• Treat AI model providers cautiously: adoption may not translate into durable pricing power, and heavy infrastructure spending could limit long-term returns.
• For richly valued private-company exposure such as Cerebras, consider trimming or hedging rather than making an all-or-nothing bet; the cited $25–30 billion valuation was hard to justify against its next four to five years of prospects.
• Compare public companies such as NVIDIA and Uber with private-market alternatives on fundamentals and valuation; the discussion viewed them as comparatively inexpensive but gave no specific buy targets.
• Seek selective opportunities in European companies with strong growth and unit economics, while weighing business quality rather than relying on lower valuations alone.
• Set position limits and a liquidity plan for concentrated or private investments, and consider gradual profit-taking when holdings become oversized or valuations appear frothy.

SpaceX

• SpaceX once represented more than 20% of WashU’s portfolio and was described as a major winner.
• The team chose to hold the position through multiple opportunities for liquidity, including continuation vehicles, and aimed to keep exposure through an IPO.
• Wilson said the university would take some chips off the table after a major holding becomes liquid, and may hedge when it considers the valuation frothy. For a position as large as SpaceX, the plan is to sell gradually over several years rather than exit all at once.

Takeaways

• SpaceX illustrates the potential payoff from allowing a high-conviction investment to compound, but also the risks of extreme concentration.
• For investors, the practical lesson is to have a plan for position sizing and liquidity before a holding becomes a large share of a portfolio. The transcript does not offer a specific recommendation to buy SpaceX.

Anthropic

• WashU has meaningful exposure to Anthropic through several investment partners.
• Wilson questioned whether AI model providers can sustain strong economics if competing models become interchangeable and free alternatives reduce customers’ willingness to pay.
• WashU has held the exposure partly because of its relationships with the investment partners; it may hedge after a public listing if the valuation seems too high.

Takeaways

• The discussion points to a key risk for AI model companies: strong adoption may not automatically translate into durable pricing power or profits.
• Investors considering this theme should distinguish demand for AI from the ability of any one model provider to defend its margins.

OpenAI

• Wilson said OpenAI could potentially have raised private capital at a valuation above $2 trillion if it had stayed private, but he was not convinced public markets would assign the same valuation.
• He also used OpenAI as an example of how large private-market valuations can make companies reluctant to list publicly.

Takeaways

• The transcript highlights a gap between private-market pricing and what public investors may be willing to pay.
• A high private valuation is not, by itself, evidence that an investment can deliver attractive returns; the eventual exit price matters.

Cerebras

• WashU had a large position in Cerebras around its IPO and took some gains and hedged part of the exposure.
• Wilson said the company’s valuation—then discussed as roughly $25–30 billion—was difficult to justify based on recent performance and expectations for the next four to five years.
• He remained positive about the company’s leadership and long-term potential, while saying the business still had a lot to prove.

Takeaways

• The discussion supports reassessing a company’s valuation against its operating results, even when confidence in its leadership remains high.
• Wilson’s approach was to reduce risk rather than make an all-or-nothing decision: take some gains, hedge where appropriate, and retain some exposure.

Nubank (referred to as “New Bank” in the transcript)

• WashU held the company from its private-company stage and continued to own it nearly five years after it went public.
• Wilson spoke positively about the company and its leadership, but did not give a price target or specific buy recommendation.

Takeaways

• The example shows that an IPO does not automatically require an investor to sell; the decision can depend on the company’s prospects and valuation.
• The transcript does not provide enough detail to assess Nubank’s current valuation or make a specific investment call.

NVIDIA and Uber

• Wilson cited NVIDIA and Uber as examples of public companies he viewed as comparatively inexpensive on fundamentals, in contrast with some highly valued private companies.
• He did not provide valuation figures or a specific recommendation for either stock.

Takeaways

• Compare public and private opportunities using business performance and valuation, rather than assuming that private companies are automatically better investments.
• The comments express a relative valuation view, not a specific buy signal for NVIDIA or Uber.

European Investments

• Wilson said WashU had recently invested, through a partner, in a fintech company in Spain and had been active in Europe more broadly.
• He argued that some European companies with strong growth and unit economics can trade at lower valuations than comparable U.S. companies.
• The discussion also mentioned a global fintech company based in London and Singapore, without naming it.

Takeaways

• The investment opportunity discussed is to look beyond the U.S. for businesses with attractive fundamentals at lower entry prices.
• A lower valuation is useful only if the company can sustain its growth and business quality; Wilson emphasized comparing price with underlying value.

China Investments

• WashU has an office in Shanghai and, according to Wilson, has been more active in China than many peers, including through co-investments.
• He said the region can offer attractive opportunities when a company can return capital to shareholders and has a pragmatic approach to operating in its political environment.
• Risks raised included geopolitical uncertainty, political intervention and the need to understand local market conditions. Wilson said investors should work through local partners.

Takeaways

• China was presented as an opportunity for selective, locally informed investing—not as a broad recommendation to invest in the market.
• Any potential return should be weighed against the political and geopolitical risks discussed in the episode.

AI and AI Infrastructure

• Wilson expected a shakeout among AI-related companies, saying it was unclear how many would still exist in ten years, how defensible their advantages were, or where long-term profits would accrue.
• He also questioned the long-term returns on the substantial capital expenditure required by some AI businesses.
• At the same time, he saw opportunities in public companies tied to AI infrastructure, noting that sharply higher chip and memory prices can boost near-term profits.
• WashU was directing liquid capital toward public-market opportunities, including non-AI areas it viewed as relatively protected from AI.

Takeaways

• The transcript suggests separating the AI theme into different businesses: model providers, infrastructure suppliers and companies less exposed to AI disruption may have different risks and economics.
• Be cautious about valuing companies on rapid growth alone. The discussion specifically raises uncertainty about defensibility, long-term profit pools and returns on heavy capital spending.

Venture Capital Funds and Private-Market Strategy

• WashU prefers concentrated exposure to a smaller number of partners and companies, and generally favors smaller funds over very large funds.
• Wilson said mega-funds may find it harder to reproduce historical investment multiples as they deploy larger amounts at higher valuations or make more investments.
• He said WashU does not invest in funds above $10 billion, while noting that it does invest in funds above $1 billion.
• Wilson wished WashU had a larger allocation to Benchmark and described Addition as one of the strongest investors in its portfolio.
• WashU stopped following Sequoia as closely after a fund-structure change it felt did not fit its co-investment model. Wilson acknowledged that passing on the opportunity may have been a mistake given Sequoia’s returns.
• The team evaluates fund exposure by looking through to the underlying companies, rather than focusing primarily on a fund’s headline performance.

Takeaways

• The discussion favors evaluating funds by their underlying holdings, valuations and access to high-conviction companies—not just brand name or past returns.
• Smaller funds may have more room for a single successful investment to drive returns, but concentrated portfolios can also magnify losses.
• WashU’s fund decisions reflect its own scale, liquidity needs and co-investment strategy; they are not a direct template for individual investors.

WashU’s Portfolio Construction and Lessons Learned

• Wilson described a broad target of roughly half public equities and half private investments, while noting that WashU had recently been above that private-market target and was moving back toward it as holdings became liquid.
• The university’s annual payout is roughly 5%, which creates a meaningful need for liquid assets. Wilson said that without required payouts, he might target 70–80% in private investments.
• More than half of the private portfolio was in co-investments and direct investments. The team spends most of its time meeting companies and typically invests only in a small fraction of those it reviews.
• Wilson said one of WashU’s biggest mistakes was failing to add to winning investments because they had already grown in value or become large positions.
• He also regretted delaying a major secondary sale in 2020–21 in hopes of getting higher prices; the market later fell, and the team missed an opportunity to sell at favorable prices.
• He cited concentration in two manufacturing investments that performed poorly as examples of painful underwriting mistakes.

Takeaways

• Liquidity needs should shape how much an investor commits to private assets; attractive long-term returns do not remove the risk of being unable to sell when cash is needed.
• The episode emphasizes both letting strong investments compound and taking gains when market conditions appear unusually favorable. Those aims can conflict, so portfolio limits and a clear liquidity plan matter.
• Direct investing requires the ability to understand, value and monitor individual businesses. Wilson described it as a skill set that has to be built, not an easy shortcut to better returns.