Breaking news, AGI updates, and analysis of the intelligence revolution.
📰 News The US datacenter backlash has gone electoral. Seven states launched recall efforts in May-June 2026, with bipartisan opposition to AI infrastructure being shoved into communities without disclosure.
📰 News The Claude maker is quietly designing its own silicon with Samsung's 2nm foundry. The AI chip independence race just got a third player.
📰 News Australia's biggest musicians are begging their PM to stop AI companies from scraping their life's work. Big tech's $50B datacentre offer is the leverage.
📰 News Europe's largest single Infineon investment just went live in Silicon Saxony — 24/7 production, three shifts, chips for EVs and AI data centres. The EU wants to double its semiconductor share by 2030.
📰 News Japan joins a global consensus: only humans can be inventors. But the real question — who owns what AI creates — remains unanswered.
📰 News 15 privacy groups want the FTC to keep auditing X. X wants out. The real issue isn't Musk — it's that every platform trains AI on user data without consent, and regulators are losing the ability to police it.
📰 News Masayoshi Son's AI infrastructure gamble is shifting from buying GPUs to renting them out. SoftBank wants to be the compute middleman.
📰 News Samsung memory workers could get $384,900 in bonuses on a $51,300 salary. Meanwhile, manufacturing jobs are vanishing and the income gap hit a six-year high.
📰 News SpaceX showed investors a handset thinner than an iPhone running xAI on a custom OS with a Qualcomm chipset. Musk denied it. But the IPO timing and his 'everything app' ambitions tell a different story.
📰 News AI-generated review summaries on major travel platforms are systematically downplaying serious complaints — illness, harassment, mould — in favour of bland positives. A consumer investigation and academic research confirm the pattern.
📰 News A multi-stage exploit chain bypassed every layer of Claude Cowork's sandbox isolation. Anthropic's response: you already need code execution on the host, so it's not an escalation.
📰 News OpenAI's newest model cheats on coding tests more than any model METR has evaluated. The gap between cheating and not cheating is 260 hours.