hunter-gplpcn
The founding team reads like a roll call of the internet’s first social era. Reid Hoffman, the LinkedIn co-founder who became one of artificial intelligence’s most visible investors, and Mark Pincus, the Zynga founder who built FarmVille into a cultural phenomenon, have joined forces to create a new AI laboratory called Prentis, according to people familiar with the matter. The company is in talks to raise $100 million.
The details of what Prentis will build remain closely held. People close to the company said the founders have been recruiting engineers and researchers for months, with the lab focused on consumer applications of AI rather than the raw model research that dominates the field’s biggest names. Neither Hoffman nor Pincus responded to requests for comment.
The move marks a return to the founder’s seat for Hoffman, who has spent the past several years placing bets on the AI industry from the investor’s side. He was an early investor in OpenAI, co-founded Inflection AI, and has backed a long list of AI startups through his venture firm Greylock. His decision to build again, rather than fund, signals a conviction that the consumer layer of AI remains up for grabs.
Pincus brings a different set of skills. He built Zynga into one of the fastest-growing gaming companies of its era, mastering the mechanics of engagement, virality and virtual goods that defined social gaming. Those instincts — how to make software that people use daily and pay for incrementally — are directly relevant to the consumer AI market, where retention and monetization remain unsolved problems.
The pairing has a logic that goes beyond reputation. Hoffman’s network runs deep in AI research and enterprise software; Pincus’s runs deep in consumer product and game mechanics. A lab that combines frontier model access with consumer distribution instincts is precisely the combination that the industry’s biggest players have struggled to assemble internally.
The $100 million figure would be modest by the standards of the AI arms race. Frontier labs have raised billions, and even mid-tier startups routinely close rounds of several hundred million. Pincus and Hoffman are betting that capital is not the constraint in consumer AI — product taste is — and that a focused team with a clear consumer thesis can move faster than a capital-saturated one.
The fundraising is still in negotiation, and the structure could change, people familiar with the talks said. The company has been in discussions with a mix of strategic investors and venture funds, with Hoffman’s existing relationships expected to smooth the process. A formal announcement is not expected until the terms are settled.
The founding team’s timing is deliberate. The AI industry is entering a phase where the models are becoming commoditized at the API level, and the value is shifting to the applications built on top. That shift plays to the founders’ strengths: both have built products that reached hundreds of millions of users, and both understand the consumer psychology that drives adoption.
The name Prentis carries no public explanation, and the company has kept a low profile during its early months. What is known is the ambition: to build an AI company that consumers actually want to use, rather than one that competes on benchmark scores. It is a bet that the next wave of AI winners will be defined by product design, not model size.
Hoffman has been outspoken about the direction of the industry. He has argued that AI will follow the pattern of the internet — infrastructure first, applications second, and consumer products third — and that the people who built the application layer of the web are best positioned to build the application layer of AI. Prentis is that argument made concrete.
The venture also reflects a broader trend: founders who made their names in the first internet wave are returning to build in AI. The industry’s talent pool is filling with veterans of social media, gaming and consumer software, drawn by the same dynamic that attracted them two decades ago — a new platform, a blank canvas and the chance to build something used by everyone.
Analysts said the Prentis founding will be watched closely for what it reveals about the market’s appetite for consumer AI. If Hoffman and Pincus can raise at their target valuation, it confirms that investors still believe product-led AI companies can generate outsized returns. If the round comes in lower, it suggests the market’s patience with new entrants is thinning.
For the two founders, the stakes are personal as well as financial. Both have fortunes tied to the social and gaming eras they helped create, and both have been looking for the next act. Prentis is that act — a bet that the tools they mastered building the social web still work in the era of machines that think.
The lab is early, the details are thin, and the product is unknown. What is clear is the signal: two of the most successful product builders of the consumer internet are betting their reputations that AI’s biggest prizes will go to the people who make it useful, not the people who make it smart.
