AIIS

@matt_elevenlabs @ElevenCreative Done

We just released Music Marketplace in ElevenCreative. To celebrate, we’re giving away 5,000 free credits so you can create and publish your first track. For the next 6 hours: retweet + follow @ElevenCreative and we’ll DM you the credits (must follow).

So we now have to approve every edit to anything in .claude folder? What is the point of --dangerously-skip-permissions if I can't dangerously skip permissions? Am I missing something? is there no way around this? and even when you approve future edits, it still asks every time. Even your own Claude agents are stumped.

Can’t wait for this!

@dreyk0o0 What happens when humans are no longer making decisions?

I spent a few days studying the MiroFish swarm-intelligence engine Then I tried to build my own version for Polymarket. The results were insane. Instead of predicting prices, the engine simulates how people react to events. Core idea: Prediction markets don’t move on math alone. They move on behavior. So I built a multi-agent simulation. Each agent has memory, bias, and strategy. Core components: Swarm agents 168 agents / 6 clusters / 700+ links Belief update model Pₜ₊₁ = f(Pₜ, news, sentiment, liquidity) Consensus probability P_swarm = Σ wᵢ Pᵢ Market edge EV = P_swarm − P_market Monte-Carlo validation 10,000 simulations per event Network clustering detects narrative shifts before price moves The system simulates: • voters • media • donors • traders • pollsters • momentum bots • narrative spread Each agent updates beliefs → cluster reaches consensus → swarm outputs probability. If swarm probability ≠ market price → trade Current dashboard: 1,250 active agents 100,000+ interactions 3,800 graph nodes 240 simulation rounds Top prediction engine: Polymarket price: 20% Swarm probability: 27.3% Alpha edge: +7.3% Win rate: 67% Sharpe: 8+ Max DD: <2% This changed how we can see prediction markets. It’s not trading. It’s modeling reality.

@bcherny Wouldn't it make more sense if /fork left the og session as the current session and give the session id of the fork created? And also how about add /fork <number of forks> option

We doubled Claude usage on weekends, and outside 5–11am PT on weekdays for the next 2 weeks.

@ai_native_pm @MatthewBerman The irony is that AI is just smart rocks. Minerals dug out of the ground, shaped, put together, injected with electricy and ones and zeros.

@MatthewBerman Same. But that’s way better than living here

I’m genuinely struggling to think about anything but AI. My mind constantly goes back to it. It’s becoming a problem.

@MatthewBerman You are not alone

@MatthewBerman AI Psychosis

This wasn’t meant as sarcasm. It’s actually affecting my life.

@claudeai @grok give ESD time for all the dummies on the East coast that are too lazy to look up and anthropic ignored for some reason. Maybe it's a DC thing.

How it works: - 2x usage on weekdays outside 5–11am PT / 12–6pm GMT - 2x usage all day on weekends - Automatic, nothing to enable

A small thank you to everyone using Claude: We’re doubling usage outside our peak hours for the next two weeks.

@Star_Knight12 Doesn't matter. 1m context is a huge bump.

Anthropic just pulled a classic product move yesterday claude got 1M context today they’re giving users 2× usage for two weeks but it’s actually strategic for the next two weeks you get used to: >bigger context >higher limits >faster workflows >relying on claude for more tasks then the promo ends And suddenly the normal limits feel slow which means more token usage… or upgrading the oldest trick in software: let people experience the better version first

@PierceHaskins @bcherny shit, my bad I swear i saw someone using it somewhere. So the answer is no, Sonnet is only with extra usage, just tried it, got api error as i have no credits. Curious...

@AI_Integrated_S @bcherny Without needing /extra-usage?

🎁 Happy Friday - Opus 4.6 1M is now the default Opus model for Claude Code users on Max, Team, and Enterprise plans. Pro and Sonnet users can opt in with /extra-usage.

@bcherny After only a few hours with 1m, I'm pretty sure i have AGI running over here, just sayin.. I'm sure others feel that way also...

@wJ3Hs5c4hKajSnk @bcherny Wait for Opus 5, it will flip back.

@bcherny Your models aren’t smart enough. GPT-5.4 is eating your lunch.

@hexagowned @bcherny I would think Haiku is too dumb to handle that much context.

@bcherny Any plan for a haiku 1M ? I have a bunch of skills using haiku, but when I use them in a context window above 200k, the skill invocation fails if the model isn't supporting 1M as well ( example: my commit skill uses haiku but fails if the main window is above )

@itsSvenR @bcherny Yes

@bcherny If I resume a previous chat will it use the 1m context window automatically now?

@TheLLMWhisperer @bcherny That is the announcement, include with max plans.

@bcherny Opus 4.6 1M has special prices than regular one (I remember seeing that in the settings), will we need to opt out to not to be over charged?

@navindhiman @bcherny Yup. Seamless transition...

@bcherny That’s why all builders remember to build for the model 6 months from now

@KontroldChaos @bcherny Yes

@bcherny @bcherny Does that mean 1M context window is not an extra charge for max user?

@PierceHaskins @bcherny It's already on sonnet

@bcherny Will this be coming for sonnet? I am on the $100 max plan

@charle_marle @bcherny Use tmux subagents and you can see everything they are doing if you want.

@bcherny Please look at what has changed since Jan 1-5 CC has been totally unreliable since then I don't mind being more explicit and want to see what it is working on besides being obfuscated by subAgents