The case for coming second (Semafor)

Satya Nadella has stopped trying to make anyone dance.

Interviewed for Semafor's Future Proof, the Microsoft chief executive called the frontier AI labs "magical businesses" while arguing this is "not an industry where your initial success means anything". Four years on from ChatGPT, he noted, AI's impact outside software development has been incremental rather than exponential, and Microsoft's job is to build what he called a sturdy dance floor rather than chase the labs on their terms. That is a notable position from the company whose investment in OpenAI helped start the boom.

There is sound strategic logic in it. If model capability keeps improving quickly and competitors keep matching it, models become interchangeable, and value moves to whoever owns the customer relationship, the data and the workflow. That is precisely where Microsoft already sits through its enterprise software.

The counter-case deserves weight. A fast follower can wait too long: if one lab pulls decisively ahead, companies built on other people's models will negotiate from weakness. Nadella is also describing the strategy his company already has, and a valuation close to US$4 trillion, as at the 21 October article, rewards a story investors are happy to hear. His scepticism about the labs and his self-interest point in the same direction, which does not make him wrong, but it is worth noticing.

For Australian executives the same choice arises at a smaller scale. Most organisations here will never build a model, so the question is where their advantage sits once every competitor has access to similar tools. If the AI model you rely on today were available to every rival tomorrow, what would still be yours?

#AI #Strategy #CompetitiveAdvantage

Source: Semafor, "Microsoft's Nadella is back in business" (Reed Albergotti, Future Proof interview, 21 Oct 2026) https://lnkd.in/gaxwPHtk