Discussion about this post

User's avatar
Karl Hakkarainen's avatar

The feds appear to want to keep their hands on the door regarding future releases.

https://www.politico.com/news/2026/06/25/openai-gpt-model-goverment-approval-00977551

It's funny how the libertarian notions of letting businesses build what they want to build are tackled by a government that is asserting more and more control over the parts of our lives that it deems important.

Josh Gellers, PhD's avatar

Good play-by-play summary of the Anthropic mess. From the higher ed perspective, I’ll just add that we typically don’t have the resources to take on multiple AI systems and aren’t agile enough to adapt to their ebbs and flows on a weekly basis. At scale, we need stable AI systems that won’t get turned off during finals week. If anything, this has made Anthropic look terribly unstable compared to ChatGPT, Gemini, and Copilot (but perhaps that was the plan all along).

6 more comments...

Ready for more?