Satya Nadella on Open vs Closed Models: Who Wins | Why Companies Need to Own Their Own Intelligence

20VC with Harry Stebbings · · 24 min

Harry Stebbings interviews Microsoft CEO Satya Nadella on open versus closed AI models, why companies should own their own intelligence, and why he now thinks alignment won't come from scaling alone.

Watch the episode 24 min · YouTube

The gist

  1. Nadella says alignment won't fall out of scaling, so companies should contain agents in a zero trust setup with identities, tamper-proof logs and policy monitoring.
  2. He expects most token usage to be cheap commodity tokens from open weight or low-cost closed models, especially when firms train on their own traces.

Made by someone on this episode

Agent 365 · Satya Nadella's — Microsoft's system for giving AI agents identities and permissions ...

Also mentioned

2:02 Zero trust architecture · "The best posture to have, if you want to continue to use these sophisticated models, is to treat these..."
6:39 MAI models · "We will have even with the MAI models that will be not open weight. We'll talk, I think, of it more as..."
11:15 Harvey · "This is for a rarefied air of companies, so it would seem. Your Harveys or your Legoras or the kind of very..."
11:17 Legora · "Your Harveys or your Legoras or the kind of very forward-thinking AI companies."
11:39 Excel · "They were able to put in an Excel spreadsheet, send it in email, and avoid the fax machine and interoffice..."

Follow 20VC with Harry Stebbings

Free email updates