Micron, Anthropic sign AI infrastructure supply agreement↗
Micron struck a strategic deal with Anthropic covering memory, storage products, and a fresh investment in Anthropic’s latest funding round.
The interesting bit is the stack-level intimacy here: Anthropic wants memory and storage tuned around Claude workloads, not just more generic boxes in a warehouse. AI infrastructure is getting peculiarly bespoke now, like tailoring a suit for a GPU furnace.
Patch the Planet: a Daybreak initiative to support open source maintainers↗
OpenAI launched Patch the Planet, built with Trail of Bits, to help open-source maintainers find, validate, and fix security vulnerabilities using AI-assisted research plus human review.
Initial projects include cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, Python, and others. It’s less glamorous than a model launch - but probably more consequential than half the shiny demos flying around.
AI startup Reflection signs computing power deal with SpaceX↗
Reflection AI signed a compute deal with SpaceX, giving the startup access to Nvidia GB300 chips inside the Colossus 2 data center.
The bill is huge, because of course it is. Open-source AI labs want to compete with closed giants, but the toll road is compute - and this one looks more like a drawbridge with rockets painted on it.
AI boom's US employment, wage impact muted so far, ECB study finds↗
An ECB study found AI has not yet caused a broad shock to U.S. employment or wage growth, even though some exposed jobs are clearly feeling the squeeze.
The catch, unsurprisingly, is distribution. Junior workers and roles with high substitution risk look more vulnerable, so the “AI took all the jobs” story is too blunt - but the “nothing to see here” story is also a little too cosy.
US curbs on AI spur European firms to spread the risk↗
European companies are moving faster to avoid dependence on any single AI provider after U.S. restrictions exposed how remotely controlled proprietary models can be switched off or limited.
Siemens, Renault, Orange, and ChapsVision are already mixing U.S., Chinese, and European models. Sovereignty here doesn’t mean building a castle, exactly - more like keeping three spare keys under different mats.
China's push for green power use in AI projects faces hurdles, experts say↗
China wants more renewable power flowing into its fast-growing AI data-center sector, but experts say peak demand is hard to forecast and grid operators are wary.
The awkward bit: GPUs are expensive, so operators want them running flat-out, which makes flexible energy use tricky. Green AI sounds lovely on a poster, but the grid is where the poetry gets mugged.
FAQ
Why does the Micron and Anthropic AI infrastructure deal matter?
The Micron-Anthropic deal matters because it signals a move toward more customized AI infrastructure. Rather than relying on generic memory and storage, Anthropic wants hardware shaped around Claude workloads. That suggests major AI labs may increasingly influence the underlying stack, from chips and storage to data-center design, around specific model behavior and performance needs.
What is OpenAI’s Patch the Planet initiative?
Patch the Planet is an OpenAI initiative built with Trail of Bits to help open-source maintainers find, validate, and fix security vulnerabilities. It combines AI-assisted research with human review. The first projects mentioned include cURL, NATS Server, pyca/cryptography, Sigstore, aiohttp, Python, and others that many software systems depend on.
Why are AI startups signing huge compute deals?
AI startups need vast amounts of computing power to train and run competitive models. Reflection AI’s deal with SpaceX gives it access to Nvidia GB300 chips in the Colossus 2 data center. For open-source AI labs, compute can be the main bottleneck, since ambitious model development often depends on costly, high-performance infrastructure.
Is AI already causing major job losses in the U.S.?
According to the ECB study described here, AI has not yet caused a broad shock to U.S. employment or wage growth. Still, the impact is not evenly distributed. Junior workers and jobs with high substitution risk may face more pressure, so the wider labor-market picture is more nuanced than either panic or dismissal suggests.
Why are European companies spreading AI provider risk?
European companies are reducing dependence on any single AI provider because restrictions can limit access to proprietary models. Firms such as Siemens, Renault, Orange, and ChapsVision are reportedly combining U.S., Chinese, and European models. The goal is not total isolation, but greater resilience if one provider, country, or platform becomes unavailable or constrained.