NVIDIA Ousts Jensen, AI Comes for Nerds & Walmart Is A Bubble?!
· Summarized
Key takeaways
- Market crashes can make years of work appear worthless, but Draper argued that failed entrepreneurs still advance the system by improving products, pressuring competitors, and creating groundwork for later companies.
- AI is lowering the cost of shipping software so sharply that small teams, and sometimes nontechnical founders, can build and test products without hiring additional developers.
- The panel argued that venture performance may depend as much on access, reputation, marketing, and deal flow as on superior company selection.
- The speakers favored founders who defend their own ideas, process evidence independently, and can reject investor advice rather than being reflexively “coachable.”
- Late-stage venture and large SPVs may be vulnerable because they often offer investors technology exposure without the risk of early discovery investing; the panel viewed seed investors as more defensible.
- AI could reduce switching costs and weaken the durability of incumbent companies’ cash flows, raising the possibility that apparently safe stocks such as Walmart and Costco are more richly valued than some technology companies.
- The panel refused to reach a firm conclusion on Jensen Huang’s removal from Nvidia, saying the board should establish the facts, follow governance procedures, and weigh both shareholder interests and national-security concerns.
Summary
Tim Draper’s crash-tested view of entrepreneurship
Tim Draper said he began venture investing in 1985, before the 1987 crash, and later lived through the 1989 crash and the dot-com collapse. He remembered thinking his work had become worthless, but concluded that entrepreneurs still “move the ball forward” even when their companies fail: MySpace and Friendster, for example, established groundwork for later products. During the dot-com crash, the Nasdaq fell from 5,000 to 1,100, while 19 of his fund’s 21 companies had gone public before the wipeout. He bought out investors who could not meet capital calls and said post-2001 venture remained difficult for four years.
The panel compares today’s market with earlier downturns
D.A. Wallach compared the current environment with biotech’s collapse beginning in late 2021. COVID created a “sugar high,” then public biotech markets fell and private financing stopped about eight months later; the sector took roughly three and a half years to recover. Asked whether current AI turbulence resembles the late 1990s or 2008, D.A. said no and compared it instead with the PayPal IPO, when a company went public during a gloomy period. He remained bullish despite possible volatility around an Anthropic IPO. Draper saw similarities to AOL and At Home: infrastructure spending and easy money, but emphasized that AI is already producing extraordinary revenue and utility.
Draper says AI lets anyone ship software
Draper described using Grok to build a fantasy-football application that teaches his family to invest. The first version worked immediately, but adding features eventually “breaks the camel’s back” and makes the application unravel. He ranked ChatGPT as the assistant that tells users what they want to hear, Grok as more truth-seeking, and Anthropic as resembling a Russian bureaucrat whose paperwork must be stamped. The panel agreed that AI has removed the longstanding startup bottleneck of needing more developers: one or two people can now ship products and potentially reach $20,000 in monthly revenue almost immediately.
Wallach’s “revenge of the jocks” theory
Wallach argued that AI marks the end of a 30-year “nerd tyranny” in which money and power flowed to programmers and technical specialists. Because nontechnical founders can now build applications with AI, he called the coming period the “revenge of the jocks” and predicted that finance could revert toward an old-fashioned Wall Street culture. He said the existential fear voiced by many AI programmers partly reflects the collapse of their status and purpose. McClure called himself an AI optimist but warned that AI could disintermediate later-stage venture investors first, since it can analyze measurable financial and operating metrics more reliably than human investors.
AI may empower unconventional founders and investors
McClure said AI has moved people through denial, anger, sadness, and perhaps eventual acceptance about job displacement. He worried that AI could become a better investor for metric-driven later-stage deals, while humans retain an edge in early-stage judgment about founders and unconventional ideas. He also said AI empowers “weirdos” by letting them test unusual ideas rapidly: a complex proposal involving GP carry, a forward contract, and a donor-advised-fund contribution took him less than an hour instead of a week. The group discussed “handoverism,” the idea that humans should eventually surrender control to AI; Sam Altman and Greg Brockman rejected that position.
A paper questions whether VC picking adds value
Wallach presented a finance paper by J.P. Bouchaud and coauthors that compared actual venture portfolios with portfolios created by randomly selecting a similar number of companies. After controlling for factors including valuation, sector, and investment size, the random portfolios had distributions of cash-on-cash outcomes that nearly overlapped real VC portfolios, including similarly extreme winners. Draper objected that randomly sampling from roughly 80,000 opportunities would include many companies that never qualified for later financing. Wallach agreed the sample was biased toward companies that had raised money and noted that entry price mattered: overpaying produced worse results.
Brand and access may matter more than selection
The panel challenged the paper’s implication that picking is unimportant. Wallach noted persistence among top managers and argued that elite investors gain access to better deals as a result of earlier successes. McClure said early winners create confirmation bias and stronger deal flow, but suggested that better access may improve returns even if the investor is merely competent. At 500 Startups, he found that making 100 investments of $100,000 generated far more publicity and deal flow than making a few $1 million or $10 million investments. Calacanis said he repeatedly publicizes breakout portfolio companies for the same reason. Wallach summarized venture’s important functions as deal flow, decisions, doubling down, and distributions.
Why the panel values difficult founders
Calacanis described a founder-scoring system based on product velocity, design quality, doggedness, communication, and productive argumentativeness. He rejected “coachable” as a positive founder trait, saying venture investors should seek people who execute their own ideas rather than meekly follow advice; McClure preferred the word “agreeable” for employees, not CEOs. Draper said he does not want founders to accept his suggestions automatically: he watches whether they think through the idea and either adopt it, reject it, or explain that they already tried it. Wallach emphasized a scientific mindset that changes opinions when strong evidence changes, while Calacanis warned that scientists can sometimes miss the broader picture.
SPVs, late-stage funds, and the search for synthetic liquidity
McClure said SPVs have grown more than threefold according to Sydecar data, partly because LPs increasingly want to choose individual deals instead of committing to blind-pool funds. Draper supported SPVs but criticized late-stage funds charging “two and 20” on enormous pools while adding little value; he said those firms “might as well be AI.” He cited Draper SPVs for Robinhood at a $1 billion valuation and Oklo at $250 million, allowing early-stage funds to participate after companies outgrew their mandate. McClure described a separate use: moving appreciated holdings into SPVs to create liquidity for funds, GPs, and LPs before an IPO or acquisition.
The alleged bubble may be in supposedly safe stocks
Wallach proposed that large allocators use SPVs to appear involved in futuristic technology without taking genuine early-stage risk. He then questioned whether the market’s bubble is actually in conventional “safe” companies: Walmart and Costco were trading at price-to-earnings ratios around 40 to 50, while Nvidia was at roughly 16 and Micron at about four times forward earnings. McClure added that AI may reduce switching costs by helping customers rebuild applications or move banking information, weakening the durability of incumbent companies’ future cash flows and therefore their valuation multiples. The panel remained skeptical of a classic AI bubble because widespread bubble talk itself can signal caution rather than euphoria.
The panel withholds judgment on Nvidia’s Jensen Huang dispute
Discussing Nvidia’s decision to remove Jensen Huang, Wallach said the facts were unclear: the board’s reasoning, Huang’s conduct, possible chip-routing through Southeast Asia, and shifting U.S. policy toward China were all uncertain. He opposed letting one powerful entrepreneur dictate national policy but also acknowledged that shareholders and major stakeholders might want Huang back. McClure focused on process: Nvidia should determine whether Huang violated laws or internal procedures and whether succession planning is adequate, rather than reacting to partisan or founder-centered narratives. Draper said boards should not lightly remove a company’s entrepreneurial “heart,” suggesting fear and internal ambition may have influenced the decision, but acknowledged that reinstatement would require the board to admit it was wrong.
Tim Draper remembers his father Bill
Draper closed by describing his father, Bill Draper, who lived to 98 and moved smoothly from being a guiding parent to being a supportive friend. Bill helped pioneer the GP-LP venture structure, promoted “venture philanthropy,” and worked in government, including at the Export-Import Bank and the United Nations Development Program. He visited more than 100 countries and advocated capitalism, democracy, women in the workforce, and environmental protection; Draper recalled him helping Mongolia after the Soviet Union’s collapse. McClure remembered Bill as relentlessly optimistic, sharp, funny, and influential beyond venture, recalling a stage event featuring three generations of Drapers.
Sponsors
- Vanta · Vanta helps startups obtain SOC 2, HIPAA, GDPR, and other compliance frameworks quickly, so founders can close enterprise deals while engineering stays focused on the product. The offer is $1,000 off.
- Plaud · Plaud’s NotePin records conversations, meetings, interviews, and calls without damaging clothing. The host gave the code TWIST for 20% off through November 30.
- Stripe Startups · Stripe offers early-stage, venture-backed startups fee credits, expert mentorship, operational tools, and access to a builder community. The panel noted that more than three-quarters of Forbes’ AI50 companies use Stripe for payments, billing, invoicing, and taxes.
- Sentry · Sentry connects errors, traces, replays, and logs in one timeline. Its SEER AI debugging agent can identify the responsible code, draft a fix, and warn when a pull request may create a problem. New users can receive $240 in credits with code TWIST26.
- GoDaddy · The host recommended GoDaddy for registering and holding premium domains, mentioning that he owns begin.com, inside.com, and mahalo.com and is selling some of them.
Show notes
This Week In Startups is made possible by:
Vanta: https://www.vanta.com/twist
Stripe: https://stripe.com/startups
Sentry: https://sentry.io/twist
Plaud: https://plaud.ai/twist
Today’s show:
Tim Draper thinks late-stage venture funds are living on borrowed time.
He notes that two and 20 on a $20 million fund is $400 million a year without adding any significant value. Those firms “might as well be AI” in his estimation, while seed investors who make the first call are actually earning their fees.
On a new TWiST VC Roundtable, Draper (Draper Associates) joins Jason Calacanis, plus Dave McClure (Practical Venture Capital) and our new regular co-host D.A. Wallach (Time BioVentures) for a lively discussion of investment bankers posing as VCs, plus more of the week’s biggest stories and trends for elite capital allocators.
That includes a new paper that found randomly selected portfolios performed about as well as real VC portfolios. Do venture returns come from making smart choices, or simply access to the best deals? Plus Dave explains how 500 Startups turned hundreds of $100K checks into a marketing machine, D.A. theorizes that the real bubble as conventionally safe stocks like Walmart and Costco, and the panel debates — what else — whether or not Nvidia’s board should Bring Back Jensen.
Guests
Tim Draper on X: https://x.com/TimDraper
Draper Associates: https://www.draper.vc/
Dave McClure on X: https://x.com/davemcclure
Practical Venture Capital: https://practicalvc.com/
D.A. Wallach on X: https://x.com/dawallach
Time BioVentures: https://www.timebioventures.com/
Relevant Links
"Do Venture Capitalists Beat Random Allocation?" (Knicker, Bouchaud, Benzaquen, arXiv): https://arxiv.org/abs/2605.03980
Joachim Klement: "Do venture capitalists beat a random benchmark?": https://klementoninvesting.substack.com/p/do-venture-capitalists-beat-a-random
Softwar: An Intimate Portrait of Larry Ellison and Oracle by Matthew Symonds: https://www.amazon.com/dp/0743225058
Sequoia's "R.I.P. Good Times" deck, and Roelof Botha on why it was no mistake (TechCrunch): https://techcrunch.com/2011/05/24/sequoias-roelof-botha-r-i-p-good-times-was-no-mistake-and-yes-they-do-read-their-email
Tim Draper compares OpenAI to AOL and predicts an AI correction (CO/AI): https://getcoai.com/news/billionaire-vc-tim-draper-compares-openai-to-aol-predicts-ai-market-correction/
Nvidia's Q3 fiscal 2027 results: https://nvidianews.nvidia.com/news/nvidia-announces-financial-results-for-third-quarter-fiscal-2027
Tim Draper wins the Silk Road bitcoin auction, 2014 (CoinDesk): https://www.coindesk.com/tim-draper-revealed-silk-road-bitcoin-auction-winner
Sam Altman warns against a future where AI, not humans, calls the shots (TechCrunch): https://techcrunch.com/2026/11/18/sam-altman-warns-against-a-future-where-ai-not-humans-calls-the-shots/
Sydecar's 2025 SPV Year in Review: https://sydecar.io/guides/2025-spv-year-in-review
SPVs and the reshaping of venture capital (Mondaq): https://webiis08.mondaq.com/unitedstates/corporate-and-company-law/1732114/spvs-and-the-reshaping-of-venture-capital
Nvidia board says it won't reinstate Huang (Yahoo Finance live): https://finance.yahoo.com/technology/live/nvidia-earnings-live-nvidia-posts-q3-beat-board-says-it-will-not-reinstate-jensen-huang-as-ceo-221428433.html
Elon Musk: hiring Jensen "would be an honor": https://x.com/elonmusk/status/2120688743111971293
Bill Draper obit (NYT): https://www.nytimes.com/2026/09/30/technology/william-draper-dead.html
Venture Capital in the Blood: Three Generations of Drapers (Computer History Museum): https://computerhistory.org/blog/venture-capital-in-the-blood-three-generations-of-drapers-in-silicon-valley/
Timestamps:
0:00 Tim started off in VC during a crash
8:20 Plaud: If your work depends on conversations - interviews, meetings, calls - you need a Plaud NotePin S. You can check it out at https://plaud.ai/twist and use code TWIST for 10% off!
9:29 Dave is reminded of the PayPal IPO
11:10 Vanta - Get $1000 off your SOC 2 at https://www.vanta.com/twist
15:01 The panel ranks the top chatbots
19:35 Stripe Startups - Stripe Startups offers early-stage, venture-backed startups access to Stripe fee credits and more. Apply today on https://stripe.com/startups.
24:35 Why AI empowers weirdos
25:58 What’s the post overrated aspect of venture?
31:12 Sentry - New users can get $240 in free credits when they go to https://sentry.io/twist and use the code TWIST26
45:01 Rise and fall of “investment banker” VCs
56:18 Are Walmart and Costco the real bubble?
1:03:55 DFJ’s SpaceX-heavy fund and the LP swap
1:08:32 “It’s murky in all directions”
Subscribe to our newsletters on Substack:
TWiST: https://twistartups.substack.com/
TWiAI: https://www.thisweekinai.ai/
TWiVC: https://thisweekinvc.substack.com/
Subscribe to This Week in Startups on Apple: https://rb.gy/v19fcp
Follow Jason:
LinkedIn: https://www.linkedin.com/in/jasoncalacanis
Thank you to our partners:
Vanta - Get $1000 off your SOC 2 at https://www.vanta.com/twist
Stripe Startups - Stripe Startups offers early-stage, venture-backed startups access to Stripe fee credits and more. Apply today on https://stripe.com/startups.
Sentry - New users can get $240 in free credits when they go to https://sentry.io/twist and use the code TWIST26
Plaud: If your work depends on conversations — interviews, meetings, calls — you need a Plaud NotePin S. You can check it out at https://plaud.ai/twist and use code TWIST for 10% off!
Check out all our partner offers: https://partners.launch.co/
Great TWIST interviews: Will Guidara, Eoghan McCabe, Steve Huffman, Brian Chesky, Bob Moesta, Aaron Levie, Sophia Amoruso, Reid Hoffman, Frank Slootman, Billy McFarland
Follow TWiST:
Twitter: https://twitter.com/TWiStartups
YouTube: https://www.youtube.com/thisweekin
Instagram: https://www.instagram.com/thisweekinstartups
TikTok: https://www.tiktok.com/@thisweekinstartups
Substack: https://twistartups.substack.com
Follow this show on AudioBrevity and new episodes get summarized for you. Create a free account