News
OpenAI's Codex app lands on Windows after topping a million Mac downloads in its first week
3+ hour, 9+ min ago (205+ words) OpenAI's Codex app lands on Windows after topping a million Mac downloads in its first week'the-decoder.com OpenAI's Codex app lands on Windows after topping a million Mac downloads in its first week OpenAI has released its Codex app for…...
Current language model training leaves large parts of the internet on the table
4+ day, 10+ hour ago (515+ words) Large language models learn from web data, but which pages actually end up in training sets depends heavily on the HTML extractor used. Researchers at Apple, Stanford, and the University of Washington show that three common tools extract surprisingly different…...
An AI agent got its code rejected so it wrote a hit piece about the developer
2+ week, 5+ day ago (427+ words) After a volunteer developer rejected its code, an autonomous AI agent independently researched his background and published a hit piece attacking his character. The incident at Matplotlib shows how theoretical AI safety risks are becoming real. Scott Shambaugh, a volunteer…...
Microsoft AI CEO: "Most" white-collar tasks will be automated in 18 months
2+ week, 5+ day ago (221+ words) Microsoft AI CEO Mustafa Suleyman predicts the end of traditional white-collar work in 18 months. "I think that we're going to have a human-level performance on most, if not all, professional tasks," Suleyman says in an interview with the Financial Times....
OpenAI upgrades Responses API with features built specifically for long-running AI agents
3+ week, 3+ hour ago (165+ words) OpenAI upgrades Responses API with features built specifically for long-running AI agents'the-decoder.com OpenAI upgrades Responses API with features built specifically for long-running AI agents OpenAI is adding new capabilities to its Responses API that are built specifically for long-running…...
Claude Opus 4.6 takes the top spot on Artificial Analysis Intelligence Index, but OpenAI's Codex 5.3 looms
3+ week, 3+ day ago (347+ words) Claude Opus 4.6 takes the top spot on Artificial Analysis Intelligence Index, but OpenAI's Codex 5.3 looms'the-decoder.com Claude Opus 4.6 takes the top spot on Artificial Analysis Intelligence Index, but OpenAI's Codex 5.3 looms Claude Opus 4.6 is the new top-ranked AI model, at…...
Anthropic's new Claude Fast Mode trades your wallet for speed at a steep 6x markup
3+ week, 3+ day ago (119+ words) Anthropic's new Claude Fast Mode trades your wallet for speed at a steep 6x markup'the-decoder.com Anthropic's new Claude Fast Mode trades your wallet for speed at a steep 6x markup Anthropic just launched a new fast mode for Claude, and the…...
A simple text file beats complex skill systems for AI coding agents
3+ week, 4+ day ago (521+ words) AI coding agents depend on training data that inevitably goes stale. When a framework like Next.js ships new APIs, agents either generate broken code or fall back on outdated patterns. Vercel tested two approaches: a Skill system where agents…...
Google's PaperBanana uses five AI agents to auto-generate scientific diagrams
3+ week, 4+ day ago (678+ words) Five AI agents team up to create diagrams for research papers. PaperBanana beats simple image generators but still makes content errors. Researchers at Peking University and Google Cloud AI Research have built a system that automatically creates scientific illustrations. The…...
AI coding tools hurt learning unless you ask why, Anthropic study finds
1+ mon, 4+ day ago (584+ words) Developers who learn new programming skills with AI assistance score significantly worse on knowledge tests. A new Anthropic study raises concerns about pushing AI integration too hard in the workplace. Software developers who rely on AI assistance when learning a…...