The Trump administration is planning to provide Cold War-era plutonium from dismantled nuclear warheads to nuclear startups that want to convert it into reactor fuel, arguing it could help address a looming fuel shortage for advanced reactors. Critics warn the idea raises serious nonproliferation, security, cost, and technical concerns. The New York Times reports: The plan has generated debate and some unease among nonproliferation experts. If finalized, it would mark the first time the U.S. government has made weapons-grade plutonium available to private companies. The Energy Department has more than 50 tons of surplus plutonium left over from nuclear weapons programs, and the agency had previously been planning to dilute much of that material and bury it. Some of the nuclear start-ups trying to obtain that plutonium say that transforming the waste into fuel is a better way to dispose of it. On Tuesday, the Energy Department said that it had selected five companies to enter into "advanced negotiations" to potentially receive some surplus plutonium. That includes Oklo, a California-based nuclear power company, which plans to partner with Newcleo, a European developer of advanced nuclear reactors. Using plutonium for fuel, Oklo and Newcleo said, could solve a looming problem: Energy firms want to build a new wave of nuclear reactors, but the United States can't yet make enough conventional fuel from uranium to supply the plants. Harvesting old plutonium stockpiles could provide a short-term fix. "A lack of fuel is one of the biggest choke points in expanding nuclear power right now," said Jacob DeWitte, the chief executive of Oklo, which is developing a novel type of small reactor intended to run on plutonium. "This will help us get more nuclear power online faster." [...] The plan is not yet final, and companies will still have to negotiate with the federal government over how to secure and transfer the plutonium. In addition to Oklo, the Energy Department said it had also selected four other companies -- Standard Nuclear, Exodys Energy, SHINE Technologies and Flibe Energy -- to enter into advanced negotiations to receive the material under its Surplus Plutonium Utilization Program, which was established last year. The program "is anticipated to help companies unlock the next level of private funding to broaden domestic nuclear fuel supplies, spur innovation on American recycling technologies, and unlock private sector funding to fuel the nation's nuclear renaissance," said Michael Goff, the principal deputy assistant secretary of nuclear energy, in a statement.

Read more of this story at Slashdot.

Apple is reportedly working to shrink Google's Gemini models enough to power parts of a long-delayed AI-enhanced Siri on iPhones. But despite Apple's best efforts to run the AI locally, "the iPhone's Gemini makeover will lean heavily on Google and Nvidia in the cloud," reports Ars Technica. That could complicate Apple's privacy-first AI messaging, especially if more complex Siri requests are routed through Google infrastructure and Nvidia's encrypted cloud-computing platform. Ars Technica reports: After inking the Google deal, Apple apparently got to work distilling Google's giant cloud-based Gemini models. Distillation is a process in which a small, less resource-intensive model learns to mimic a large, expensive one. With enough time, this can reliably transfer useful capabilities while pruning less important weights from the model. That may enable Siri to handle some tasks with private local compute, but a cloud component looks inevitable. Processing users' AI data in the cloud could be a problem for Apple. At WWDC, the company will probably promote its years of experience designing chips and how well that positions it for AI. However, The Information claims that Apple has struggled to even get Google's massive undistilled Gemini models running on its custom Private Cloud Compute infrastructure, which is built on on M-series Mac chips. When the smarter Siri rolls out, it will probably route more complex tasks to Google's cloud infrastructure instead of Apple's, but it won't be running on Google TPUs. Apple has reportedly signed a deal with Nvidia to use its Confidential Computing platform for this purpose. Confidential Computing keeps data encrypted on Nvidia GPUs while it's being processed in the cloud, which could help Apple claim it's still sensitive to user privacy concerns. It might even retain its own Private Cloud Compute branding for the system. The iPhone probably won't tell you which version of Gemini is handling individual Siri requests. Device makers designing hybrid systems that rely on local and cloud-based AI like to talk about making the experience feel "seamless." There might be clues, though.

Read more of this story at Slashdot.