VCs on AI investing: M13 and Basis Set share pricing strategies

Key Takeaways

• ChatGPT's revenue jumped from $1B to $40B in six months, resetting expectations for what 'good growth' means in AI startups
• M13 and Basis Set focus on 'friction as a moat' — regulated industries where hyperscalers won't compete quickly
• Infrastructure below and applications above AI models offer the best defensibility for early-stage companies

Two veteran AI investors laid out their playbook for pricing deals in a market moving faster than anything venture capital has seen. Carter Reum of M13 and Chang Xu of Basis Set Ventures, speaking at TechCrunch's StrictlyVC event in Los Angeles, agreed on one thing: the old math doesn't work anymore.

M13 manages $2.5 billion and has backed 17 unicorns at seed or Series A. Basis Set, which launched in 2017 as one of the first funds focused exclusively on AI, is deploying its fourth fund with nearly $1 billion in cumulative assets. Both see opportunities in the current environment. Both admit they're worried.

Is AI infrastructure in a bubble?

Xu's answer: yes and no. The growth curves are real. ChatGPT went from $1 billion to $40 billion in revenue in six months. One Basis Set portfolio company, Open Art, hit $1 million ARR in its first year, $10 million in year one, and $70 million in year two. Cash-flow positive the whole time. Twenty employees.

"The bar for what is good growth has totally changed," Xu said. "When you have this possibility of compounding accelerant growth, the valuations don't seem so crazy because you price that into the terminal value."

But here's the catch: if you price every deal assuming that trajectory, your portfolio blows up. Not every company will compound at that rate. Most won't.

Reum pushed back on the idea that this cycle is entirely new. Cloud, the iPhone, even the automobile in the 1920s created similar panic. Jobs were lost. Life went on. What's different now is the competition.

"Past cycles had innovators competing with innovators," he said. "Zuck versus Evan, Travis versus John Zimmer. In this cycle you have innovators competing with innovators, competing with the largest, most well-funded innovators the planet has ever seen, and competing with the ten largest tech companies on the planet."

For the first time, Reum argues, incumbents have the advantage. They have the tech, the capital, the data, and the talent. Startups can rise fast. They can fall faster.

How do you price a deal when revenue growth is real but defensibility isn't?

Reum still does what he calls "cocktail napkin math." He recently looked at an AI software company targeting brands. His questions: How big were the winners last cycle? Will there be more brands in the world? Will they pay double or triple for software? The math didn't check out. M13 passed.

Xu's framework is more technical. Basis Set stays close to what she calls "defensible technical differentiation." The problem: that frontier moves every quarter, sometimes every week.

Her mental model splits opportunities into two zones. Below the AI layer sits infrastructure. Databases, version control, deployment tools. All of it was built for human developers. Agents need something different. Last year, Xu wouldn't have predicted anyone would need a new GitHub. This year, she counts multiple strong teams building exactly that.

Above the AI layer, where applications sit, the market gets crowded fast. The only question that matters: what's defensible long-term?

Where won't OpenAI and Google go?

Reum's thesis centers on friction. He loves regulated industries. M13 had a near-billion-dollar exit from a company disrupting 911 call centers with AI. The hyperscalers could go there eventually. But for a few-billion-dollar outcome, it's not worth their time.

Healthcare fits the same pattern. Google and OpenAI will enter eventually. Regulation slows them down. That window matters.

"What keeps all of us up at night is that it can change on a dime," Reum said. "You used to see them coming in the rearview mirror. I tell every founder: you need a microscope in one eye and a telescope in the other."

The SpaceX factor and what it means for L.A. tech

The conversation touched on a coming shift for the Los Angeles tech scene. SpaceX is preparing for an IPO, and both investors expect it to create a wave of newly liquid wealth. Some of that money will flow into angel investing and early-stage funds. Some will chase AI.

L.A. has traditionally lagged San Francisco in venture density. A SpaceX liquidity event could narrow that gap, particularly for hardware-adjacent AI companies where Southern California already has concentration.

What investors are actually worried about

Neither Reum nor Xu claimed to have the market figured out. The honest answer: this is harder than previous cycles. The upside is bigger. The downside is faster. A company that looks defensible in January may be commoditized by June.

Basis Set's edge comes from focus. Eight years of AI-only investing means the team has seen technical shifts before they hit the mainstream. M13's edge comes from pattern recognition across 17 unicorns. Both admitted those edges may not be enough.

The paradox Xu described is real. Every deal priced to perfection is a bet that this company, specifically, will be one of the winners. History says most won't be. But the few that are will generate returns that make the math work anyway.

Frequently Asked Questions

How much does M13 manage in assets?

M13 manages $2.5 billion in assets and has been a seed or Series A investor in 17 companies that reached unicorn status.

What is Basis Set Ventures' investment focus?

Basis Set launched in 2017 as one of the first early-stage funds focused exclusively on AI. The firm is now investing from its fourth fund with nearly $1 billion in cumulative assets under management.

How do VCs avoid investing in companies that hyperscalers will crush?

Both M13 and Basis Set look for "friction as a moat." Regulated industries like healthcare and emergency services create barriers that slow down Google, OpenAI, and other large players.

What growth rates are AI startups achieving?

ChatGPT reportedly grew from $1 billion to $40 billion in revenue in six months. Basis Set portfolio company Open Art went from $1 million to $70 million ARR in two years while staying cash-flow positive with 20 employees.

Where are the best AI investment opportunities in 2026?

According to these investors, infrastructure "below" AI models (developer tools rebuilt for agents) and applications "above" AI models in regulated verticals offer the best defensibility.

Source: Venture Capital News | TechCrunch / Connie Loizos