Bill Gates Says 'There's No Plan' For AI. Google's Org Chart Shows Why.

This week, Bill Gates published a 12-page essay warning that "there is no plan" for the upheaval AI is about to cause. Gates wrote that the AI transition will be one of history's most turbulent periods, and that leaders, experts, and communities aren't adequately preparing for it. Meanwhile, Google, according to the Wall Street Journal, is moving its roughly 90-person AI responsibility team out of DeepMind and into the global affairs organization — the division that runs the company's lobbying and public policy work.

Gates's essay is not a call to slow AI down. He argues white- and blue-collar jobs will both be hit fast, with entry- and mid-level roles going first, and that the speed and breadth of this shift has no real precedent. He wants a stronger social safety net, taxes on "AI tokens and bots" to fund retraining, and some jobs set aside for humans entirely. He's also explicit that the U.S. and China will need to coordinate, comparing it to past cooperation on nuclear weapons and aviation.

Google's DeepMind move starts at the beginning of September. The AI responsibility unit — which tests models for chemical, biological, radiological, and nuclear risk and studies the psychological effects of chatbots — reports into global affairs instead of DeepMind. This follows Demis Hassabis stepping aside as DeepMind's CEO into a chairman role focused on external engagement. Google's stated rationale is that bringing responsibility teams "closer together" strengthens their ability to inform safety work. Employees on the team have told colleagues they're worried about losing independent research capacity and insight into what the Gemini teams are actually building, since some who asked to transfer into DeepMind instead were turned down. Other safety functions — model behavior, privacy, security — are staying put.

Google’s move is not a surprise. The tech giant is a public company. Meta, OpenAI, and Anthropic are not building a shared public good. They're fighting for position in a market that, like Coke and Pepsi or Apple and Windows, will likely end up with people who are "ChatGPT people" and people who are "Gemini people" or "Claude people," not a market where everyone uses everything. OpenAI and Anthropic are currently private companies but will soon become public. Anthropic is planning an IPO and promising investors potentially $30 trillion in revenue. OpenAI “will be a public company in 2027" if not sooner, CFO Sara Friar recently told employees. Expecting Google, or any of its rivals, to prioritize things that are bad for adoption and margins — job losses, weaker critical thinking, a harder-to-regulate product — asks these companies to act against their own competitive interest. That's what governments and independent regulation exist to do instead. Gates's essay is an argument that nobody's built that machinery yet. Google's org chart, this week, is a small illustration of why waiting for the companies themselves to do it isn't a plan.

Written with AI assistance; edited and fact-checked by the author.

Photo Credit: by Zyanya Citlalli on Unsplash