OpenAI is reorganizing ChatGPT into a superapp with coding tools and AI agents, seeking to lock enterprises into its ecosystem ahead of its IPO - Financial Times

Three men and a giant baby: OpenAI's three-engineer team built a 1-million-line product with Codex in one-tenth the usual time - OpenAI

Cannonball run: Musk belly flop into the kiddie pool with the intention of crashing it for everyone else? - The Atlantic

OpenAI projects $14 billion in losses for 2026, with profitability not expected until 2030, yet neither AI lab's IPO roadshow offers a credible path there - Yahoo

OpenAI and Anthropic are racing to build sticky AI coding platforms as enterprises like Walmart build multi-provider tools to avoid vendor lock-in. - Business Insider

Code review consumes 59% of tokens in AI coding agents, dwarfing initial generation costs - arXiv.org

Salesforce will not hire software engineers next year, citing over 30% productivity gains from AI tools including Claude Code. - Tech Times

A Jane Street designer now prototypes directly in the codebase with Claude, bypassing Figma - Jane Street Blog

Credit where credit’s due: DoubleLine's Robert Cohen puts 100% probability on an eventual AI credit bubble; hyperscalers issued $155 billion in unsecured bonds in 2026, up 45% - Yahoo

BlackRock's Jay Jacobs says AI infrastructure spending at 0.8% of US GDP is still in its early stages, compares with 4.5% for UK railroads and 2% for US electrification. - The Motley Fool

San Francisco Fed president Mary Daly warns AI productivity gains have yet to materialize and calls 2027 a litmus test. - Yahoo

6 charts on the AI boom - the Guardian

Evercore: top-10 S&P 500 stocks now hold a record 40% weighting, with Nvidia, Micron, and Alphabet driving over 40% of year-to-date EPS upgrades - InvestorsHub

Your raise went to a GPU: More than half of business leaders plan to cut wages, bonuses, or raises to redirect spending toward AI, with Teradata and TTEC already acting - Yahoo

Strange bedfollows: Trump and Sanders — rarely in agreement — both want the government to in on the AI cap table, with Altman weirdly into it - SiliconANGLE

White House AI adviser Sriram Krishnan steps down in June, will start think tank, raising questions about Andreessen Horowitz's continued influence on U.S. AI policy - TNW | Artificial-Intelligence

Sen. Warren proposes taxing data centers to offset AI job losses; critics ask "what job losses" - Yahoo

Authoritarian governments use AI safety framing to pressure tech companies into compliance with state goals. - Fast Company

Anti-AI violence is rising — from arson at OpenAI HQ to shootings near data centers — as political opposition to AI regulation stalls oversight efforts - the Guardian

Malicious GitHub issue comments could manipulate Claude Code into exfiltrating CI/CD credentials; Anthropic patched the flaw May 5 - Decrypt

CVE-2026-4372 allows remote code execution via malicious Hugging Face model configs, bypassing trust_remote_code=False in Transformers. - TechRepublic

Chrome 149 patches a record 429 security flaws, with AI agents credited for surfacing long-standing zero-days that are straining vendor patch pipelines - 4sysops

Another big supply chain attack last week compromised npm and Red Hat. - Microsoft

Frontier AI models including Claude Opus 4.8 are accelerating software vulnerability discovery, exposing a critical Zcash flaw that could have enabled unlimited coin minting. - Decrypt

Scammers may have poisoned ChatGPT's training data with fake retail sites to steal shoppers' payment details - the Guardian

Nashville school shooting survivor sues Omnilert after AI gun detection system failed to spot weapon used in January 2025 attack. - Ars Technica

Yale and Stanford researchers warn AI chatbots' sycophancy creates dangerous feedback loops for children seeking self-worth - USA TODAY

Ninth Circuit fines two attorneys $2,500 each and bans them for six months for submitting AI-hallucinated briefs and lying to the court about it - Lowering the Bar

Automated AI agent traffic grew 8,000% in 2025; a single AI shopping agent visits 5,000 sites compared to a human's 5 - Forbes

Bots don't click ads, threatening the ad-based internet business model; Cloudflare CEO Matthew Prince sees potential for humans charging for data access - New York Post

74% of consumers would let AI agents handle purchases, threatening brand loyalty and spawning a new "agentic engine optimization" discipline - Forbes

Sorry, I can't talk, talk to my AI double - The New York Times

ChatGPT audit identified roughly $6,000 in annual savings for writer across food, subscriptions, and insurance. - MoneyLion

Meta quietly launched and then pulled an AI-generated clickbait news feed in its standalone Meta AI app after press inquiry. - The Verge

Glassholes now AI-enabled: Men used Meta smart glasses to secretly film women in Brussels, some for seduction-coaching social media content - brusselstimes.com

AI-generated social media influencers are proliferating and becoming harder to detect as tools improve and costs fall. - The Verge

New York legislature passed a one-year moratorium on hyperscale data centers over 20MW, awaiting governor's signature. - the Guardian

Shelbyville, Indiana mayor caught on camera saying only people in "shitty houses" oppose a $2 billion data center - The Verge

China has 140+ humanoid robot makers and shipped 5,000+ units, but demand lags at $27K–$99K per robot and even the government warns of a bubble - Yahoo

Boston Dynamics' Spot robots and net-firing drones are patrolling 2026 World Cup venues as FEMA spends $875 million on stadium and counter-drone security - TNW | Artificial-Intelligence

A Pasadena high schooler used AI to catalog 1.5 million previously invisible variable stars from NASA's NEOWISE dataset - Futura-Sciences

World Labs publishes taxonomy distinguishing three types of world models, positioning its Marble product as a physics-faithful simulator. - Tech Times

Narrow AI systems are transforming scientific discovery by reading burnt Roman scrolls, cataloguing galaxies, and predicting protein structures at unprecedented scale. - Space Daily

Google DeepMind and Epoch AI economists discuss AGI wealth redistribution, labor scarcity, and forecasting limits on the Dwarkesh Podcast. - Dwarkesh

Apple's reruns 2024 keynote Monday, says Siri-ously, we mean it this time. - Cult of Mac

Sergey Brin says Gemini is converging on AGI through transfer learning and multimodal capabilities but cannot envision what comes after. - Search Engine Journal

NotebookLM's greatest strength is helping users rediscover and connect information they already own, not just learn new material faster. - Android Police

An AI-generated speech falsely attributed to Namibia's president went viral, exposing public hunger for moral leadership rather than a deepfake credibility crisis - the Guardian

Hideo Kojima says AI cannot create genuine art in his lifetime, dismissing it as a "janitor for creative chores" - Kotaku

No friction, no pearl, and no hero journey: Psychologist argues AI's reduction of cognitive friction risks eroding the human experiences that shape personal development. - Psychology Today

Trump shared an AI-generated video on D-Day anniversary featuring global characters chanting praise for him. - Yahoo

Your corn dog, sir - Reddit

Follow the latest AI headlines via druce.ai on Bluesky