AI Investing in a Fast-Moving Market: VCs Discuss Bubble, Pricing, and Defense Strategies

At a recent TechCrunch StrictlyVC evening in Los Angeles, two prominent AI-focused investors discussed the challenges of venture capital in a market moving faster than ever. Carter Reum, co-founder of M13, and Chang Xu, a partner at Basis Set Ventures, covered topics from the AI infrastructure bubble to how startups can survive competition from hyperscalers.

On whether there is an AI infrastructure bubble, Xu offered a nuanced view. She noted that while the growth curves of companies like ChatGPT are unprecedented, not every deal will follow that trajectory. Reum drew parallels to past technology cycles but emphasized that this cycle is unique because incumbents have advantages in technology, capital, data, and talent, making investing harder but potentially more rewarding.

Regarding deal pricing in a fast market, Reum described using 'cocktail napkin math' to assess numbers. He passed on an AI software for brands because the math didn't work. Xu stressed the importance of technical differentiation, as the frontier changes rapidly—sometimes weekly. She divides investments into 'below the AI' (infrastructure) and 'above the AI' (applications), noting new tools like version control for AI agents emerging.

To avoid being crushed by OpenAI, Anthropic, or Google, Reum advises targeting regulated industries where friction acts as a moat. He cited an AI company disrupting 911 call centers that achieved a near-billion-dollar exit. He told founders to balance daily execution with long-term vision. Xu uses a framework of 'depth markets' (where hard things remain hard) versus 'velocity markets' (where speed is key). She mentioned a portfolio company using transgenic chickens for drug manufacturing as a depth market example.

On the novelty of current ideas, both agreed that while many pursue consensus categories like AI agents for finance or healthcare, the most interesting opportunities often initially seem like no business. Xu pointed to OpenArt, which grew from $1 million to $70 million ARR in two years despite early skepticism.

Finally, the investors discussed the upcoming SpaceX IPO and its impact on Los Angeles. Reum predicted the liquidity event would spark a second wave of startups, similar to previous LA cycles producing Riot Games, Tinder, and Snap. He believes LA's strength in brand, content, and culture will become more valuable as AI shifts from technical waves to waves requiring taste and emotional resonance. Xu added that the next frontier in AI is not more compute but taste—something LA has in abundance.