Sam Altman wants to slow AI, but still promises rapid progress

Sam Altman is now saying out loud that AI development should slow down a bit. That's worth noting, because he heads one of the companies pushing hardest for progress.

Concretely, OpenAI has introduced safety checks before it launches its biggest training runs. In other words, an extra look under the hood before turning up the power.

According to the outlet The Information, OpenAI has also for several months been talking with competitors Anthropic and Google about a form of joint self-regulation. The idea is that the big players find common ground before someone from outside sets the rules for them.

For you as a user or company, this doesn't change anything here and now. No brakes are being applied to the tools you'll use tomorrow.

What's worth watching is whether the talks between the big AI companies lead to concrete agreements. If they do, it could in the long run mean slower, but hopefully more carefully considered, updates to the AI tools many Danes and companies already use.

Until then: no reason to change anything at all in your own use of AI. Just make a note of Altman's name the next time he says something about speed versus caution: it tends to matter for the rest of the industry.