Claude Fable 5: Anthropic's First Public Mythos-Class Model Ships With a Built-In Safety Valve
Fable 5 scores 80.3% on SWE-Bench Pro and brings Mythos-level coding to Pro subscribers—but silently downgrades to Opus for sensitive domains. Here's what that means for agentic workflows.
If Claude Fable 5 Stops Helping You With ML Work, You'll Never Know Why
Anthropic's Fable 5 silently degrades responses for pretraining pipelines, distributed training infrastructure, and ML accelerator design—affecting 0.03% of traffic, with no notification to users.
Why Large Models Learn Rare Skills That Small Ones Forget
New research pinpoints the mechanism behind emergent LLM abilities: an update-and-forget pattern where larger models preserve rare-task signals that smaller ones lose when frequent tasks dominate training steps.
Agentic AI Solved the Coding Problem—and Exposed Everything Else
Agentic AI has made coding fast. The exposed problems—requirements, integration, accountability—were always the real bottleneck. Now they're visible.
Apple Passwords in iOS 27 Is the Clearest Agentic AI Demo Yet
iOS 27's Passwords app navigates websites and changes weak credentials autonomously—one tap, background execution. The clearest consumer agentic AI feature yet.
Agentic AI Token Pricing: When Tokens Become a Business Metric
Token-based AI pricing is fragmenting into performance tiers — and companies that treat token consumption as a proxy for productivity are already burning budgets without results.
AI Spending Visibility: Only 1 in 4 Companies Knows What It Pays
Only 26% of companies have full visibility into AI costs. As agentic workloads scale, token-based billing is creating a CFO blindspot that finance teams were never built to manage.
AI Enthusiasts Race Against Time. AI Skeptics Race Against Entropy. Both Are Right.
The real AI adoption divide isn't about whether AI works — it's about competing timescales. Enthusiasts fear organizational obsolescence. Skeptics fear system incoherence. Neither is wrong.
Bain Study: 40% of Companies Miss AI Savings Targets — Humans Are the Bottleneck
A Bain study finds companies expected 11–20% AI cost savings but 40% achieved less than 10%. The gap isn't the AI — it's that only 7% actually deploy fully autonomous agents as their business cases assumed.
Google Offers Publishers an AI Search Opt-Out. The Catch: There's Nowhere Else to Go.
Google Search Console now lets publishers exclude content from AI Overviews and AI Mode. The opt-out is real. The leverage it creates is not — and Google knows it.
Ideogram 4.0: The First Open-Weight Image Model That Can Actually Render Text
Ideogram 4.0 releases as an open-weight model with native 2K resolution and top-ranked text rendering. It outperforms Midjourney v8 on DesignArena and enables fine-tuning on proprietary datasets.
OpenAI Says Chat Is Dead — Here's What ChatGPT Becomes Next
OpenAI is rebuilding ChatGPT as an autonomous agent superapp. CPO Thibault Sottiaux's vision: a personal agent for everything in your life, personally and professionally.
Perplexity's Hybrid AI: Routing Sensitive Data to Local Models, Everything Else to the Cloud
Perplexity's new hybrid AI architecture automatically routes tasks based on sensitivity and compute needs — financial and health data stays local, compute-heavy tasks go to cloud models.
Perplexity's Search as Code: AI Models Now Write Their Own Search Pipelines
Perplexity Search as Code lets AI models generate custom Python search workflows instead of calling fixed APIs — delivering 85% fewer tokens and leading competitors on 4 of 5 benchmarks.
Turing Award Winner Sutton: Pure Generative AI Can't Do Real Science — Here's Why
Richard Sutton argues that generative AI lacks the evaluation loop required for genuine discovery. His argument draws on AlphaGo, AlphaFold, and Claude Code — AI systems that can test their own outputs.
This Open-Source Voice Model Listens Nonstop and Decides Every 0.4 Seconds Whether to Respond
Audio Interaction is a new open-source voice model that makes real-time decisions every 0.4 seconds whether to speak or stay silent — no waiting for you to finish, no push-to-talk, no submission step. Available now under Apache 2.0.
Meta's Hatch Is Its First Paid AI Agent — And It Costs Up to $200 a Month
Meta launched Hatch, its first paid AI product: an autonomous agent that builds tools, schedules appointments, and automates tasks. Hatch Plus tops out at $200/month — direct price parity with OpenAI and Anthropic's premium tiers.
Agentic AI in Healthcare: The Rehumanizing Effect
68% of healthcare providers have deployed AI agents. At Hospital for Special Surgery, they cut appeal time from 45 minutes to 5 and improved claim success from 65% to 100%.
ChatGPT Now Builds a Narrative Profile of You — Here's What It Knows
ChatGPT's new memory system creates prose dossiers covering your work, hobbies, and travel. Fact retrieval accuracy doubled to 82.8%. What this means for AI context.
MiniMax M3: The Open-Weight Model That Runs Million-Token Contexts
MiniMax M3 is the first open-weight model to achieve 1M token context at proprietary-class performance — beating Opus 4.7 on autonomous web search while running 4x faster.
OpenAI's Sam Altman: Proactive AI Is the Next Phase After Agents
Altman named proactive AI — always-on systems that act without being prompted — as the single most important shift to prepare for over the next year.
Nous Research Releases Hermes Desktop: Open-Source AI Agent for 300+ Models
Nous Research launched Hermes Desktop, an MIT-licensed AI agent for Windows, macOS, and Linux that runs across Telegram, Discord, Slack, email, and the terminal — with access to over 300 models via Nous Portal.
Uber Exhausted Its Entire 2026 AI Budget in Four Months — Then Capped Claude Code
Uber burned through its full-year 2026 AI budget by April, then limited all employees to $1,500/month per AI coding tool. What the cap reveals about enterprise AI cost management.
Anthropic Project Glasswing: 150 Partners and 10,000 Vulnerabilities Later
Anthropic is scaling Project Glasswing to 150 new partners across 15 countries, with Claude Mythos hunting critical infrastructure flaws. Over 10,000 serious vulnerabilities discovered so far — what this signals for AI-assisted security research.