• Eli Lilly (LLY) is the clearest publicly traded long idea: the speaker is bullish on its drug-development capabilities, scale, and direct-to-consumer ambitions, though no price target was provided.
• Figma (FIG) is a higher-risk long: AI could strengthen its collaborative design platform, but competition from AI-native tools and valuation remain key risks.
• The speaker is more short-biased on HubSpot (HUBS) and SaaS broadly due to AI-enabled software clones, but gave no specific short target; assess each company’s ability to defend its differentiation.
• The speaker is positive on Bitcoin (BTC) over the long term, but offered no entry point or allocation guidance, so treat it as a selective, high-risk holding.

Salesforce (CRM)

• The investor was neither long nor short, saying the company’s inertia may be stronger than people expect.
• The broader concern is that AI could pressure some software businesses, but the speaker did not view Salesforce as an obvious short.

Takeaways

• Treat Salesforce as a case where customer and business momentum may provide resilience, while recognizing that the transcript offers no specific valuation or recommendation.

HubSpot (HUBS)

• The speaker was more short-biased on HubSpot than on Salesforce.
• This view was part of a broader skeptical stance toward SaaS companies, based in part on the potential for AI tools to enable software alternatives and clones.

Takeaways

• The discussion points to disruption risk for SaaS, but does not provide a price target or a specific short recommendation.

Figma (FIG)

• The speaker said he was long Figma, while acknowledging that it was not among the firm’s top four positions.
• The bullish case is that AI could enhance creative work on a shared design canvas and increase product usage, potentially supporting usage-based pricing.
• The speaker cited Figma’s product expansion and argued that the stock had fallen significantly from its IPO, though it was still not cheap.
• The bear case included AI agents replacing design tools, weak or “sloppy” AI-generated output, executive turnover, competition from Anthropic’s Claude Design, and valuation risk.
• The hosts also discussed Figma’s Model Context Protocol (MCP) integration as a potentially useful way to connect AI tools with design workflows.

Takeaways

• Figma is presented as a contested AI-disruption case: the upside depends on its products becoming more valuable with AI, while the downside is that AI-native tools could displace it.
• The speaker’s positive view is not a blanket endorsement; he explicitly recognized valuation and competitive risks.

OpenAI and Anthropic (Private)

• Asked about a roughly 24-month investment horizon, the speaker said he was neither long nor short on OpenAI or Anthropic.
• He compared them to SpaceX in the sense that they may be interesting over a 10-year horizon, but were difficult to assess over a shorter period.
• The speaker believes leading AI labs could capture value in some markets and undermine many companies built on top of their models.

Takeaways

• The discussion favors caution about short-term outcomes: the labs could become powerful platforms, but the transcript does not identify a clear near-term investment position.
• For companies dependent on AI labs, consider how their products and margins might hold up if model providers expand into adjacent products.

NVIDIA (NVDA)

• NVIDIA was mentioned in the context of large projected AI infrastructure spending and demand for compute.
• The speaker also warned that AI infrastructure businesses may face finite GPU supply, consolidation, and pressure on margins as competitors expand across the value chain.

Takeaways

• The transcript highlights the scale of AI compute demand but does not express a specific view on NVIDIA stock.
• Investors considering AI infrastructure exposure should distinguish long-term demand from the risk that competition, supply constraints, or changing economics affect individual companies.

Eli Lilly (LLY)

• The speaker said the firm was long Eli Lilly and suggested it could become a multi-trillion-dollar company.
• He pointed to the company’s large cash resources, expertise, and efforts to move closer to direct-to-consumer distribution.
• More broadly, he expects AI to help create valuable new biotech assets, while established pharmaceutical companies may benefit from their experience with clinical trials and regulation.

Takeaways

• The positive view rests on Lilly’s scale, drug-development capabilities, and potential distribution advantages; the transcript does not give a price target.
• The speaker also emphasized that smaller biotech companies could create valuable drugs, so established pharma may not capture all the gains.

Novo Nordisk (NVO)

• The speaker contrasted Novo Nordisk with Lilly, noting that Novo was trading below the level it had reached when GLP-1 drugs first gained prominence.
• This was used as an example of the sharp reversals and uncertainty that can occur even in successful drug markets.

Takeaways

• The comparison underscores that strong products do not eliminate volatility or guarantee continued share-price gains.
• No specific recommendation or price target for Novo Nordisk was given.

Robinhood (HOOD)

• The speaker said he would not invest in Robinhood, criticizing its marketing and use of push notifications to encourage user activity.
• He specifically objected to the gamification of investing and the promotion of short-term trading behavior.

Takeaways

• The speaker’s concern is about the company’s business practices and incentives, rather than a stated valuation or earnings outlook.
• Investors may want to consider whether a platform’s user-engagement model aligns with their views on customer outcomes and long-term business risk.

Bitcoin (BTC)

• The speaker described himself as a big Bitcoin person.
• More broadly, he sees crypto’s underlying principles as likely to persist, even if the sector’s popularity and use cases change over time.

Takeaways

• The transcript expresses long-term interest in Bitcoin but gives no price target, timing, or explicit portfolio allocation.
• The speaker also noted that crypto has struggled to attract builders and attention while technology talent has shifted toward AI.

Hyperliquid (HYPE)

• Hyperliquid was named as one of the larger crypto positions in the firm’s portfolio.

Takeaways

• This indicates an existing investment position, but the speaker did not explain its thesis, give a target, or describe specific risks.

HairDAO (HAIR)

• HairDAO was named among the firm’s crypto holdings.
• The speaker also expressed interest in crypto funding models for rare-disease research and other areas where traditional sources of capital may be limited.

Takeaways

• The discussion suggests interest in decentralized funding for specialized research, but it does not provide a specific thesis or valuation for HairDAO.

Helium (HNT)

• Helium was described as a crypto asset the firm had historically owned; the speaker did not say it was a current holding.

Takeaways

• Treat this as historical portfolio context, not an indication of a current position or recommendation.

Kinetic (Crypto Asset)

• Kinetic was named among the firm’s crypto holdings, without further detail about its investment thesis.

Takeaways

• The transcript confirms a position but provides no basis for assessing its outlook, risks, or valuation.

Chainalysis (Private)

• The speaker said he had been trying to buy Chainalysis shares for about a year and a half, and noted that the company had traded at a steep discount in its last round.
• His positive thesis is that the company could benefit if crypto remains important and cybercrime continues to grow. He views Chainalysis as having a strong position in crypto security and difficult-to-win customer relationships.
• He described the company as having only a small number of meaningful competitors.

Takeaways

• The speaker sees Chainalysis as an attractive private-market opportunity, but this is not a publicly traded stock.
• The thesis depends on continued crypto adoption and rising cybercrime; the transcript provides no valuation or expected return.

Other Crypto Projects and Businesses

• Vibe Bio and other decentralized-science projects were discussed as possible ways to fund rare-disease research.
• The speaker said the firm was an investor in HairDAO and expressed interest in the broader category, but did not state a position in Vibe Bio.
• The speaker described crypto as potentially useful for organizing capital and creating more efficient systems through smart contracts.
• He also said tokenization can weaken long-term incentives, a concern when evaluating token-based business models.

Takeaways

• The discussion points to possible crypto applications in research funding and financial coordination, while highlighting the need to assess whether token incentives support durable value creation.

Private AI and Deep-Tech Companies

• The speaker cited several portfolio companies as examples of investments in technically complex businesses:
  • Wayve (referred to as “Wave” in the transcript), a self-driving company, and Runway, an AI-model company, were described as having grown into multi-billion-dollar companies.
  • Prime Intellect began with distributed computing and now works on post-training open-source models.
  • General Intelligence Company is developing an agent platform intended to help enable single-person or agent-native companies.
  • Orbital Materials uses AI and laboratory work to develop new materials; one application is a modular data center designed around cooling GPUs.
• The speaker favors companies with durable advantages across multiple innovation cycles, rather than businesses that may benefit only from the current surge in AI spending.
• He warned that AI demand can make companies look strong even when they are growing more slowly than the overall market, and that infrastructure businesses may face consolidation, margin erosion, and competition from firms that expand across the stack.

Takeaways

• The investment framework emphasized technical depth, long-term durability, and a clear path from research to a commercial product.
• For AI businesses, consider whether growth is driven by lasting differentiation or by a rapidly expanding, subsidized market.

Biotech, Healthcare, and Drug Development

• The speaker expects AI to contribute to major medical advances and estimated that around half of the major diseases that kill people today might not be fatal in 20 years. He also suggested that people alive today could live to 120 or more; these were personal forecasts, not established outcomes.
• He highlighted drug-discovery platforms, new drug types, lab automation, and mental-health care as areas of interest.
• The firm’s latest investment was described as an AI-oriented psychiatry business. The speaker also cited Talkspace as an earlier investment in video- and text-based therapy.
• He expects AI to make drug development safer and faster, potentially lowering costs and enabling regulatory changes. He also cautioned that biological knowledge remains difficult to replicate and that self-experimentation could have negative consequences.
• Human challenge trials—where participants are exposed to a disease and then given a treatment under controlled conditions—were described as an underdeveloped opportunity, mainly used today for respiratory viruses.
• The speaker expects regulatory reform to create opportunities for new biotech companies, while recognizing that established pharmaceutical firms have advantages in clinical trials and regulation.

Takeaways

• The discussion is broadly optimistic about AI-enabled medicine, but the potential benefits remain dependent on scientific evidence, clinical outcomes, and regulatory pathways.
• The speaker specifically flagged the risks of insufficient biological understanding and unsupervised experimentation; biotech investments can also depend heavily on clinical results.

Energy, Materials, and Industrial Technology

• Pantelosa, a portfolio company, was described as generating energy from ocean waves.
• The speaker sees energy, food supply, and new materials as important areas, while noting that energy investing has become crowded.
• He described a broader opportunity in developing new materials that can enable new products, rather than investing only in materials for their scientific novelty.
• Examples discussed included materials for data-center cooling and alternative manufacturing inputs. The speaker said the firm had considered, but did not invest in, a company developing new bullets.

Takeaways

• The investment case is strongest when a technical breakthrough is tied to a product and a route to market.
• The speaker warned that deep-tech businesses can depend heavily on suppliers, customers, or other intermediaries, any of which can constrain commercialization.

SaaS and AI-Driven Software Disruption

• The speaker expressed a generally short-biased view toward SaaS, but said that some companies could remain attractive.
• He described seeing a sharp increase in open-source projects that clone public software products and argued that improved AI coding tools could accelerate this trend.
• He cautioned that rapid growth in AI infrastructure or software does not necessarily indicate durable competitive advantage: a company growing at 6× while its market grows at 10×, for example, may be losing ground despite impressive headline growth.

Takeaways

• Evaluate software companies on their ability to outperform their markets and maintain differentiation, not just on fast growth in a rapidly expanding category.
• The transcript does not support treating every SaaS company as equally vulnerable; the speaker explicitly allowed for a smaller subset that could remain compelling.

Crypto Market and Investment Themes

• The speaker said crypto had been in a difficult period for nearly two years and that each market cycle requires a stronger reason to bring participants back.
• He viewed the lack of builders and talent as a key challenge, with many people choosing AI instead.
• He said regulatory uncertainty had eased somewhat, but identified token design and long-term incentives as continuing concerns.
• He sees some appeal in liquid crypto assets whose tokens are beginning to accrue value in more understandable ways, while noting that venture opportunities are limited outside areas such as new perpetual-trading platforms.

Takeaways

• The discussion suggests a selective approach to crypto: look for clear use cases and mechanisms through which activity can create value for the token.
• The speaker’s concerns about talent, adoption, and long-term token incentives remain important risks; he did not give a market-wide price forecast.