The Google Has a New Sheep Herder: An AI Boss to Nip at the Heels of the AI Beasties
December 17, 2025
Another dinobaby post. No AI unless it is an image. This dinobaby is not Grandma Moses, just Grandpa Arnold.
Staffing turmoil appears to be the end-of-year trend in some Silicon Valley outfits. Apple is spitting out executives. Meta is thrashing. OpenAI is doing the Code Red alert thing amidst unsettled wizards. And, today I learned that Google has a chief technologist for AI infrastructure. I think that means data centers, but it could extend some oversight to the new material science lab in the UK that will use AI (of course) to invent new materials. “Exclusive / Google Names New Chief of AI Infrastructure Buildout” reports:
Amin Vahdat, who joined Google from academia roughly 15 years ago, will be named chief technologist for AI infrastructure, according to the memo, and become one of 15 to 20 people reporting directly to CEO Sundar Pichai. Google estimates it will have spent more than $90 billion on capital expenditures by the end of 2025, most of it going into the part of the company Vahdat will now oversee.
The sheep dog attempts to herd the little data center doggies away from environmental issues, infrastructure inconsistencies, and roll-your-own engineering. Woof. Thanks, Venice.ai. Close enough for horseshoes.
I read this as making clear the following:
- Google spent “more than $90 billion” on infrastructure in 2025
- No one was paying attention to this investment
- For 2025, a former academic steeped in Googliness will herd the sheep in 2026.
I assume that is part of the McKinsey way, Fire, Aim, Ready! Dinobabies like me with some blue chip consulting experience feel slightly more comfortable with the old school Ready, Aim, Fire! But the world today is different from the one I traveled through decades ago. Nostalgia does not cut it in the “we have to win AI” business environment today.
Here’s a quote making clear that planning and organizing were not part of the 2025 check writing. I quote:
“This change establishes AI Infrastructure as a key focus area for the company,” wrote Google Cloud CEO Thomas Kurian in the Wednesday memo congratulating Vahdat.
The cited article puts this sheep herder in context:
In August, Google disclosed in a paper co-authored by Vahdat that the amount of energy used to run the median prompt on its AI models was equivalent to watching less than nine seconds of television and consuming five drops of water. The numbers were far less than what some critics had feared and competitors had likely hoped for. There’s no single answer for how to best run an AI data center. It’s small, coordinated efforts across disparate teams that span the globe. The job of coordinating it all now has an official title.
See and understand. The power consumption for the Google AI data centers is trivial. The Google can plug these puppies into the local power grid, nip at the heels of the people who complain about rising electricity prices and brown outs, and nuzzle the folks who:
- Promise small, local nuclear power generation facilities. No problems with licensing, component engineering, and nuclear waste. Trivialities.
- Repurposed jet engines from a sort of real supersonic jet source. Noise? No problem. Heat? No problem. Emission control? No problem.
- Brand spanking new pressurized water reactors built by the old school nuclear crowd. No problem. Time? No problem. The new folks are accelerationists.
- Recommissioning turned off (deactivated) nuclear power stations. No problem. Costs? No problem. Components? No problem. Environmental concerns? Absolutely no problem.
Google is tops in strategic planning and technology. It should be. It crafted its expertise selling advertising. AI infrastructure is a piece of cake. I think sheep dogs herding AI can do the job which apparently was not done for more than a year. When a problem becomes to big to ignore, restructure. Grrr or Woof, not Yipe, little herder.
Stephen E Arnold, December 17, 2025
Comments
Got something to say?
-
- Subscribe to Beyond Search
Feature archive
News archive
-
