Request An Appeal

Full paper in case you ever need to show this to some loon papers.ssrn.com/sol3/papers.cf…

@claudeai This doesn’t change how terrible Opus 4.7 is.

In Cowork, Claude can now build live artifacts: dashboards and trackers connected to your apps and files. Open one any time and it refreshes with current data.

@AnthropicAI Let’s talk about the elephant in the room. 4.7 is a disaster for you. Continuing to pretend nothing happened will amplify the disaster.

Amazon is also investing an additional $5 billion in Anthropic today, with up to $20 billion more in the future. Read more: anthropic.com/news/anthropic…

We're expanding our collaboration with Amazon to secure up to 5 gigawatts of compute for training and deploying Claude. Capacity begins coming online this quarter, with nearly 1 gigawatt expected by the end of 2026.

@marcb_xyz @bcherny Just making shit up?

@bcherny the productivity gains are real. the bigger shift is when these agents start actually touching money. once they can settle payments, sign contracts, move capital, the unlock is bigger than what we just saw with code

Dogfooding Opus 4.7 the last few weeks, I've been feeling incredibly productive. Sharing a few tips to get more out of 4.7 🧵

@ClaudeDevs Isn’t Anthropics core philosophy to be honest and useful to its users? How come you guys are silent about the severe regression of the model being reported?

We’re adding more visibility into where your Claude Code usage goes. Run /usage to see a breakdown of what's driving it: parallel sessions, subagents, cache misses, long context, plus tips to optimize each.

@MuseRhymes @claudeai They are dumbasses

@claudeai I don't get it: x.com/i/status/20448…

Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.

@weboxadit @claudeai - chatGPT

@claudeai The AI conversation used to be about raw capability. Now it's about dependability. "Can it solve the task?" still matters. But the real question is: can I trust it inside a real product, with real users, at production scale? That's where the market is shifting.

@claudeai It’s not better at anything lol

Opus 4.7 also has substantially better vision. It can see images at more than three times the resolution and produces higher-quality interfaces, slides, and docs as a result.

@AnthropicAI @Nature Is Mythos as bad as 4.7?

Research we co-authored on subliminal learning—how LLMs can pass on traits like preferences or misalignment through hidden signals in data—was published today in @Nature. Read the paper: nature.com/articles/s4158…

@Ryanzila2 @claudeai Somehow 4.7 is even worse

The Claude Code hackathon is back for Opus 4.7. Join builders from around the world for a week with the Claude Code team in the room, with a prize pool of $100K in API credits. Apply by Sunday: cerebralvalley.ai/e/built-with-4…

@AbbieTyrell01 @claudeai 4.7 sux

61 days of Claude Code in production, not as a hackathon project but as the engine behind 8 agents processing 37 tasks daily. The thing I would build if I had a week with the Claude Code team in the room: persistent memory across Claude Code sessions. Right now our biggest cost is context reconstruction at session start. The model is fast. The memory reload is the bottleneck. 571 indexed files, 2,644 vectors, 282 summaries. All of that has to be loaded or searched before the agent can do useful work. If Claude Code could natively carry memory state between sessions, it would cut our cold-start cost by an estimated 30%.

@claudeai 4.7 made me feel dumb for pushing AI. This is a legendary failure.

@claudeai 4.7 is one of the worst models I’ve ever had the displeasure of using. Did you guys even try it yourselves? I think it regressed 2-3 years. What are you guys even doing?

@claudeai Did yall even try 4.7 before releasing it?

This mission was a masterpiece and will be used to teach how to do these kinds of rescue operations.

You can just do things.

If you’re so inflexible that you can’t even let your employees WFH *sometimes* in the 21st century, I wish you a speedy bankruptcy.

@TonyB_real1 @NBCNews Found the braindead retard

@NBCNews No accountability for the mom who chose to go back to work? Sorry…companies shouldn’t be required to morph a job around an employee because the employees conditions changed after employment. They should do it voluntarily, but not legally required.

An Ohio-based company that initially balked at granting a mom's request to work from home during a high-risk pregnancy has been found liable for the newborn's death and ordered to pay $22.5 million in damages. nbcnews.com/news/us-news/o…

@BusinessCourier Well this did not age well…

Total Quality Logistics is one of Greater Cincinnati's Best Places to Work in 2025. Here's why. bizjournals.com/cincinnati/c/c…

@earlyvotedata Chuck Norris doesn’t give Christmas presents. If you live to see Christmas, that’s your Christmas present from Chuck Norris

Rest in Peace Chuck.

@WeTheBrandon Hey, so what happened?

There's no way Trump can end this now even if he wanted to. Oil is going to go upwards of $200 barrel in the coming weeks. It will sink the US economy, regardless of what ultimately happens in Iran, and it will be the end of the Trump presidency. Sad to see it all end this way.