Reading up on Opus 4.5
8 deep · digging since nov 28, 25
- Better AI models enable more ambitious work
A University of Chicago study of 500 companies found AI usage rose 44% as models improved, with developers shifting to more complex and cross-system tasks.
- Scaling long-running autonomous coding
Cursor's multi-agent system, using planners and workers, scaled to run hundreds of autonomous coding agents for weeks on projects like building a browser, finding simpler coordination beats complex distributed-computing approaches.
- When AI writes almost all code, what happens to software engineering?
AI coding tools like Opus 4.5 and GPT-5.2 now generate 90%+ of code, fundamentally reshaping software engineering roles and skills.
- 200k Tokens Is Plenty - Amp
Amp's CEO argues that short LLM threads under 200k tokens produce better results, lower costs, and simplify complex coding tasks compared to monolithic long-context sessions.
- Horses: AI progress is steady. Human equivalence is sudden
AI progress resembles horse-and-buggy obsolescence — steady efficiency gains mask that human-equivalent disruption arrives suddenly, not gradually.
Takes
We've gone really quickly from "local models are dogshit" to "local models are good actually" (like, a 12 month window from A to B). I don't think they're actually good ENOUGH yet. We need an Opus 4.5 quality local model. When that happens, I think the world will spill over. Opus 4.5 is/was amazing, and is more than good enough for almost all tasks still as long as you pair with a frontier-level planner/judge. It'll still require a hugely expensive machine to run it, I'm sure, like a $5K or more laptop or mac studio. But, that's going to be pennies compared to the API costs plus all the benefits of guaranteed privacy and so on.
@mitchellh
Since we launched Claude Code, we saw people using it for all sorts of non-coding work: doing vacation research, building slide decks, cleaning up your email, cancelling subscriptions, recovering wedding photos from a hard drive, monitoring plant growth, controlling your oven.… https://t.co/KNc4YgK6nC
@bcherny
Opus 4.5 is insane. Just shipped v2 of my compounding engineering plugin—watch the video for my full thoughts on the model.Compounding engineering plugin v2: https://t.co/6LI5u1ZHThThis wouldn't have worked a week ago. Previous models would derail after the second parallel… pic.twitter.com/ujUodM4qus
@kieranklaassen