Prentis: Рид Хоффман и Марк Пинкус запускают AI-лабораторию и привлекают $100 млн

New AI lab Prentis, co-founded by Reid Hoffman and Mark Pincus, is in talks to raise $100 million as of July 24, 2026 — TechCrunch reports. The startup is betting that automating routine computer tasks will soon overtake code-writing as the largest use case for artificial intelligence.

Who is behind Prentis

Prentis was founded by two Silicon Valley veterans — Reid Hoffman and Mark Pincus. Hoffman created LinkedIn in 2003 (Microsoft acquired the social network for $26 billion in 2016), remains a partner at Greylock, was an early investor in OpenAI, and in 2022 became a co-founder of the AI startup Inflection AI. Pincus founded gaming giant Zynga in 2007 and took it public in 2011. Hoffman also served on OpenAI's board of directors until 2023, so for him a bet on applied AI is a continuation of years of experience.

• Name — Prentis, a new AI lab (a "neolab")
• Co-founders — Reid Hoffman (LinkedIn, Inflection AI) and Mark Pincus (Zynga)
• Round — in talks for $100 million, according to TechCrunch
• Date — July 24, 2026
• Focus — automating routine computer work, not code generation

Why the bet isn't on coding

Prentis believes that automating everyday computer operations will become a more massive AI application than today's trendy AI coding. As of mid-2026, developer tools — GitHub Copilot, Cursor, Claude Code — remain the most visible and commercially successful niche of generative AI, but the Prentis team calls it narrow compared to the billions of ordinary office workers.

The neolab is betting that automating routine computer tasks will soon

overtake code-writing as the largest AI use case.

— TechCrunch's phrasing, July 24, 2026

This refers to a class of "computer-use agents" — systems that see the screen, click the mouse, fill out forms, and carry out multi-step tasks in place of a human. Such agents promise to serve not the developer niche, but the most massive audience — accountants, managers, support specialists who spend their whole day in a browser and office programs. OpenAI with its Operator mode, Anthropic with its Computer Use feature, and Google are already competing for this market — and this is exactly where Prentis is directing its $100 million.

For now, computer-use agents remain unstable: they stumble on CAPTCHAs, payments, and complex interfaces. It is precisely on this immature but potentially enormous territory that Prentis plans to build its research program.

How much it's raising and what a neolab is

According to TechCrunch, as of July 24, 2026, Prentis is in talks for a $100 million round. The company's valuation, lead investor, and deal stage have not been publicly disclosed — only the target amount and the founders' names are known.

TechCrunch uses the term "neolab" to describe a new wave of compact AI labs that don't build a product around a chatbot, but instead immediately declare a fundamental bet and raise large rounds for research. A sum of $100 million at an early stage is a signal of that scale: that much is needed not by an app, but by a "model factory" with expensive researchers and compute power. By comparison, product-focused AI startups at the seed stage typically raise single digits or the first tens of millions of dollars — Prentis's ask is immediately an order of magnitude higher.

What this means

Prentis's emergence shows where venture interest is shifting in 2026: from narrow AI coding toward universal agents that automate any computer-based work. The names Hoffman and Pincus, together with an ask of $100 million right off the bat, mean that the new "neolabs" intend to compete with the big players not through a single product, but through a bet on changing AI's primary use case. If Prentis's thesis is correct, the next wave of AI products will be measured not in lines of code, but in the number of routine operations taken off the hands of ordinary employees.