Alex Davis on Concentrated Investing, AI Compute, and Private-Market Risk

Alex Heath asks investor Alex Davis where he sees the largest technology opportunities and risks. Davis points to US open-weight AI models and the infrastructure needed to make scarce compute available.

He says neoclouds can play an important role even at high valuations, provided they can actually deploy and operate compute. Financing is another part of the opportunity, in his view.

Heath asks whether announcing a $10 billion growth fund at a time of high valuations gives Davis pause. Davis argues that a concentrated approach is particularly suitable when a small number of private companies attract much of the capital and talent.

He says the universe of highly valued private companies has grown too large for any investor to cover comprehensively. Disruptive instead tries to follow a much smaller set of companies it considers exceptional.

As leading companies stay private longer and require larger amounts of capital, Davis believes they need investors able to write concentrated checks. He contrasts that with spreading small investments across a very large portfolio.

Davis attributes some of the apparent froth to the cost of competing in AI. A company seeking very large GPU clusters cannot finance them with a small amount of friends-and-family money, he says.

Heath asks whether improvements in model architecture could reduce compute needs dramatically. Davis welcomes efficiency but says he has not yet seen convincing evidence that frontier competition will become inexpensive.

Davis says benchmark claims about very cheap models need to be tested against the requirements of real institutional or sovereign customers. He presents this as an investor’s judgment, not a technical proof that efficiency gains are impossible.

Disruptive began by investing in private companies that appeared to be on a path toward public markets. Davis describes organizing individual opportunities for friends and institutional investors before building the larger firm.

His approach is to look for companies that already show product-market fit or technical progress, rather than select the earliest unproven ideas. He acknowledges that the earlier-stage task is harder and that his strategy depends on prior venture investment.

The team follows a short list, gets to know the companies and looks for ways to help beyond funding. Davis says it has generally done better when it proactively found a company than when it relied on unsolicited pitches.

For the new fund, he describes a target of roughly ten investments and a theoretical average of about $1 billion each over the deployment period. Those are the fund’s stated intentions, not guaranteed outcomes.

Heath asks why Disruptive is shifting from individual special-purpose vehicles to a discretionary fund. Davis says the private-market landscape has changed, and companies increasingly want clarity about who is providing capital.

He worries that some entities market access to popular private companies without adequate transparency. In his view, committed fund capital can make the source of money clearer to a company choosing among oversubscribed investors.

Asked whether investors may be hurt by some SPVs, Davis says yes, though he does not supply a specific proven example. He recounts brokers allegedly offering shares his firm already owns and says he fears misrepresentation or fraud in parts of the market.

The long period between investment and a private company’s exit can delay discovery of problems. Davis says regulatory custody checks help, but he remains concerned about arrangements that do not receive comparable scrutiny.

Heath asks whether more transparent and accessible private markets could reduce information asymmetry. Davis replies that young companies often keep information close while preparing to become public, and says the current private-market structure suits his firm.

He notes that companies go public only once, whereas an investor sees many such decisions. He says that repeated exposure to milestones, growth and listing decisions helps his firm judge how companies are maturing.

Davis says Disruptive updates a list of roughly 15 to 20 potential investments. A company may enter or leave the list as the team’s understanding changes; the firm does not treat a higher later valuation as an automatic reason to pass.

He says the firm can become more confident and write a larger check at a higher valuation when a business has made progress. His argument is based on expected outcomes and risk, rather than a claim that the entry price never matters.

The other half of the work, Davis says, is helping companies reach customers, enter countries and navigate important firsts. He believes an investor needs to understand a company deeply enough to explain its value in plain language.

He describes using longstanding relationships to introduce portfolio companies to senior decision-makers. Those introductions can save time, he says, although the usefulness of the approach varies by company and market.

Heath turns to Palantir, an early and important investment for Disruptive. Davis says his confidence centered on the team and its ability to recruit engineers, even when potential customers did not yet have budgets for its products.

He recalls traveling with the company and watching it work through internal resistance at customer organizations. The team’s challenge was not always an external competitor, but explaining why a new kind of product was needed.

Around the planned public listing, Disruptive invested more alongside large partners. Davis says the decision reflected years of familiarity with the company and his view that the public would come to understand its leadership.

He particularly describes Alex Karp’s effect on the team and customers as a reason for conviction. Davis acknowledges the risk of losing key employees during a listing, and says his investment was ultimately a bet on people and their execution.

The interview moves to Groq, the inference-chip company. Heath describes its licensing arrangement with NVIDIA and the movement of core team members; Davis explains why he became involved in the remaining business.

Davis says he supported the company’s shift toward an accelerator and the people leading it. After the licensing arrangement, he looked at the remaining infrastructure and whether it could support a new business rather than simply winding down leases.

He says he helped oversee shareholder distributions and took responsibility for the board, then became executive chair. His account emphasizes both the unusual transaction and the practical work required after it.

The plan he describes is to upgrade existing capacity, adapt facilities for newer cooling and chips, and build an inference-oriented neocloud. He says portfolio companies with compute constraints could be customers, alongside other organizations.

Davis says he does not know whether he will run Groq over the long term and expects additional leadership to formalize its future. He describes the role as an unusual level of operational involvement for an investor.

On neocloud competition, he argues that demand for usable compute is abundant. The challenge is delivering working capacity on time, with credible funding, power, hardware and operational expertise.

He warns that customers may pay deposits against promises of capacity that prove difficult to deliver. In his view, ready-to-use clusters can command a premium, while execution failures can harm customers and financiers.

He distinguishes experienced owner-operators from speculative data-center landlords. Someone with land but insufficient funding or knowledge can be exposed to construction delays, changing chip requirements and debt before the project produces usable compute.

Heath asks about Davis’s upbringing and his grandfather, entrepreneur Marvin Davis. Davis says early exposure to business and risk-taking was formative, and he now tries to explain his own deals to his children.

He recalls wanting to build his own business rather than rely on his family name. He attended Northwestern and spent time learning about markets and auctions, then left college after his grandfather died.

Looking ahead, Davis sees opportunity in defense technology as well as AI. He predicts newer private companies can become important defense suppliers and says safety and protection of corporate data may also become large markets.

He thinks open models may have a role in defense, while acknowledging that how the pieces fit together remains uncertain. For founders raising capital, he recommends a specific ask and a credible explanation of how the money will be used.

Davis says venture investing is difficult and necessary to his own later-stage business. Early investors take many risks before a few companies emerge as candidates for larger growth investments.

He worries that incentives can favor gathering assets and management fees over investment performance. He says Disruptive emphasizes carried interest and returns, and argues that managers should not become wealthy solely from management fees.

In a concentrated portfolio, Davis says each company matters. Rather than automatically abandoning an underperformer, his team tries to determine whether its relationships and operational help can improve the outcome.

Asked about AI at his firm, Davis describes investing as a relationship-based business. He says staff use generative AI to improve their work, but he does not plan to replace the team with agents.

Heath asks how he thinks about present valuations versus exponential future growth. Davis says regulation can worry him more than a small difference in entry price when he has conviction in an exceptional company.

He clarifies that price still matters, but argues that a small discount or premium has not been the main driver of his strongest historical outcomes. That is his investment philosophy, not a general assurance that high prices are safe.

Looking back, Davis says he might have supported some model companies more heavily, but is satisfied with the team’s work. He closes by describing the next few years as an opportunity to help companies and decide where to invest next.