Four things that happened in tech worth knowing. Two lines each, link to the source.
Grok 4.5 enters private beta at SpaceX and Tesla
Elon Musk announced that Grok 4.5, built on the 1.5-trillion-parameter V9 foundation model, now runs in private testing inside SpaceX and Tesla, ahead of any public release. xAI has not published independent benchmarks, so the performance claims remain unverified. The idea is clear though: train the model on real operational data, rocket trajectories and manufacturing lines, before handing it to anyone else.
The White House races to lock down rules for frontier models
The US administration is close to a voluntary agreement with OpenAI, Google, and Anthropic that would give the government access to the most powerful models up to 30 days before public launch, for a safety review. It is not a law, it is a voluntary protocol negotiated company by company. Still, it marks the first real attempt to put government oversight before a model ships, not after.
Nvidia now gets paid in a slice of AI cloud revenue too
Since July 1, Nvidia supplies Grace Blackwell GPUs to cloud operators like Sharon AI and Firmus not just for cash upfront, but also for a share of the future revenue that compute capacity generates. The first batch alone is 210,000 GPUs. It helps its cloud customers scale faster, but it also ties Nvidia's own fortunes to their balance sheets.
Microsoft sends 6,000 engineers inside client companies
Microsoft launched Frontier Company, a $2.5 billion unit with about 6,000 engineers and industry specialists who work directly alongside clients to make AI actually work in their processes. The move follows MIT research showing that 95% of enterprise generative AI pilots deliver no measurable impact on profit and loss. In short: companies have the AI tools but not the know-how to make them pay off, and cloud giants now sell the hands that run them too.
If one of these is useful for a project, get in touch.
- this week in tech
- ai
- hardware