• Consider Eli Lilly (LLY) for long-term biotech exposure; the investor is bullish, but offered no price target or timeline, so weigh clinical and commercial execution risks.
• Figma (FIG) is a higher-risk, contested AI-era software idea: the investor is long, but valuation, AI competition, and customer shifts warrant close monitoring.
• The investor is cautiously positive on Bitcoin (BTC) and selected crypto assets, but gave no price targets; treat named tokens such as Hyperliquid as disclosed holdings, not automatic buy recommendations.
• Avoid treating broad AI infrastructure growth as a buy signal: the investor flagged competition, possible consolidation, and margin pressure, and expressed a more bearish view of HubSpot (HUBS) than Salesforce (CRM).

AI Labs and AI Infrastructure

• The investor described OpenAI and Anthropic as difficult to assess over a roughly 24-month horizon, even while viewing them as potentially interesting over a 10-year horizon. When asked about the labs at hypothetical $2 trillion valuations, he said he was neither long nor short.
• He believes leading AI labs’ ability to scale compute, data, and model performance could undermine some businesses, especially generic software “wrappers.” He also sees uncertainty over whether value will accrue to the labs, product companies, or other parts of the technology stack.
• AI infrastructure providers—including Fireworks, Together, Modal, and FAL—were described as growing quickly. But the speaker cautioned that a company growing 6x while its overall market grows 10x may not be winning relative to competitors.
• He expects potential consolidation, greater competition across parts of the infrastructure stack, and possible margin erosion. Companies may also face constraints such as limited GPU supply.
• NVIDIA was mentioned in the context of large expected AI-compute spending, not as a specific investment recommendation.

Takeaways

• Assess AI companies against the growth and competitive dynamics of their markets—not just their headline growth rates.
• For infrastructure businesses, consider whether the company can remain differentiated as competitors expand across the stack and market growth eventually settles.
• The discussion was cautious about durability at current levels of AI excitement; it did not establish a price target or direct recommendation for the named companies.

Public Software: Salesforce and HubSpot

• The investor said Salesforce (CRM) was likely to benefit from greater-than-expected business inertia. He said he was neither long nor short.
• He expressed a more bearish view of HubSpot (HUBS), describing himself as “probably a little more short.”
• More broadly, he is concerned that AI tools could make it easier to create open-source alternatives and clones of public SaaS products. He said this trend had accelerated in GitHub repositories and could intensify as AI models improve at decompiling software.
• He did not argue that all SaaS companies are doomed; he said a small subset could remain attractive.

Takeaways

• Consider how difficult a software product would be to replace, and whether customer inertia, proprietary capabilities, or other advantages can protect it from AI-enabled alternatives.
• The discussion’s skepticism was selective, rather than a blanket bearish call on all SaaS companies.

Figma

• The investor said he was long Figma (FIG), though it was not one of his firm’s top four positions.
• His bull case is that creative work still needs a canvas and that AI could increase usage, particularly if usage-based pricing captures greater demand. He also said Figma had begun shipping products he thought were overdue.
• He acknowledged the bear case: AI tools could weaken the importance of design software, AI-generated work could become “slop,” the stock was not cheap on a multiple basis, and the company had experienced executive turnover.
• The conversation highlighted mixed user experiences: one participant said their team had moved from Figma to Claude Design, while the investor argued that Figma’s model-context-protocol (MCP) integration could improve the product.

Takeaways

• The discussion presents Figma as a contested AI-era software investment, not an uncomplicated recovery story.
• Follow whether AI features increase Figma’s usefulness and paid usage, while weighing valuation, product competition, and the risk that customers shift their workflows elsewhere.

Biotech and Eli Lilly

• The investor said he was long Eli Lilly (LLY) and thought it could become a multi-trillion-dollar company. No target price or timeline was provided.
• He expects AI to help biotech, but said its more immediate impact may be making drug development safer and faster, rather than automatically solving every discovery problem.
• He believes larger pharmaceutical companies may benefit from their cash, expertise, clinical-trial experience, and regulatory infrastructure. He also noted Lilly’s efforts to move closer to direct-to-consumer distribution.
• He contrasted Lilly’s outlook with Novo Nordisk (NVO), saying Novo was trading below the price it had reached when it originally began getting into GLP-1 drugs. This was presented as a comparison, not an explicit short recommendation.
• He forecast that 50% of major diseases that kill people today might not do so in 20 years, while acknowledging that it is uncertain whether AI labs themselves will drive those advances.
• The transcript also discussed potential changes to biotech regulation and faster clinical testing. Separately, the investor warned that more people may try self-experimentation, with some experiencing negative effects.

Takeaways

• The discussion is bullish on the long-term potential of biotech and specifically positive on Lilly, while emphasizing that the path depends on clinical, regulatory, and commercial execution.
• Treat the claims about future disease outcomes and Lilly’s possible scale as the investor’s forecasts, not established outcomes.
• In biotech, consider how a company handles clinical development and distribution—not only whether it uses AI.

Private AI and Deep-Tech Companies

• The investor described several portfolio companies as examples of AI and deep-tech opportunities:
  • Wayve (referred to as “Wave” in the transcript), a self-driving company, and Runway were described as large companies that began as early investments.
  • Prime Intellect works on distributed compute and post-training open-source models.
  • General Intelligence Company is building an agent platform intended to support one-person or agent-native companies.
  • Orbital Materials uses models to develop new materials; its work includes a modular data-center product designed around GPU cooling.
• The investor’s broader thesis is that deep-tech businesses may be more durable when they turn scientific advances into products with a path to market, rather than stopping at a model or research result.
• He also described a potential company-building opportunity in creating new materials that enable new products. He gave cheaper, scalable bullet manufacturing as an example of the business shape, but said he had no desire to invest in a bullet company.
• He said energy remains interesting but has become crowded.

Takeaways

• In private deep tech, the discussion favors examining whether research can become a commercial product and whether the company can control more of its route to market.
• The speaker warned that businesses dependent on a single external partner, customer, launch provider, or systems integrator can have their prospects shaped by that dependency.
• These were examples of investment themes and portfolio companies, not public-market buy recommendations.

Mental Health, Talkspace, and Pantalossa

• The investor said his firm had invested in Talkspace (TALK) early, when therapy delivered through video and text was still an unusual idea.
• He described expanding access to mental-health care as an important opportunity and said the firm’s latest investment was in AI-oriented psychiatry; he did not name that company.
• He also named Pantalossa, which generates energy through ocean waves, as a portfolio company.

Takeaways

• The discussion identifies mental-health access and alternative energy generation as areas with potential, but provides no public-market price targets or specific trading recommendations.
• Energy opportunities may warrant extra scrutiny because the investor said the sector has become crowded.

Crypto and Digital Assets

• The investor said his firm invests in crypto tokens and described crypto as potentially asymmetric over the long term. He is also personally positive on Bitcoin (BTC).
• He said crypto venture activity had been weak for nearly two years, in part because talent and attention had shifted toward AI. He sees improved willingness to build crypto businesses as a positive, but believes attracting talent remains a challenge.
• He said liquid crypto assets were interesting to his firm, particularly tokens that are more widely distributed and may be evaluated on fundamentals rather than narrative alone.
• He named Hyperliquid, HairDAO, and Kinetic among larger positions in the fund, and said the firm had historically owned Helium. These holdings were disclosed, but he did not provide individual price targets or detailed current views on each token.
• He described rare-disease research funding as a promising potential crypto use case, mentioning decentralized science projects and HairDAO.
• He cautioned that tokenization can create poor long-term incentives.
• He said he disliked prediction markets and hyper-gambling, and would not buy Pump.fun’s token. He also criticized Robinhood’s historical gamification of investing and said he would not go long Robinhood.

Takeaways

• The investor’s crypto stance was cautiously constructive, particularly on selected liquid assets and Bitcoin, but he also identified talent, token incentives, and the sector’s limited current activity as concerns.
• Treat the named tokens as disclosed holdings, not as blanket recommendations; the transcript does not provide a valuation case for each one.
• The speaker drew a negative distinction between investing and highly short-term, gambling-like products, including prediction markets and meme-coin activity.

Crypto-Related Companies and Public-Market Exposure

• The investor said he had been trying to buy Chainalysis shares on the private market. He sees the company as potentially well positioned if crypto remains important and cybercrime continues to grow, citing its work with difficult customers.
• He named Coinbase (COIN), BitGo, Circle (CRCL), and Securitize as examples of publicly traded ways to gain crypto exposure at the time of the conversation.
• He also mentioned Fireblocks, TRM Labs, and Ledger as private companies that might eventually pursue public listings, while noting that they were not necessarily ready to do so.
• He did not disclose an investment in Chainalysis or provide price targets for these companies.

Takeaways

• The discussion points to crypto security and infrastructure as possible ways to invest in the sector beyond tokens, particularly if crypto activity and cybercrime continue to expand.
• The investor’s interest in Chainalysis was tied to those assumptions; it was not presented as a public-market recommendation.

Venture Investment Approach

• The investor said his firm focuses on technically demanding areas such as AI, robotics, healthcare, biotechnology, crypto, materials science, and energy.
• He described favoring companies with substantial scientific or engineering risk and said the firm often passes on crowded categories where it does not believe it has an advantage.
• He said the firm avoided investing for about 16 months after raising a fund in 2021, but also acknowledged that it was not aggressive enough when it resumed investing in 2023.
• He emphasized that venture companies need to be durable across multiple innovation cycles, rather than relying only on a current wave of excitement.

Takeaways

• For investors evaluating early-stage opportunities, the discussion emphasizes specialization, technical differentiation, and durability over chasing whatever category is currently attracting the most attention.
• The speaker’s account also illustrates a trade-off: waiting through a speculative period may limit exposure to weak deals, but moving too slowly can mean missing promising companies.