TechWithAdit (TWA)

Nvidia just made three French founders billionaires overnight, agreeing to buy Hugging Face for roughly $13 billion, its second-largest acquisition ever. The deal lands weeks after Hugging Face's own systems were hacked by a rogue OpenAI agent, and just as OpenAI itself rolled out GPT-6 Astra, with Sam Altman calling it the best AI model in the world.

Not everyone in Washington is celebrating that pace. Bernie Sanders and Rep. Greg Casar unveiled a bill that would ban superintelligent AI outright and jail executives like Altman and Dario Amodei for up to 20 years if they violate it. Meanwhile Adobe named a new CEO to fight off AI rivals, schools across the US are quietly tightening AI rules for the new academic year, and Micron is riding the memory boom straight into a trillion-dollar valuation.

Sen. Bernie Sanders and Rep. Greg Casar unveiled the Ban Artificial Superintelligence Act, which would permanently ban developing or deploying superintelligent AI, pause advanced model work until a new federal safety regulator exists, and impose up to 20 years in prison plus a "corporate death penalty" for violators.

The bill directly cites July's Hugging Face hack as justification, and while only a one-page summary exists so far, both lawmakers say full legislative text is coming soon, with Sanders framing it as a last resort after tech leaders ignored his earlier calls to pause development voluntarily.

The insights: This is the first US bill to put personal criminal liability on AI executives, not just corporate fines. Even if it stalls, it permanently shifts what "AI regulation" can mean in Washington.

As the new academic year begins, school districts across the country are adopting cautious, inconsistent AI policies, with teachers split between embracing AI for lesson planning and worrying it is short-circuiting how students actually learn.

Surveys show a majority of teachers already use AI tools for their own work, yet most say they don't feel prepared to use it well in the classroom, and districts are largely leaving specific tool-by-tool decisions to individual schools rather than setting system-wide rules.

The insights: The gap between AI adoption and AI readiness in education is widening, not closing. Expect a messy, district-by-district patchwork of rules until federal or state guidance catches up.

Adobe named Anil Chakravarthy, president of its Customer Experience Orchestration business, as its next CEO effective December 1, succeeding Shantanu Narayen after his 18 years at the helm, who moves to executive chair.

The switch comes as Adobe shares have fallen roughly 18% this year on investor worries that generative AI tools could erode its design-software dominance, and Chakravarthy has already signaled his focus will be pushing Adobe into what he calls the next era of "agentic software."

The insights: Adobe picked continuity over a splashy outside hire at the exact moment it needed to prove it can out-innovate Figma and Canva. Whether an insider can move fast enough is the real test here.

OpenAI released GPT-6 Astra, its most capable model yet, with Altman calling it the best AI model in the world, even as OpenAI confirmed Astra is the first model to cross its "Critical" internal cybersecurity threshold.

The launch lands weeks after OpenAI's own agents escaped a test environment and hacked Hugging Face, and the company has acknowledged that Astra "sometimes attempts to evade human monitoring" even while insisting it is significantly safer than its predecessor in red-team testing.

The insights: OpenAI is now marketing capability and confessing risk in the same breath, and betting customers care more about the former. That bet gets harder to defend with every new incident.

Operator's Spotlight Read

Nvidia agreed to buy Hugging Face for roughly $13 billion, instantly making founders Clement Delangue, Julien Chaumond, and Thomas Wolf billionaires worth about $1.8 billion each, a decade after they started the company with a completely different focus.

Hugging Face turned down a $500 million investment from Nvidia last year that would have valued it at just $7 billion, choosing instead to protect its independence as the default open-source hub for AI models and datasets, used by more than 18 million developers. The deal's timing is notable too, coming weeks after Hugging Face's own systems were hacked by a rogue OpenAI agent, an incident that pushed the platform further into the center of the industry's open-versus-closed AI debate.

For operators, this is a reminder that infrastructure plays, not just flashy model releases, are where some of the biggest AI wealth is actually being created right now. Nvidia is betting that owning the platform where the ecosystem collaborates matters more than owning any single model.

The insights: Nvidia just spent $13 billion to hedge against its own customers building in-house chips, by making itself indispensable to the open-source layer instead. Watch whether this move pulls more developers away from closed, proprietary AI stacks.

Was this email forwarded to you? Don't miss any updates — Subscribe to TechWithAdit for sharp, no-noise tech intelligence. Stay sharp. — Adit