/Dispatch #106

Comms for Week 5, Jul 2026

[WHAT EVERYONE'S TALKING ABOUT]

Dario Amodei clarifies Anthropic's stance on open-weight models amid industry pressure

Amodei says Anthropic never sought to ban open-weight models and calls safe ones a public good — yet he alone refused the industry's open-weights letter, citing bioweapon and authoritarian-regime risks.

The bottom line: AI's most safety-cautious lab now stands alone against the open-weight tide.

Dario Amodei clarifies Anthropic's stance on open-weight models amid industry pressure

Jensen Huang takes open AI models fight to Capitol Hill after first-ever social media post

Mark Zuckerberg says US should not block Chinese AI models, warns against regulatory capture

Anthropic launches Claude Opus 5 AI model, matching Fable 5 power at half the price

DON'T BE THE LAST TO KNOW

[Altman Rewrites the Workweek Promise]

Sam Altman says AI won't shorten the workweek because humans secretly enjoy being busy

For years, the promise was that technology would buy us leisure. Now Sam Altman says forget it: humans secretly enjoy being busy, so a 'post-superintelligence world' leaves us more overworked, not less. It's a sharp reversal — real four-day-week trials have lifted revenue 8% and cut stress 35% — and Altman's own line on AI and jobs has swung between optimism and gloom every few months.

[CONTINUE READING ABOUT]

TECHNOLOGY BUSINESS STARTUPS POLICY

[WHAT ELSE IS TRENDING]

Google DeepMind dismantles Nobel Prize-winning AlphaFold team in strategic shift to Gemini

Google DeepMind is folding its Nobel-winning AlphaFold team into Gemini. Nearly a quarter have left — and co-winner John Jumper is defecting to Anthropic with two colleagues.

Tl;dr: The team that won a Nobel is being broken up.

Google DeepMind dismantles Nobel Prize-winning AlphaFold team in strategic shift to Gemini

AI-Generated Doctors Spread Fake Cures on TikTok, Experts Warn of Public Safety Risks

AI-generated 'doctors' appear in 40% of TikTok's top health videos — up to 84% for 'health tips' — pushing invented supplements and cancer myths, sometimes cloning real physicians' faces.

Tl;dr: AI 'doctors' are pushing fake cures to millions.

AI-Generated Doctors Spread Fake Cures on TikTok, Experts Warn of Public Safety Risks

Claude AI shared chats exposed in Google search results, raising AI chatbot privacy concerns

Roughly 500+ shared Claude conversations — including medical records and NSFW role-play — got indexed by Google and Bing. Anthropic blames users; Google blames Anthropic. It's happened before.

Tl;dr: Hundreds of private Claude chats turned up in Google.

Claude AI shared chats exposed in Google search results, raising AI chatbot privacy concerns

+23

Technology

Black Forest Labs unveils FLUX 3 multimodal AI to generate video, images, and robot actions

Germany's Black Forest Labs unveiled FLUX 3, generating 20-second video with audio from a prompt — and driving robot hands at Audi. It beat Runway and Luma in early tests.

VentureBeat, Decrypt

Moonshot AI releases Kimi K3 open-weight model with 2.8 trillion parameters for public download

Moonshot released the weights for Kimi K3 — a 2.8-trillion-parameter open model — free to download. It rivals Claude and GPT on coding at a fraction of the cost.

VentureBeat, Digital Trends and 4 more

Business

Nvidia's $250 billion OpenAI backstop ignites debate over circular AI financing risks

Nvidia may guarantee up to $250B of OpenAI's borrowing on its own strong credit — so OpenAI can buy more Nvidia chips. Investors balked, wiping ~$250B off Nvidia in a day.

CNBC, NYT and 10 more

SK hynix operating profit surges 557% on AI boom, but shares tumble as expectations outpace reality

SK hynix, maker of the memory chips AI needs, posted a record profit up 557% — yet its stock fell 10%, because investors had expected even more.

Tom's Hardware, The Register and 11 more

Startups

Nvidia bets billions on Ilya Sutskever's Safe Superintelligence to scale AI research

Nvidia is investing billions in Ilya Sutskever's Safe Superintelligence — no product, no revenue, a $32B valuation — after rare access to its research. The compute jumps tenfold.

TechCrunch, FT and 3 more

Andrew Ng launches LearnVector with $100 million to train workers as AI reshapes jobs

Andrew Ng launched LearnVector with $100M from Coursera to reskill white-collar workers using AI tutors. His bet: AI won't cause a job apocalypse — it'll shift which skills matter.

Reuters, Axios

Policy

AI Kill Switch Act gives DHS power to shut down rogue AI systems after OpenAI security breach

A bipartisan bill would let Homeland Security shut down catastrophic or rogue AI, with fines of $20M a day — Washington's answer to the OpenAI breach.

Ars Technica, The Verge and 22 more

EU AI Act labeling rules take effect Sunday, requiring companies to mark all deepfakes

New EU rules take effect this weekend requiring companies to label AI-made content — deepfakes, chatbots, synthetic media. Meta, Google, and TikTok are already tagging, but watermarks are easy to strip.

TechSpot, PetaPixel and 4 more

Don’t drown in AI news. We cut through the noise - filtering, ranking and summarizing the most important AI news, breakthroughs and research daily. Follow topics that matter to you and stay ahead.

Explore

News Categories

Technology Business Policy Startups Health Science Entertainment

Terms Privacy Content Contact Us