Silicon Valley venture capitalists are questioning whether research-first AI startups can justify funding targets as high as $50 billion amid competition from established frontier labs.

Can neolabs make the economics work?

Neolabs are lavishly funded, research-first AI companies founded by talent recruited from OpenAI, Google DeepMind and similar institutions. Their rounds now carry multibillion-dollar price tags, increasing the burden of turning laboratory breakthroughs into durable businesses.

Jeff Dean, Google's former chief scientist, is seeking a valuation as high as $50 billion for his new AI science startup. Safe Superintelligence, founded by OpenAI cofounder Ilya Sutskever, is worth $32 billion, while OpenAI alum Mira Murati is raising money for Thinking Machines Lab at around $40 billion.

The division among investors is widening. Some see neolabs as the only credible path to owning the next layer of AI models or hardware; others doubt that companies can earn adequate returns after burning multibillion-dollar rounds and fighting frontier laboratories with deeper resources.

One venture capitalist at a top firm, speaking anonymously, called neolabs the most intensely debated investment category. He said:

“Some would say you shouldn't invest in any, and others would say this is basically all you should be investing in because software's dead, and you either have to own models or hardware, and even if you enter at a billion, the company could be worth a trillion.”

At Business Insider's Growth Mode conference, Menlo Ventures partner Matt Murphy immediately named neolabs as the most overhyped sector, a pointed answer because Menlo has invested in several of them. Murphy, an early Google board adviser, compared Google's early success to throwing a match into a bonfire.

He warned: “The opposite thing you want to do is start a match with a bonfire. That's the risk of what's going on now with companies raising such large rounds.”

Menlo partner Deedy Das wrote last month about the difficult game of making neolab economics work. Das appeared taken aback by the reaction to his post and later said many startups will fail while some will work, restating the power-law wager rather than abandoning it.

Radical Ventures partner Aaron Rosenberg called neolabs a power-law bet taken to an extreme. Investors should expect most to burn enormous sums on compute and talent without becoming profitable, while a rare winner could become a foundational AI company worth hundreds of billions or more.

Rosenberg remains confident the strongest neolabs will attract ongoing support because leading talent continues moving from established laboratories and academia into founder roles. He sees no sign that flow slowing.

New Enterprise Associates wrote a bullish neolab assessment in June. Partner Andrew Schoen said his view had not changed and reported that neolabs he knew of were finalizing equity rounds above $1 billion, with the processes moving smoothly and quickly.

Reporting from Venice Beach, Business Insider chief correspondent Ben Bergman plans to interview Nadia Carlson in New York next week at The Long Play: Stories of Reinvention. Carlson is trying to turn Allbirds into an AI infrastructure company.

What did Menlo's anniversary event show?

Menlo Ventures marked 50 years with a black-tie gathering at the Ritz-Carlton in San Francisco on Saturday. Editor at large Katie Roof reported that public-company and startup leaders attended alongside investors and bankers.

DreamWorks alumnus and venture capitalist Jeffrey Katzenberg jokingly urged the audience to pray for a blockbuster Anthropic IPO, a reference to Menlo's early investment in the AI company.

Morgan Stanley technology banker Michael Grimes attended, alongside CEOs from Snowflake, Zscaler, Legora and OpenRouter. Menlo partners Shawn Carolan, Matt Murphy and Venky Ganesan also attended.

Ganesan joked that the room held more assets under management than many countries, reflecting the attendees' personal wealth and the investment capital they managed.

Menlo was not the first: Index Ventures recently staged a large celebration for its 30th anniversary, and an attendee at a rival venture firm expressed disappointment that their own shop did not hold a comparable event. Menlo closed its party with concerts by two secret Grammy Award winners, with the newsletter joking that they were there to cash in on AI money.

Which deals and startups stand out?

• Fireworks AI is seeking new funding at a valuation of $30 billion to $40 billion, a sharp increase only months after closing its previous blockbuster round.
• The documentary You Can See Everything is drawing attention, and Ben Bergman called it ingenious but somewhat long. Ben expects the film to enliven this year's Oscars race and recommends a theatrical viewing before online spoilers.
• Rya Jetha examined the promise and risks of investing in robotics startups, which, like neolabs, command large valuations despite having little to show for it so far.
• Foursquare's $0 acquisition produced no payout for employees or investors, according to founder Dennis Crowley.

How quickly are decacorns being created?

An October 21, 2026 post from Rory O'Driscoll counted only 130 US technology companies, public or private, valued at $10 billion or more. That equates to roughly 20 to 30 new companies per decade across the last five decades.

With one quarter of 2026 left, 42 decacorns had already been minted privately. O'Driscoll argued that the pace could represent the most economically transformative event since the industrial revolution, or that investors are getting ahead of themselves.

For neolab backers, the unresolved test is stark: identify the rare winner before compute and talent spending consume its path to profit.