AI Turns Specialists into Generalists

"We all had to be specialists until AI. Now we're all generalists."

The talent agent and dealmaker Michael Ovitz states the claim in its sharpest form: for his entire career, business ran on vocational specialization, in entertainment you picked film or television or books, then a subdivision of film, and stayed there. Today, decades later, he works simultaneously in consumer products, high technology, defense, medicine, and biopharma, fields he calls himself "totally unqualified for," because AI lets him "learn about it really fast and have enough vocabulary to understand what these brilliant young minds are creating."¹

The measured compression

The claim comes with a number. Preparing to meet the C-suite of Matsushita in the late 1980s, a company with $15 to $20 billion of balance-sheet liquidity, took four weeks of analog work: a stockbroker's office, the library, a physical copy of the 10-K. The equivalent preparation for two current transactions took six to seven hours across three AI platforms over two days, layered on decades of instinctual knowledge of the business. That is roughly a hundredfold compression of the learning-curve tax that used to force specialization.¹

The boundary condition matters as much as the claim: businesses that require extraordinary computational depth, such as Palantir and Anthropic, still demand deep-science specialists. AI generalizes the business layer, not the frontier-research layer.

Ovitz is also careful to separate temperament from tool. Asked how much of his range is AI-driven, he answers "none, I've been like this my whole life," then immediately grants that AI gives someone with that disposition "superpowers, even more additional leverage, absolutely." The generalist instinct predates the technology; the ability to act on it across five industries at once does not.

The founder version

Eric Glyman makes the same argument from the opposite direction. His determined-generalist thesis holds that large language models have read more code, case law, and filings than any single human specialist alive, so determination replaces craft boundary as the binding constraint on what a person or a company can attempt. Where Ovitz demonstrates what the shift does to an individual career at the far end of seniority, Glyman argues it changes who a company hires and how it organizes itself. The two arrive at the same mechanism from opposite ends of a career, an aging dealmaker and a fintech founder, which is part of why the claim is worth taking seriously rather than treating as generational bravado.