came back in 2024, instantly fixed pricing and other issues. That worked before AI. Like immediately his first week, he he fixed
35:30
Unidentified speaker (speaker_1): the endless price increases and customer ripoffs that existed before him. Then he went all in on agents and AI in the very beginning. It worked, right? However, even with all that, even with the incredible blood, sweat, and tears of the founders, staying the whole journey, Owen coming back, they only got it to a hundred milder plus million of agentic revenue. Did it really work? We'll never know as part of Salesforce. We'll
35:50
Unidentified speaker (speaker_1): never get the next card of the story. And every VC that says this is a case study, I think it's an example, but I don't even think it proves the case that this works, that a approved proof goes all the way. We didn't get to see intercom go to a billion of agentic agentich revenue. We never saw the next act of Owen and Friends. And that's okay. People made money. But this was never truly proven that this works,
36:10
Unidentified speaker (speaker_1): that this coming back. We just we just know it works to some extent. And so it's it's it's effing hard. Well,
36:15
Harry Stebbings: that's what was my gonna be my question, which is like what is enough? You said that enough several times. I love Daniel and UiPath, I love Dave. But like, I think it's all TBD. It's our
36:26
Unidentified speaker (speaker_1): are you can you become a growth stock again? Or do you become Zoom or or and Dropbox?
36:31
Harry Stebbings: Well, can you tell me one that has become a growth stock again?
36:34
Unidentified speaker (speaker_1): No. I mean there are there are some that look like it. There's the Octas and Twilios. I mean the growth isn't astronomical, but but the stock prices show so much benefit from a gentic spend.
36:44
Unidentified speaker (speaker_0): Agreed. There's no one above 30% thirty percent in the public markets of that category of pain. pay. Obviously, Palantir is it. it's doubling, but to your point, there's no one from the SaaS era who has reaccelerated back above thirty percent. I kind of agree with your comment, Jason. I I think it's it is really hard to dramatically change the trajectory of a billion dollar revenue
37:04
Unidentified speaker (speaker_0): company. You're probably playing for some version of instead of having only five thousand customers Was going down to two 2,000, thousand, I can get it back up to eight 8,000, thousand, I can do better, I can point the growth rate up to twenty-five 25%, percent. and I can survive and be a successful, viable, interesting public company that matters. But there'll be a five-year period where the startup, the Sierra, growing
37:24
Unidentified speaker (speaker_0): 4X year on year, will have the growth and the sizzle and the cachet. That's just very hard to compete with. So yes, I agree. And it's worth, by the way, one random comment on this going back to the Jensen thing. It's kind of funny. It is worth noting when three or four boards felt their backwards back to was the wall and they have to save the company, they break the glass and get the founder back. Just saying
37:44
Unidentified speaker (speaker_0): to the board, hold that thought. Because it is noticeable for all the faults that founders can bring to the table, it is just worth noting that, and in every case it was the right move. Dines, I think it'll be right with Brian, I think it was definitely right with Owen. Dev makes a ton of sense. It is noteworthy that that's the person you need, that early-on early on person who knows what to do here. But I agree, it is it is hard. especially
38:03
Unidentified speaker (speaker_0): in the apps level. I mean, I think I I don't know where the current Mongo growth rate is. Some of the things like Datadog, if you can co-attach to the infrastructure boom, you can get your growth rate up. But fundamentally, I mean, going back to your comment, Jason, I mean there's two different things. At the app level, if you're selling a SaaS based app, the problem you're up against is it could be that the new thing that the customer
38:23
Unidentified speaker (speaker_0): buys in the AI first world is just so different. that, you know, to a rounding error, all you're doing is you know building a side better version of yesterday. And that's why I thought the I don't know if I love the positioning, but uh Brian Halligan's comment, they're going to be the customer relationship harness, I think is what they said. Which is uh clearly an attempt to position themselves in the new world, which is smart.
38:44
Unidentified speaker (speaker_0): I'm not sure that the average HubSpot HumpSpot customer, and remember we were investors in the series, I think, C back in 2010, so we did very well. I'm not sure the average Humpspot customer of a 50 to 100, 250 person company really has an opinion on harnesses. But clearly what they're doing is trying to position in that kind of new product world. But it's hard. It is really hard because even if your
39:03
Unidentified speaker (speaker_0): new business it goes great and I think the intercom new business was going really well, you're still trapped in this bigger, slower growing thing. See, you know, Wix slash base 44 for details. It's very hard to dramatically change those trajectories.
39:16
Harry Stebbings: I always say to Daniel, I'm like, dude, why would you go back? It's so brutal. Like, you have a nice life now. Like, it's just, oh.
39:26
Unidentified speaker (speaker_0): Man loves to struggle.
39:28
Harry Stebbings: Man loves to struggle. And actually, do you know what I see? I see a level of aggression in all of them where it's like, they've made so much money already. Like it It doesn't matter like it used to. And so they go with just unlimited. aggression, confidence that they didn't have before.
39:42
Unidentified speaker (speaker_1): I think so. I I will say I I think that's true. And and no, it's very hard for an outside CEO to do that, right? It's certainly hard for the CFO, the that you know, promoted to CEO two days ago to do it. I will say I think that that of course that's true. It's a level of aggression that you need, and that's why I think it will work. I think Brian will work. I think they will all work to what Rory and I said. But I do see a hint of
40:02
Unidentified speaker (speaker_1): hesitation. And what I mean, maybe hesitation is the wrong thing. I hear a hint when I talk with a lot of C founder CEOs, not necessarily that went back, but are there or whatever, is they also recognize there may be a limit to Rory's point. 30% may be the best I can do. I already have 320,000 customers at HubSpot. You know, Salesforce hasn't even disclosed its customer account for like four years.
40:22
Unidentified speaker (speaker_1): That's telling, despite being a 50 billion in revenue. They don't, they stopped stop disclosing it. That whenever what someone someone stops disclosing a metric, it's telling one way or the other, right? So Salesforce is clearly a master of selling you five to ten expensive products, but not so much in bringing in net new logos. I think folks, and I'm sure even Daniel, you can ask him, who are off the charts smart.
40:42
Unidentified speaker (speaker_1): I think he also probably has a limit. Okay, I can only do this so so much vis-a-vis agents and LLMs. No matter what I do with UiPath, it's never gonna grow like it did in the past. And there's a certain logical thinking here to try to maximize the reality outcome, upgrade the team, do the layoffs, do whatever. But there's a hint of hesitation that maybe whatever
41:02
Unidentified speaker (speaker_1): limit is as good as it can be, right? After 20 years, 15 years and the you know, 25 years. So I don't know. And
41:08
Unidentified speaker (speaker_0): to some extent, that's not surprising. I mean, one of the things we had said back, I think in 2018 is that we foresaw the SAS SaaS slow. slowdown. And it's not because AI is going to replace it as much. That was the second issue we said at the time. But the more important thing is the TAMS get saturated. People have a CRM, people have a HubSpot. Some of these markets are more finite than one thinks. It it's hard to grow a customer count
41:28
Unidentified speaker (speaker_0): infinitely, especially in the enterprise side. So I'm not surprised at the the Salesforce comments. But
41:32
Unidentified speaker (UNKNOWN): well
41:32
Unidentified speaker (speaker_1): look, that's true, but I would say for what it's worth, there is a mystery to me here. The fact that Monday and HubSpot have seen their net new customer account count decline. so radically over the last three quarters, that is a fundamental change on the way SMBs are buying software today.
41:45
Unidentified speaker (UNKNOWN): Yeah.
41:46
Unidentified speaker (speaker_1): This is radical change. And I don't think it's vibe coding your own software. It's that's ridiculous.
41:51
Unidentified speaker (UNKNOWN): So,
41:51
Harry Stebbings: Jason, who who are the winners here? Is this like Atio and Zero and the New York City?
41:55
Unidentified speaker (UNKNOWN): If you look at the
41:56
Unidentified speaker (speaker_1): ramp data, I guess Attios Adios benefited, but they're so small in the grand scheme of things, it's a rounding error.
42:01
Unidentified speaker (speaker_0): I I don't think so. I think the way winner, candidly, is probably Claude, a topic. In the sense of for small SMBs where the workflow and the auth and all the kind of corporate legacy stuff isn't that important, you can crank something out now that will suffice as a CRM app, right?
42:19
Unidentified speaker (speaker_1): No.
42:20
Unidentified speaker (UNKNOWN): No, no,
42:21
Harry Stebbings: no, no. The people who use Monday are not going in vibe coding with core code, their own CR.
42:27
Unidentified speaker (UNKNOWN): You can do a sub but
42:28
Unidentified speaker (speaker_0): you
42:28
Unidentified speaker (UNKNOWN): know
42:28
Unidentified speaker (speaker_0): look, I was I would have said look, I said now on the show six, nine months ago, and then I have you know, every single associate and principally has cranked out their own front end using Claude for our Salesforce instance, and no one lugs into Salesforce anymore. They've built
42:40
Unidentified speaker (UNKNOWN): their own thing.
42:45
Harry Stebbings: They're not Caesar, the pizza maker.
42:47
Unidentified speaker (speaker_1): I think the answer, if you want my opinion, I think it's in the middle, okay? To Harry's point. I don't think your average SMB that is not tech savvy is vibe coding their own front end to Postgres or Salesforce or I just don't I see no evidence that's happening. I'm pretty close to the repli team. I can tell you CRM is not one of their core verticals, okay, at a billion in revenue. Okay. It's just not. I briefly asked Anton, he never brought it up as
43:07
Unidentified speaker (speaker_1): something that they're doing. I don't think it's a big deal. But here's what I do think is happening to Rory's point. I just think it's subtler. I think folks are just trying not to buy this software for SMBs. They're just saying it's not so much that I'm buying it. They're sitting around like, and in the old days when we started, you'd bring in a VP of sales, and what's the thing he he or she would do their first week? Bring in Salesforce. First week, I gotta roll it out.
43:27
Unidentified speaker (speaker_1): And I got I'd probably have to hire a consulting firm to help us, small, medium, and large. That was a day one kind of thing. I think that you can't argue with the data. I there's not an is Addio benefiting? Yeah, Ramp says a little bit, another 10 million of revenue. That's not. enough to explain what's happening. SMBs are saying, I don't want to buy this crap. I just don't want to buy it. I'm gonna, whether it's a little bit of data analysis in ChatGPT
43:47
Unidentified speaker (speaker_1): without knowing I'm doing data analysis, whether it's running it in Google Sheets for another six months because with Chat GPT it's good enough. It's just in my small business, I don't want to spend 40 grand a year for HubSpot right now in this world. I just don't want to do it, I think is what's happening. Or Monday. And e Monday's not even explaining. expensive, but and this is something I did not anticipate that SMBs would just say, I don't want to
44:07
Unidentified speaker (speaker_1): buy this stuff. On the other hand, I can say my portfolio selling to SMBs, anyone that's automating away front office, back office humans is is exploding. My God, it is the easiest thing. Nothing's easy to sell. But I have three portfolio companies selling five $500 hundred to a $1,000 thousand dollars a month front office, back office agentic solutions that as long as they
44:27
Unidentified speaker (speaker_1): can get a prospect on the phone, they can close it that day. But they don't want to buy these software products. It's not enough. They're just gonna bumble along with what they do. It's not a P1 or a P0. You know what my P0 is? My receptionist quits every Wednesday. So oh, you have an agent that will do that 24-7, reschedule Schedule all the appointments, handle billing,
44:47
Unidentified speaker (speaker_1): handle everything for me, and it actually it actually works $800.
44:51
Unidentified speaker (speaker_1): I'll sign this afternoon, kiddo. That's what people want to buy. Not not Kanban crap. So I didn't anticipate this would happen because we grew up in a world where you couldn't survive without the software, right? You couldn't survive without it. Clearly, people can, and they want to put that energy into things that are different. And that's what's happening. They may buy HubSpot or Monday or someone or Or Adio, Addio, but they may put it off a year or
45:11
Unidentified speaker (speaker_1): two and and and that may end up being never. Like we don't
45:14
Unidentified speaker (UNKNOWN): know.
45:14
Unidentified speaker (speaker_1): That's my theory. because I looked at the ramp day, looked at others. It's not substitution, not sufficient to explain this. They're just done. They want to do stuff that has high ROI. This used to be high ROI software, it isn't today.
45:25
Harry Stebbings: I feel we've got to talk about Microsoft going down for two days. I mean, seriously, my Rory, you'll be surprised that I'm still dating the same girlfriend. She stayed with me short. for She wakes up and she's like, I got no emails. I don't know what
45:38
Unidentified speaker (UNKNOWN): I'm saying.
45:39
Harry Stebbings: to do. Stay. Teams is down, uh co-pilot's down, Outlooks is down, down for 36 hours or so. And then we had uh flights grounded in the US, FAA problems in terms of L3 Harris and one of the contractors.
45:56
Unidentified speaker (speaker_0): And worth pointing out, based on the initial comments, that in neither case is there an allegation of you know severe understanding, it's not a cyber issue, it's not AI cyber, it's not even directly speaking AI. These are old school problems that people just missed. So in fact, you know, in the case of um the FAA, it was a router issue that kicked off the
46:16
Unidentified speaker (speaker_0): whole thing. In the case of Microsoft, I don't fully know, but it appears to be some kind of network configuration thing. I mean, I have seen a post that basically says: it turns out if you whack all the senior engineers and think the AI is going to gonna figure it out, this is the kind of thing you miss. I can't comment on that, I don't know. But they appear to be fundamental classic IT screw-ups that, you know, have
46:35
Unidentified speaker (speaker_0): reverberated quite a lot. So
46:37
Harry Stebbings: But like, you know, the NSA today went down as well for their internal comms and is down now. And I'm just like, hang on a minute. The core systems that run our society, air traffic control, must Microsoft. God, we've never had the fragility of systems in my lifetime, it feels like this. I
46:57
Unidentified speaker (speaker_1): think they've been more fragile, but what here's how I think about it. I think it's DDoSing 2.0 or DDoSing for the AI era. Okay. You could always take down a website back in the day. I mean, websites used to be so fragile, you could DDoS anything. You could hire some nefarious agents offshore, and I could take down Scale's website if I was pissed off, right? It was pretty easy to do because we could bring down that pay. penny, those Pentium servers uh pretty darn
47:17
Unidentified speaker (speaker_1): quickly, right? With agents and other things, we we can find however you want to DDoS it. And this Microsoft thing apparently was a load-related bug that it existed for four months and they hit some sort of maximum load they had in model before and it went down. And I completely agree with Rory as our DevOps team's decline in quality and SecOps team decline
47:37
Unidentified speaker (speaker_1): in quality, i.e., optimizing. optimized, This is gonna happen more. But also, not only is everyone gonna get hacked, like we've talked about for months, and now we're seeing it happen constantly, right? Google said we saw the first truly autonomous agentic hack this week. Everyone's gonna be DDoSed. Everyone's gonna be overloaded their servers. It's too easy. It's too easy to spin up agents, it's too easy to break Microsoft, it's too easy to just break,
47:57
Unidentified speaker (speaker_1): and every system can be broken and is fragile. So it's just another thing we didn't anticipate being able to do, which was attack anything. Was it really just an old level three Linksys router from 1998 that took it out, or was it a load issue, or was it someone probing the edges of a network, right? I think every network's gonna be probe. It's gonna be probed with traffic, with load, with
48:16
Unidentified speaker (speaker_1): hacking. All the vectors are gonna be hit. And so everything's gonna be hit. And it's not all gonna be truly security and stealing credentials and API keys, it's gonna going to be load-based load based too. Anything that can be taken out with massive. load, you're going to spool up agents and you're going to see if you can take it down. Let's take down everything. Iran will spool up a thousand 10,000 agents on Qwen or whatever and see what they can take down.
48:36
Unidentified speaker (speaker_1): It's not that hard. And this is a tough one. This cascading, if you've ever run a B2B service, there's not few things more stressful than cascading web servers because all of a sudden it goes, you have a thousand servers, ten go down, it's really stressful. And then you sort of affect fix it, another 10. ten go down, another now now you spread the load, a thousand loads on nine hundred, eight hundred, seven hundred, six hundred, five hundred, and like you want to
48:56
Unidentified speaker (speaker_1): jump off the roof because it it it gets it gets even as you find the solution, it gets worse and worse. It's gonna happen to everybody.
49:02
Unidentified speaker (speaker_0): And I think even though those two examples weren't at least obviously AI cyber related, we've definitely seen a plethora of other companies wrestling with you know smaller cyber breaches in the last three, six months. that clearly probably are related to just the ability to hack almost anything more quickly. So there's a lot going on. And to your point, I mean it just
49:22
Unidentified speaker (speaker_0): highlights the this, not perhaps P doom or any of that stuff. Th this is one of the big risks in the AI boom, which is when you see what it's like to be with the you know I
49:31
Unidentified speaker (UNKNOWN): mean,
49:31
Unidentified speaker (speaker_0): I I frankly I could live without, you know, Outlook. I didn't even notice and I d I don't love teams. But I think, you know, when you see D FAA go down for a day. or can you imagine a week? I mean, that would start to get economically very painful. You know, when you start to think about that, it does highlight the
49:45
Harry Stebbings: Dude, the weekend was economically very painful. By the way, teams and teams and outlook, I'm with you completely. The devastation that causes for four hours of downtime.
49:55
Unidentified speaker (speaker_0): It's probably going to mean all these companies Is really having to think about, you know, what level of competence is required to deliver these systems. And it's there's probably going to have to be a subject. substantial uptiering in, I want to say the seriousness, the professionalism, the the the capabilities of any company that's offering a kind of mission critical offering in today's world. I remember I remember
50:15
Unidentified speaker (speaker_0): when DocuScience suddenly started to become huge. And I would really go all the board members who had a cyber background, just basically sitting everyone down and saying, look, now you're going to get hacked. This is what you have to do. And they did it. We spent the money to be, you know, as reliable as it takes. And I think all these comp I mean, you can't continue to have the FAA run on a shoestring and a bunch of nuts. I mean, did you see the thing? They just announced they're getting rid of floppy disks.
50:35
Unidentified speaker (speaker_0): Ironically, actually, floppy disks are probably impervious to many of these issues.
50:39
Unidentified speaker (UNKNOWN): They
50:39
Unidentified speaker (speaker_1): are air gapped.
50:40
Unidentified speaker (speaker_0): There is going to have to be a lot of spend on making these work and making these work Work and and durable, because this is not y you we can't be doing this two weeks in a row. Like if the FAA went down again this weekend, I think people would lose their collective shit. And they'd probably, you know, put Sean Duffy's head on a spike.
50:57
Harry Stebbings: It's fucking wild. Have you seen Die Hard 2, where like they're unable to connect air traffic control to planes? That's what was happening.
51:06
Harry Stebbings: Fucking terrifying.
51:07
Unidentified speaker (speaker_0): Yeah. I mean, I read the article on how they got them all. about There's 4,000 planes in the sky, and you have a national wide ground stop.
51:12
Unidentified speaker (speaker_1): I know that one's scary. I just think more tactical. tactically related to venture and startups. I I think in some ways the optimizely thing is worse. I don't mean in terms of global because because Roy made a really important point. On the one hand, we're in the era of can one person build a billion dollar startup, right? On the other hand, we're in an era where I don't think we can trust anything except
51:32
Unidentified speaker (speaker_1): companies with S-tier security teams. And optimizely is a moldy-oldy moldy oldy of B2B, right? And it gets prompt injected or whatever happens. CNN, the BBC, Zoom, everyone's using it, stealing credit cards, stealing PAI. Every mediocre third-party plug-in and app is a security issue. That's the that's what happened to fireworks too, right? And then
51:52
Unidentified speaker (speaker_1): and then and then and then Notion and who else blame fireworks by name, which you never see before, right? Before you'd always see one of our vendors had an issue. It impacted less than 1% of them. Notion sent to their entire million dollar customer base attacking fireworks. And so I'm not picking on them. I mean, I'm a little bit, but
52:12
Unidentified speaker (speaker_1): like where this plays out, on the one hand, we want one person billion dollar startups with no, frankly, no security. Okay. Nothing. Nothing. Have you ever built a real production app yourself? You have no, you know, you can do security audits on the one hand, right? Uh So we want this massive. innovation, but these products are going to be totally insecure. And I don't think you can trust any old product that's a plug in. I just don't think
52:32
Unidentified speaker (speaker_1): you should use any web analytics. I don't know if you should use any new startup payment startup besides Stripe or Ad Yen. It's just just the risk here is so high in the in the automated agenti hacking. Right. And so how that plays out for startups is confusing to me. On the other hand, our s we can scale like never before. On the other hand, your security risk is like never before. Or, you know, unless you're the
52:52
Unidentified speaker (speaker_1): best of the best of the best.
52:54
Harry Stebbings: It goes to the point though that you said, Jason, which is actually the scary thing, is the fireworks team is exceptional. The engineering team is truly best in class. If their security team and Secorps team is 10,000 times better than 99.9%,
53:09
Harry Stebbings: then I really am quite worried. Yeah,
53:11
Unidentified speaker (speaker_1): they're I'm sure they're great. It's their job. And it happened in a corner of their application too. And that's where it often happens. That's what that's what, whether it's a human or AI, you're looking for the edges in the corners. It's not like all of fireworks data was exposed, right? It was this corner for testing where they stored some keys that they found their way into, right? And that's with probably a hundred great 50
53:31
Unidentified speaker (speaker_1): great people in SecOps and DevOps, right? Oh yeah, cursor called them out too. It's pretty bad, right? Well, it just shows the stakes, right? The notion That notion
53:38
Unidentified speaker (UNKNOWN): of the thing.
53:39
Unidentified speaker (speaker_1): cursor and Harvey would call out a vendor. I remember when Harry and I were in uh London about a year ago at Saster AI London a year ago, and OpenAI had some breach with a third-party vendor right while we're there. I can't remember which one it was. I remember they didn't name them. I mean they did cancel the contract.
53:55
Unidentified speaker (speaker_1): I forget who it was, but it was a classic. We've had a small issue. It was only with our CRM and external, it didn't access any of our data, but man, now it's fire. fireworks, Notion Cursor, and Harvey, fireworks, fireworks, fireworks.
54:07
Harry Stebbings: Dude, I think we've just lost all elegance though in this business. Like before, VCs didn't bash each other as much, companies didn't bat. Now it's just like all-out brawl. I do want to touch on one MA that seemingly kind of flies under the radar, but I actually think it's quite interesting. It's OpenAI buying a company called Odesia.
54:27
Harry Stebbings: It's like instinct, but for travel specifically. And it's the founder of Sonder, Francis Davidson. My question that I thought was interesting off the back of that is: does this signal the end of verticalized personal assistance? Your financial personal assistant, your travel personal assistant, and it's just a horizontal play. Any thoughts on this? Well,
54:47
Unidentified speaker (speaker_1): who founded Sonder? Saunder.
54:48
Harry Stebbings: Yeah.
54:48
Unidentified speaker (UNKNOWN): I
54:48
Unidentified speaker (speaker_1): love Sonder. Saunder. It just ended up being a terrible business. model. So I'm happy for him because it was one of my favorite products. Like I use so many right in Covent Garden, I stayed at a three-bedroom in in in Saunder. Every time I was there it was a gem, right? So I'm glad he had a win in the end. Seriously, it makes my heart warm as a founder, right? I'll tell you my one learning I because everyone comes on, you know, we had Jack Altman and others.
55:08
Unidentified speaker (speaker_1): Why aren't you using Use InstinctJason? an instinct, Jason. I'm so deep in Muse now. You don't know all the workflows that we are running across Saster and Muse now. Muse is our X,
55:16
Unidentified speaker (UNKNOWN): our
55:16
Unidentified speaker (speaker_1): horizontal agent. Muse, we're so deep on Muse.
55:19
Unidentified speaker (UNKNOWN): Wow. What do you use it
55:20
Harry Stebbings: for?
55:20
Unidentified speaker (UNKNOWN): Everything.
55:21
Unidentified speaker (speaker_1): We use it for outbound. We use it for competitive review. We use it to I use it to edit all of my content. I use it to create clips and other things. I use it to find all the gaps and everything we're doing. We use it to do analysis of our web traffic, of our Saster annual attendees. It does everything constantly around. the clock. And I have so many workflows in
55:41
Unidentified speaker (speaker_1): it that the burden to use another one is pretty high. Like it is a shallow moat, but man, it's become wide with like 20 or 30 workflows, right? Any single gap we have, and we have twenty we have so many agents at Saster, right? 21 agents for humans. Any gap we can't fill, we spool a muse up and our muses talk to each other. Mine and Amelia's talk to each other. They have their their little web chat channel,
56:01
Unidentified speaker (speaker_1): their email address, they don't tell us, they talk. to each other. So whatever I'm working on and she's working on, they can collaborate together. Our two muses, Carson and Friday, are their names. And any gap, any gap we have, it also anything it reviews every single inbound venture deal I get. It re ranks them automatically. It sends me an update. It reads the ones that I didn't read and it reminds me if there's any good ones that I didn't get
56:21
Unidentified speaker (speaker_1): to and get any attention. It does it every single night.
56:23
Unidentified speaker (UNKNOWN): So
56:23
Harry Stebbings: you've given it full access to your email. Yeah,
56:25
Unidentified speaker (speaker_1): full access for full time.
56:27
Harry Stebbings: Credit cards?
56:28
Unidentified speaker (speaker_1): No, it did order, it did order flowers for a big deal closing on the team without telling me. So maybe that's a thoughtful reminder about the flowers. Um so no credit card yet. But it's so useful that I probably would go to the effort of locking down a Mercury account, like with like a thousand dollar cap if I needed it to buy stuff right now. We don't need it. We have so many other systems for that.
56:48
Unidentified speaker (speaker_1): We have Bill and Brex and others, but I would if we didn't have Bill and Bill
56:52
Unidentified speaker (UNKNOWN): and Brexit.
56:57
Unidentified speaker (speaker_1): What I am worried about for startups and others, as humans, we go out to buy a vendor and we we basically, if we're early adopters, we see what the hell's out there, bless their souls, bless those CIOs with emerging vendor uh groups and stuff like that. Most of us look at the two leading vendors as humans. We do our best we can to evaluate when we buy. one. And maybe over two years we turn over.
57:17
Unidentified speaker (speaker_1): Agents have a different way of doing it. But once one really works, you're not gonna change unless there's a reason to change. It gets locked in. So there is, I don't know if I'm more I think Mark was right to push this at the time he did, because I think January would be much harder than now. He's gonna have a 20 million person install base to build on. And I don't think people wanna want to go
57:37
Unidentified speaker (speaker_1): from Muse to instinct to grok to Dots and back and forth. There'll be folks in tech that do it. But once you have these workflows running and they're really, really good, you don't want to move them. You can, but you don't want you don't want to.
57:48
Harry Stebbings: I'm with you. I have like 50 automations on Instinct running and I'd be a real bull aid to have to change them all.
57:54
Unidentified speaker (speaker_1): You would do it if there's something that you can't do in Instinct an that's important to you. If you for example, Muse is very, very good at at headless browser interactions. So you might find there's something that is so good at headless browser. I'm just saying, I'm not saying this is true of the instinct team, but you might find something that's so good at headless. Like, for example, it can like Muse is the best thing. We still get a million views a month on our WordPress content.
58:14
Unidentified speaker (speaker_1): Nothing else can interact with with WordPress that I found as an agent. Muse does it perfectly. So let's say you had a WordPress, you might do it just for this one one and then give it another, but it's because it could do one thing that is at least a P2 you couldn't get. otherwise. But otherwise you're not going to move off instinct.
58:29
Harry Stebbings: Jason, how much do you think you cost
58:31
Unidentified speaker (speaker_0): exactly
58:31
Harry Stebbings: Zuckerberg? You said four to five bucks last week.
58:35
Unidentified speaker (UNKNOWN): We spend twenty
58:35
Unidentified speaker (speaker_1): and we get whatever the the pro is. How much I don't know, two hundred, five hundred dollars a month, a thousand dollars a month for twenty bucks? I mean I'm sure with Claude with my two hundred dollars I'm spending ten thousand dollars a month where I mean it's it's it's it's
58:46
Unidentified speaker (UNKNOWN): the
58:46
Unidentified speaker (speaker_0): good news for both companies is you're you are not the typical user. Their business model will survive. and the poor dude
58:52
Unidentified speaker (UNKNOWN): just
58:52
Unidentified speaker (speaker_1): maybe, or maybe I'm the future, just in the present.
58:55
Unidentified speaker (speaker_0): Could well be that you think that of yourself.
59:02
Unidentified speaker (UNKNOWN): I mean it
59:03
Unidentified speaker (speaker_1): to learn from. I mean to learn from, right? No,
59:05
Unidentified speaker (speaker_0): I exactly What percentage of people will have that dynamic a set of interactions Actions on an ongoing basis is the question. I
59:11
Unidentified speaker (speaker_1): think what will happen in B2B is you'll wake up, you get the quote from the HubSpot rep, it's $30,000, right? And you wake up and you realize Muse can do it. Not all of it, but more than you thought. And you might put off that purchase a month because Muse, it turns out Muse can go to your email list and email all the 20 VC listeners. It can go and identify, you know what Muse is
59:31
Unidentified speaker (speaker_1): really good at? Go into the scale database into Salesforce, find the top 200 founders we want to invite to a VIP dinner and in cocktails, and invite them in a and it does it. And you're like, why do I need marketing automation software? You might do that without realizing you're replacing marketing automation software because you just asked Muse. I guarantee you, if your Muse is hooked up to any CRM database, it can do just what I said because we've
59:51
Unidentified speaker (speaker_1): already done it. Invite the top 200 COs to 20 VC's office tomorrow night. I'd probably instinct and do it too. And that's that's what you pay $100,000 a year for marketing automation software to do. Except the humans had to figure out the list and pick it and get it wrong. and then invite the failed CEOs
1:00:04
Unidentified speaker (UNKNOWN): on Action.
1:00:05
Unidentified speaker (speaker_0): Look, we don't want to get a the interesting question is it's so funny, Muse, uh the the the the Facebook, the ultimate consumer company, and what you're basically saying is it's powerful enough to effectively be a a small business agent. Right. And the question is, you know, will there be a better small business agent you'll switch to, or is in fact this the correct answer for you? Aaron
1:00:22
Unidentified speaker (UNKNOWN): Powell
1:00:22
Unidentified speaker (speaker_1): I think for me technically, at a pro the interesting thing that people don't get about Muse, okay? It is so. uh over engineered.
1:00:28
Unidentified speaker (speaker_0): Agreed.
1:00:29
Unidentified speaker (speaker_1): It is so overly because of the team building it, because of Alex and all these great guys, right? The best of the best built this, right? They talk about this thing being a little consumer app to make reservations, but it is the ultimate expression of open claw, and it is amazing software. It is the best software I've used since since like Anthropic 3.x got good. It is so over over-engineered engineered for the
1:00:49
Unidentified speaker (speaker_1): use cases in the popular media. It's epic.
1:00:51
Harry Stebbings: Jason, can I ask you a really hard one? But for someone today who doesn't have Muse, what would you tell them would be a great first use to see the power of it, to get started? What would you tell
1:01:03
Unidentified speaker (UNKNOWN): me? It's easy.
1:01:03
Unidentified speaker (speaker_1): It's easy. First of all, Muse is free for it's the most generous free category I know of. So sign up, it's free. If you don't know, ask, ask. If you have any friends that have Facebook, ask them for an invite. It's pretty easy to get. Okay. And here this is. this is the magic. Look around your little business, your little office. What isn't getting done, Harry? In your general business workflows, what isn't getting done? Okay. And just ask Muse
1:01:24
Unidentified speaker (speaker_1): if it can do it. Probably can't. This post didn't get done. This email didn't get done. We're not following up on customers. We're not following up. We had it, we had a dinner last night at scale. Did you follow up with the founders and ask them how I did it? No, I asked my my secretary, but didn't want to. do it because you didn't want to figure it out. Ask Muse. Muse will figure out everyone that came to the scale dinner last night
1:01:44
Unidentified speaker (speaker_1): and it will craft a really good email and ask them how it was. Think about something you just didn't get done and ask Muse. You will be shocked that I think five times out of ten it will do a better job than someone in whatever humans you have left in your office, it will do a better job. I try this. Your jaw will drop because between the memory, the browser agent, how it handles everything, the quality of the LLM. LLMs, the way the
1:02:03
Unidentified speaker (speaker_1): LLM is tailored to around these workflows, right? Rather than arguing with you or being developer focused. Um, it's very task focused. You will be shocked what it can do. It can do, hey, you know what? You want to get know a little bit more about AEO or GEO? Just ask. Ask Muse. How's my AEO?
1:02:19
Unidentified speaker (speaker_0): Jason, does it make you more or less bullish on instinct?
1:02:23
Unidentified speaker (speaker_1): As I said in RIC, I there's gonna be some vendors. I just worry about runaway infrastructure costs. There have to be some limits to it. There can't be just Rory's endless hand waving of great founders figuring it out. I mean slightly facetious, but I'm I'm not bullish on it today. But I haven't seen all the justification about how simple this will get. I'm not I'm
1:02:43
Unidentified speaker (speaker_1): not, because I do think I do think this product gets maimed if I can't run this thing, this gift from meta, which is that I can run. these agents 24-7.
1:02:52
Harry Stebbings: It's got a Cash Cal that powers the business.
1:02:54
Unidentified speaker (UNKNOWN): And if you are instinct.
1:02:55
Harry Stebbings: Just comment is if your typical user of these products is a Jason who wants a whole bunch of B2B
1:03:01
Unidentified speaker (speaker_3): functionality, then by definition he's going to gravitate to the company who can afford to give him all that functionality for one tenth of the price it costs them to deliver it, which is probably going to be meta rather than instinct, right? But the question maybe the other question is if the vast majority
1:03:14
Unidentified speaker (UNKNOWN): of the thing is.
1:03:15
Unidentified speaker (speaker_3): users today, and I accept your point, Jason, don't need that level of compute and aren't running their entire SMB on the infrastructure,
1:03:23
Unidentified speaker (speaker_0): then at least Instinct has a chance to compete and maybe can win on the basis of you know moving faster, being standalone. But still it is scary when
1:03:31
Unidentified speaker (speaker_3): a
1:03:31
Unidentified speaker (speaker_0): large company with five billion, four billion of human
1:03:34
Unidentified speaker (speaker_3): reach is producing a product that arguably technically is more overpowered than your product and has distribution up the wazoo. Now didn't work for threads, but there you go. This came a long way from that little question about that one person travel agency.
1:03:48
Unidentified speaker (UNKNOWN): And I freaking
1:03:49
Unidentified speaker (speaker_0): love the diversion. That was awesome.
1:03:50
Unidentified speaker (UNKNOWN): Oh
1:03:51
Unidentified speaker (speaker_1): sorry, why why did what what happened with the Dessia?
1:03:53
Unidentified speaker (UNKNOWN): Oh he
1:03:53
Unidentified speaker (speaker_0): called
1:03:54
Unidentified speaker (UNKNOWN): the company
1:03:56
Unidentified speaker (speaker_1): for the agent for agent
1:03:57
Unidentified speaker (UNKNOWN): for agents.
1:03:58
Unidentified speaker (speaker_1): I guess my comment is I know this is Captain Obvious, but I I think I think we're past the era of agentic apps. We're past the era of AI, CX, and crap. We are truly in the agentic era next year, like these horizontal agents, these autonomous agents. And this this is the big win for most of us at a lay level. Most of us are not. I mean, Cloud Code is the greatest app of
1:04:18
Unidentified speaker (speaker_1): all time, but most of us as humans are not cloud coding all the time. This is the it gets good next year. It gets good with agents for, and so we're gonna see 20 of these Odesia Edesia deals. to try to win.
1:04:28
Unidentified speaker (speaker_0): The biggest challenge I find is actually workforce mindset, which is like we have big teams here now. And actually, just getting people to go. Could an agent do this? What can I replace with agents? What are we not doing that agents could do? It's not a primary mentality for a human brain that's worked for 10 years to absorb. Yeah,
1:04:47
Unidentified speaker (speaker_1): and those and those people will disappear. Because what will happen is this is a transition phase. where you're trying to get humans to to be agents. When you do what I just did and you ask Muse to do it first, then what's happened is you just won't hire that person next. I mean our agents do all of our collections, okay? So our previous collection contractor who billed us three or four thousand dollars a month, we get billed one hour. Okay.
1:05:07
Unidentified speaker (speaker_1): So it will just get inverted and we will stop trying to get humans to work with agents and we'll just start asking agents to do it and we'll we'll we'll plug humans in into the gaps. It's a transition phase. And what'll happen is you'll then you won't even fire them. Like they'll they'll think they get a better job and they'll quit. And you don't you'll stop backfilling people, is what happens.
1:05:26
Unidentified speaker (speaker_0): I sat down with one of our guys and I said, just to let you know, I agree right now, the clips that you make, we can't do with AI. In a year, we will be able to. You need to get better or you will be replaced. I
1:05:37
Unidentified speaker (speaker_1): think that's right. I tell you, we went from 20 people to to three humans and 21 agents. You say that and it's true. You know what it requires today? It basically requires managing your agents for even more time. Six hours a day, eight hours a day, ten hours a day. Do you want us to be sitting in front of Muse for eight hours a day? It doesn't have to be you, but it'll be someone on your team, right? So it's not it's
1:05:57
Unidentified speaker (UNKNOWN): just. not I just
1:05:57
Unidentified speaker (speaker_0): wish I was Manlow Manlo with the carry. Fuck off.
1:06:02
Unidentified speaker (speaker_3): It it's good to see you doing it for the love of the game, Howie.
1:06:06
Unidentified speaker (UNKNOWN): Oh,
1:06:07
Unidentified speaker (speaker_1): come
1:06:07
Unidentified speaker (UNKNOWN): on.
1:06:08
Unidentified speaker (speaker_1): Yeah. I'd rather be sound sound ventures or Menlo. Who would you rather be? I'd rather be Sound Ventures.
1:06:13
Unidentified speaker (speaker_0): I would too. 435 million to 7 billion. Ashton Cook.
1:06:18
Unidentified speaker (UNKNOWN): Two GPs?
1:06:19
Unidentified speaker (speaker_1): Two GPs, call it a day. Don't have to work for anybody you don't want to.
1:06:22
Unidentified speaker (speaker_0): Amen. Norwest raises four billion dollar fund from one LP. Wells Fargo. Jesus, I wish wish we all had a Wells Fargo in our lives. SpaceX guts its Grok video TV.
1:06:34
Unidentified speaker (UNKNOWN): If
1:06:34
Unidentified speaker (speaker_3): you're gonna do I'm gonna stop it there. You're you're you're you're you're at the tail end of the agenda. If you're gonna do SpaceX, you can't talk about gro you just gotta give them credit for Starship. Just wanna say
1:06:42
Unidentified speaker (UNKNOWN): that we can do
1:06:43
Unidentified speaker (speaker_0): Starship. Starship,
1:06:44
Unidentified speaker (UNKNOWN): it's
1:06:44
Unidentified speaker (speaker_3): just not much to say except that is a fricking amazing technical achievement. And you know, honestly, I personally find uh the interpersonals on the guide to be a pain, the politics to be a pain, but that's a stunning, stunning, stunning achievement. And when you look back on the top five things of the year that matter most, that's gonna be one of them. You know, all the kind of orbital costs
1:07:04
Unidentified speaker (speaker_3): plummeting down by virtue of that. So I thought it was just cool. And a sustained effort over yeah, half a decade to make it happen.
1:07:12
Unidentified speaker (UNKNOWN): No, it it's
1:07:13
Unidentified speaker (speaker_3): not it's not gonna sound like.
1:07:14
Unidentified speaker (speaker_3): but it is stunning to realize despite that that now the whole rocket launch business is sub ten percent of the SpaceX market cap.
1:07:23
Unidentified speaker (speaker_0): Rory, would you buy SpaceX?
1:07:25
Unidentified speaker (speaker_3): Not at this price, because it's not about this rockets. The truth is the rockets, which are amazing, unique and differentiated, make the Starlink, which is very amazing, reasonably differentiated, especially for kind of the non-crowded non crowded area stuff. But then most of the business and most of the market cap and most of the TAM, according to their own S1, is all about
1:07:44
Unidentified speaker (speaker_3): compute. So really you're saying to yourself, would you buy compute? And that gets into the whole dynamic of forecasting compute for the next five years, which is just a different question. I might buy it, I might not buy it, but I wouldn't buy it because of this amazing technical achievement. You buy SpaceX the stock entirely on the basis of what you think Entropic will be paying you. one year from now for that compute.
1:08:04
Unidentified speaker (speaker_3): Will they be canceling it or will they be buying more? It's the only question. Well
1:08:07
Unidentified speaker (speaker_0): that sounds like a long only investor right there. No, I think you buy SpaceX because you read Sean Maguire's Excel model.
1:08:16
Unidentified speaker (speaker_3): I thought, you know, it's it I I I saw the comments on that. I thought, you know, it's kind of
1:08:20
Unidentified speaker (UNKNOWN): great. I I but
1:08:21
Unidentified speaker (speaker_0): I didn't mean that in stock sockets. as well.
1:08:22
Unidentified speaker (UNKNOWN): No, I know you didn't. I just
1:08:23
Unidentified speaker (speaker_3): want to say I want to say it too, though. I was like, everyone was like, Oh, you didn't take into account tax, dude. it's the big picture. How much money five years from now for me to the three markets? It was you don't have to believe in it to say, yeah, that's that's a perspective. All
1:08:33
Unidentified speaker (UNKNOWN): right,
1:08:33
Unidentified speaker (speaker_0): boys. Anything I've missed?
1:08:34
Unidentified speaker (speaker_1): Should we record a little bit at the end in case uh Jensen's back by the time we do it? I wasn't boys, I wasn't surprised at all that Jensen would come back. I told you it would happen by Thursday. Altman was back in what, 80 hours? How long was it for Jensen? A hundred? Do
1:08:47
Unidentified speaker (UNKNOWN): you
1:08:47
Unidentified speaker (speaker_0): know what Jason? That's a really interesting question. So I predict that Jensen. Jansen will come back. Yeah. In
1:08:53
Unidentified speaker (UNKNOWN): the
1:08:53
Unidentified speaker (speaker_3): next week or over a period of time? Oh
1:08:55
Unidentified speaker (speaker_0): no, I think Jansen will come back in the next three days.
1:08:58
Unidentified speaker (speaker_1): So let's let's record it now so that so that so that you're right either way.
1:09:01
Unidentified speaker (speaker_0): He's got to come back in the next three days, otherwise the chorus dies, anthropic IPO happens, Trump does something else, and it fizzles. This is the moment. This is the mother load.
1:09:12
Unidentified speaker (speaker_3): I don't think that'll happen. I mean, because it'd be great, but
1:09:15
Unidentified speaker (UNKNOWN): that
1:09:15
Unidentified speaker (speaker_0): if it doesn't happen, it won't happen. Well,
1:09:17
Unidentified speaker (speaker_3): yeah, some things don't happen, Harry.
1:09:19
Unidentified speaker (UNKNOWN): And
1:09:19
Unidentified speaker (speaker_0): that's the mistake of Jensen and the board, which is news cycles have never been faster. And I think if they'd just put it on chill, then anthropic IPO would have happened, Daria be outside the stock exchange, woo-hoo, woo-hoo. And it would have fizzled.
1:09:33
Unidentified speaker (speaker_3): Okay.
1:09:34
Unidentified speaker (speaker_0): Boys, what a joy, as always. It's a wrap.
1:09:39
Unidentified speaker (speaker_0): But before we leave you today, base 44 is one of those products I used, and I just thought straight away, man, I wish I'd had this when I started 20 VC with $50 and no contacts in Venture. I wanted to see what it could do. So I built a guest research hub for the show. Upcoming guests, company notes, previous interviews, sponsor contacts, contests,
1:09:58
Unidentified speaker (speaker_0): follow-ups, all in one amazing location. I described it in plain English. A few minutes later, I had the product open in front of me. Database, login, backend, and hosting already handled and just all in one. It made it so easy. And that's what I loved. I didn't write a long spec. I didn't stitch five different tools together. I built the thing and I started using it. For any founder
1:10:18
Unidentified speaker (speaker_0): listening, that's the point. The faster you turn an idea into something real, the faster you launch apps, dashboards, internal systems, AI agents, whatever. you need to build, base44 gets you there. Go to base44.com forward slash 20 VC. That's base44.com slash 20vc. While base44 helps you build,
1:10:38
Unidentified speaker (speaker_0): AlphaSense helps you find answers really, really fast. Quick one for anyone doing diligence in venture right now, AlphaSense just launched Super Analyst and it's one of the more genuinely useful AI products that I've seen come out of this space. It's an always-on agent that works like a power user inside AlphaSense, but on your behalf. So check this out. It pulls from
1:10:58
Unidentified speaker (speaker_0): over 500 million premium sources, covering both public and private companies. That's filings, earning calls, broker research, expert transcripts. And it doesn't stop when you close your laptop, point it at a sector, a company, a thesis you're tracking, and it keeps working in the background. And here's the part that I really care about. It remembers your process, it
1:11:18
Unidentified speaker (speaker_0): learns your templates, your firm's logic, and it applies that to every new task. So it compounds, it gets better the more you use it. Isn't that the value? And every single output links back to its source. So there's no black box, there's no trust me, and you can see exactly where something came from. So you have complete traceability. Pick the deal memo or the market map.
1:11:39
Unidentified speaker (speaker_0): you've been putting off and hand it over to Superanalyst instead. Start your free trial at alpha-sense.com forward slash 20vc. That's alpha-sense.com forward slash 20vc. While Alpha Sense finds insights, intercom helps you serve customers. As AI agents become more common in customer
1:11:58
Unidentified speaker (speaker_0): experience, teams often end up juggling multiple silo tools for every job. Well, Fin was built to change that. It's a single unified agent that works across your entire customer experience, from service to sales to success and beyond. Fin is the agent making perfect customer experiences possible for thousands of customers. It's powered by custom
1:12:18
Unidentified speaker (speaker_0): models, trained on years of real customer interactions, so it understands the nuance and complexity of customer service better than any other agent. That means faster resolutions, more consistent support, and just better experiences for every customer. It's also designed to be fully self manageable so you can easily improve and adapt it as your business evolves. No
1:12:38
Unidentified speaker (speaker_0): third parties required. Leading companies like Gamma, Asana, DoorDash, and Crypto.com already use and Lovefin love fin to deliver better customer experiences. So for a limited time, you can get $500 a month in things. Fin credits for your first three months. Learn more at fin.ai forward slash 20 VC.
This page was produced automatically: the audio was transcribed with nvidia/parakeet-tdt-0.6b-v3+sortformer-v2.1 and interpreted by Qwen3.6-35B-A3B-UD-IQ3_S.gguf. Transcription and interpretation can contain errors; always confirm against the original episode. Nothing here is investment advice.

More from The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch

The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch

20VC: Jev Creator on Why Data Is More Important Than Compute | Will US Open-Source Models Catch Chinese | The Cost of Intelligence and Tokens: What No One Sees But Needs to Know with Diogo Almeida

Episode discussion: Jev Creator argues AI reliability and data quality trump compute, predicting a shift from front-end apps to background automation.

Speakers’ claims: Jev Creator (Diogo Almeida) argues that the primary bottleneck for AI automation is not compute or speed, but the reliability of intelligence, defined as the consistency of decision-making rather than system uptime. He posits a 'bitterer lesson' where task definition and data quality are more critical than algorithms or compute, which he claims have plateaued in effectiveness since GPT-4.5. Creator criticizes current market hype, benchmark-maxing, and the 'AI coworker' model, predicting a 'SaaS Palooza' where incumbents win by integrating AI into workflows rather than replacing them. He also discusses TypeSafe AI's fundraising strategy, hiring philosophy, and the geopolitical landscape of open-source models.

AI ReliabilityData QualityVenture Capital
1h 14m2 assets22 claims to verify
The Twenty Minute VC (20VC): Venture Capital | Startup Funding | The Pitch

20VC: Taking on NVIDIA: Fractile CEO on the Future of Chips | What Everyone Gets Wrong About Capex Spending | The Winners and Losers in Neo Clouds with Walter Goodwin

Episode discussion: Walter Goodwin on Fractile's 3D-stacked memory chips, AI infrastructure bottlenecks, and NVIDIA's future valuation

Speakers’ claims: Walter Goodwin, CEO of Fractile, argues that frontier AI intelligence is not yet commoditized, distinguishing it from general cloud compute. He positions Fractile's 3D-stacked memory architecture as a solution to the memory bandwidth bottleneck that limits LLM inference speeds. Goodwin contends that land and power, rather than silicon, are now the primary constraints for AI infrastructure, favoring energy-efficient designs. He predicts NVIDIA will reach a $10 trillion valuation within five years, driven by the finite nature of data center power and the need for chip replacement cycles.

SemiconductorsAI InfrastructureVenture Capital
1h 16m3 assets25 claims to verify
All insights