Am I the Only One Who Watched Terminator?
Feb 10, 2025
AI is getting the nukes now. Cool, cool, cool. Google quietly deleted a key section from its AI Principles, the pledge against developing AI for weapons or surveillance tech.
TL;DR Takeaways
- Google quietly dropped its explicit ban on developing AI for weapons and surveillance
- Critics warn this could make AI-enabled warfare feel normal
- OpenAI is also partnering with U.S. National Labs for “nuclear research.”
The AI Arms Race Is Real
There was a time when Google made loud promises not to build AI for warfare or surveillance. Way back in 2018, Google employees revolted over Project Maven, a military project using AI to analyze drone footage. Thousands protested, some quit, and Google promised to steer clear of weapons projects. That line in the sand is now gone.
Google just updated its AI Principles page. Out went the explicit ban; in came a blog post from DeepMind CEO Demis Hassabis and SVP James Manyika about “responsible” AI partnerships with democracies. That is a pretty big rhetorical shift from “we will not build weapons” to “we will work with governments we trust.”
The Pentagon now has an office dedicated to AI systems for warfare. Tech companies are pouring billions into AI infrastructure and eyeing military contracts to monetize it. Alphabet alone plans $75 billion in AI investments.
Meanwhile OpenAI, once the poster child for AI safety, will now provide GPT models to U.S. National Labs for nuclear weapons research, “to reduce nuclear risk,” they claim. Critics aren’t convinced.
The Tech World Is Split
- Andrew Ng supports the move, citing national security.
- Meredith Whittaker and Geoffrey Hinton warn against weaponizing AI.
- Jeff Dean previously opposed autonomous weapons.
- Margaret Mitchell isn’t buying it, warning it opens the door to AI “that can kill people.”
And inside Google not everyone is on board. Apparently memes on internal boards joked “Are we the baddies?” I’m curious if we will see any protests or mass resignations soon, but in today’s not-so-hot job market I wouldn’t be surprised if it’s much more subdued than last time.
Shall We Play a Game?
This isn’t just about tech companies backtracking on their ethics. It’s about normalizing machine-speed warfare, where systems react to systems before a human can really understand what is happening. That is where mistakes get terrifying fast, and where accountability gets very fuzzy. I want to believe, I need to believe, that technology can be an overall force for good. Self-driving cars will save lives. AI will diagnose patients and cure diseases. I want that Star Trek future. But that hinges on us not blowing ourselves up…