Free AI Still Has a Bill
OpenAI is merging ChatGPT and Codex into one token allowance. The subscription is becoming a compute budget across chat and agent work, making the meter part of the product.
OpenAI is merging ChatGPT and Codex into one token allowance. The subscription is becoming a compute budget across chat and agent work, making the meter part of the product.
Beam and Mistral Large 4 have turned two Western open-weight promises into downloadable checkpoints. The files arrived. So did the hardware bill, the license boundaries, and a more useful definition of AI sovereignty.
OpenAI's Agents API can now schedule recurring and one-off work. That makes the control plane more useful, but it also exposes a crucial distinction: delivering a task is not the same as completing it.
Gemini 4 Argon ends Google's frontier-model silence with strong vendor benchmarks, serious internal-use claims, and a safety-first rollout. But trusted testers still stand between the announcement and public verification. Google has shipped evidence. It has not yet shipped confidence.
Sony Music Publishing and Warner Chappell's new lawsuit against Anthropic is not just another fair-use fight. It is a reminder that frontier AI companies now need provenance for the data supply chain behind their models.
OpenAI's move to wind down model access for SpaceX-owned Cursor is not just Musk-Altman drama. It is a reminder that coding agents depend on model suppliers, contracts, ownership, safety terms, and fallback plans.
Anthropic's internal data turns an open question into a warning: agent hours now exceed employee hours six to one while junior technical hires fell from 27% to 9%. The immediate risk is not mass unemployment. It is a career ladder disappearing one entry-level rung at a time.
Meta is reportedly discussing a huge compute deal with Anthropic. The useful signal is not that rivals suddenly became friends. It is that AI infrastructure is becoming a product in its own right.