YESTERDAY, STILL WORTH KNOWING

Still
Crunchy.

A day of sharp contrasts saw political rivals unite against AI risks while tech giants raced to build faster chips and more accessible tools.

Political Rivals Unite on AI Safety

In a rare display of bipartisanship, Bernie Sanders and Steve Bannon joined forces in Washington to demand tighter government limits on artificial intelligence. The unusual coalition, which also includes religious leaders and artists, called for construction freezes on data centres and presidential executive orders, arguing that the technology poses a dangerous and unprecedented risk to humanity. This rare alignment highlights a growing consensus that current regulation is insufficient to address the scale of the threat.

Meanwhile, OpenAI has backed the FRONTIER Act, a federal safety bill. This move suggests that even the leading developers of the technology are seeking external oversight. The backing of the FRONTIER Act could set a precedent for mandatory third-party safety audits for major AI developers, potentially forcing companies to prove their systems are safe before deployment.

Hardware Race and Memory Shifts

The race for superior AI hardware showed no signs of slowing. NVIDIA unveiled its Vera Rubin NVL72 system, which debuted in the MLPerf Inference v6.1 benchmarks with up to 3.7x higher throughput than its predecessor. These gains are attributed to a full-stack codesign that includes enhanced Tensor Cores and Transformer Engine acceleration. The high scaling efficiency suggests that Vera Rubin could significantly lower the cost per token for AI inference workloads.

Separately, PieceMakers, a Nanya-backed DRAM designer, has listed on Taiwan’s Emerging Stock Board with a strategy to supply custom-designed memory that fuses DRAM stacks directly to processors using hybrid bonding. The company argues that edge AI inference memory diverges from the high-bandwidth memory used in data centres, positioning its products between Nvidia’s SRAM-only Groq LPU and traditional HBM solutions.

AI in Daily Life and Gaming

The integration of AI into everyday life continued to expand. Google is rolling out early access to a new integration for Google Home that lets third-party AI agents control connected devices, review camera summaries and access smart home activity using natural language. This move follows a broader trend of opening up smart home platforms to external tools, potentially making home automation more conversational and flexible.

In the gaming world, Capcom scrapped a contentious microtransaction in Monster Hunter Wilds that previously charged players to edit their hunter and Palico characters. The change means players will soon be able to alter their appearance as often as they wish without incurring extra cost. This decision improves player freedom and addresses a common criticism of the game's monetisation strategy.

BRIEF INDEX

15 briefs behind this edition