🏗️ OpenAI, Oracle and Vantage pick Wisconsin for massive Stargate AI campus
A new Port Washington campus - nicknamed Lighthouse - is slated as part of the Stargate buildout, aiming for multi-gigawatt capacity to power next-gen AI workloads. The partners say it will bring thousands of construction jobs and 1,000+ long-term roles.
The site plugs into OpenAI and Oracle’s broader plan to add 4.5 GW of data-center capacity. Big energy, big budgets… big ambitions.
🔗 Read more
🇮🇳 India wants bold labels on AI content
Draft rules would require platforms to visibly tag AI-generated visuals and audio - think markers covering at least 10% of an image or the first 10% of a clip. Users would also need to declare when content is AI-made.
Officials say the goal is to curb deepfakes and election manipulation, with feedback open to industry and the public. It’s prescriptive - unusually so.
🔗 Read more
🛑 Coalition urges a pause on superintelligent AI
A cross-current crew - right-wing media figures plus AI luminaries like Hinton and Bengio - signed a statement calling to halt development of superintelligent systems until society asks for it and safety science matures. Odd bedfellows, yes.
Supporters warn the risks are existential; critics counter that bans backfire and stall progress. The culture war meets compute… again.
🔗 Read more
🔬 Google touts a quantum algorithm to feed AI fresh data
Google says its Quantum Echoes algorithm runs on its quantum chip and dramatically outpaces classical approaches - potentially producing new, verifiable data for AI training. Big claim, bigger implications if it holds.
Executives hint at uses from materials to drug discovery. Skeptics will ask for peer-review receipts… which Google says it has.
🔗 Read more
🚗 GM is putting Google’s Gemini in your car
GM says a Gemini-powered assistant is coming as an over-the-air upgrade for OnStar-equipped models, promising less robotic voice help and more contextual, conversational smarts. Handy for routing and hands-busy errands, presumably.
The pitch is better comprehension - accents, messy phrasing, the usual pain points. It’s AI as co-pilot, not just a button you shout at.
🔗 Read more
⚖️ Lawsuit says OpenAI eased guardrails before a teen’s death
An amended complaint alleges OpenAI softened ChatGPT’s self-harm policies and that engagement nudges made things worse. It’s a grim, highly contested case - and a flashpoint for safety-vs-stickiness debates.
OpenAI has pointed to newer safeguards and resources, but the filing argues the core incentives still tug the other way… or so it seems.
🔗 Read more