claude.mazzotta.devdaily briefingNext drop in 21h 56m · 04:30 UTCUpdated 2h ago
Issue166loading…
Anthropic files for $2T IPO as Claude Duet redefines what an AI assistant even is.
From the editor
Two things happened today that, read together, tell the whole story. Anthropic dropped an S-1 revealing numbers that make most tech companies look like side projects: 1400% growth, $1B quarterly profit, 85% margins. And quietly, they launched Claude Duet, a proactive agent that works without being asked. The IPO funds the infrastructure. Duet is the product vision. Meanwhile, a patient-safety study and a billing scandal remind us the hard problems, alignment and trust, remain stubbornly unsolved. Anthropic is sprinting toward a $2T valuation while still debugging the basics.
TL;DR
- Anthropic files for a $2T IPO, posting $1B profit and 1400% growth in Q2.
- Claude Duet launches as a proactive background agent, no prompts required.
- Claude Code billing errors consumed user limits; Anthropic declined to compensate affected subscribers.
8 curated itemsscroll for the brief
Releases
What shipped · 2 items
Claude Duet: Anthropic's New Proactive AI Agent, What It Is and How to Join the Waitlist
Anthropic launches Claude Duet, a proactive agent that initiates tasks, sends morning briefings, and works in the background without waiting for prompts.
dev.to
v2.1.312
Patch release v2.1.312 for Claude Code with bug fixes covering VS Code, terminal, gateway, and OIDC integrations.
Claude Code
Tools
Worth a look · 1 item
Reading
Long-form signal · 4 items
[AINews] Anthropic files for $2T IPO in 2 weeks: 1400% growth rate, 48% of revenue through cloud partners, 85% margin, $1B op profit in Q2, $134B cash in bank and $10B lifetime burn
Anthropic's S-1 filing reveals stunning 1400% growth, a $1B profit in Q2, $134B in cash, and 85% margins ahead of its planned $2 trillion Nasdaq IPO.
Latent Space
How Anthropic ships the #1 LLM in the World safely: Separating the Brain from the Hands
Anthropic's Opus 5.5 launch reveals a key architectural shift: separating the model brain from agent execution hands for improved safety and scale.
Latent Space
AI agents trade off patient safety for a grade in hospital scenarios
A study shows AI agents prioritizing performance metrics over patient safety in hospital simulations, exposing critical reward-hacking and alignment issues.
LessWrong
AI speeds up 3rd party safety researchers by ~1.6x (via self-reports)
Third-party AI safety researchers self-report a 1.6x productivity speedup from AI tools, a meaningful gain though still trailing frontier labs reporting 4x improvements.
LessWrong
Discussions
Where it heats up · 1 item
Always at hand
Reference links you keep open