981–990 of 1001 posts
Funny, I'm just doing my normal coding workflow with Claude Code, and after every change that compiles it keeps suggesting that we're at a good stopping point, and should pick up again tomorrow. It's done this before, but usually doesn't. I bet they're giving it some kind of throttling signal due to high load from today's announcement.
> Software engineering. During early testing, Stripe reported that Fable 5 compressed months of engineering into days. In a 50-million-line Ruby codebase, the model performed a codebase-wide migration in a day that would otherwise have taken a whole team over two months by hand. How was it measured? How was the output of this magnitude verified over a period of couple of days?
> In light of the ability of recent models to accelerate their own development, we’ve implemented new interventions that limit Claude’s effectiveness for requests targeting frontier LLM development (for example, on building pretraining pipelines, distributed training infrastructure, or ML accelerator design). Using Claude to develop competing models already violates our Terms of Service, but enforcing this restrictio…
1: https://www-cdn.anthropic.com/d00db56fa754a1b115b6dd7cb2e3c3...
It's interesting that we're seeing these gains when it seems Mythos/Fable is "just" a scaled up version of their existing architecture[0]. When GPT 4.5 launched, the gains compared to the model size didn't seem that great, leading some to believe that the only progress we'd see would come from RL. This model certainly has quite a "substantial amount of post-training and fine-tuning", but it's also based on a new pret…
There’s nothing much new about the architecture. The real gains come from the usage traces. It turns out that having a text based interface for a text-trained model creates a very nice feedback loop. Right now as we speak, people are generating text traces on anthropic and OpenAI servers that teach their models to do everything under the sun, text wise. So people right now getting super mad at how dumb the model is w…
But it also really, really wants to burn tokens. I asked it to look into a fairly straightforward database bug in my RN app, and while I was off getting coffee it decided to spin up an android emulator unprompted and started navigating the app by reading screenshots and injecting touch events. There went my entire week's tokens. There was no reason to even start the emulator, the bug wasn't graphical, so I have no clue what it was doing.
I've spent enough time with this now in Claude Code (and Claude.ai and Claude Code for web) to have an opinion on Fable 5: it's a beast. I'm throwing some VERY difficult problems at at - things I've been dragging my heels on for months - and it's crunching through them very happily. One that I'm willing to share (albeit from just a week ago) - I built a Python library last week that bundles MicroPython compiled to WA…
As an additional check, I just submitted it to Fable, and it eviscerated it. Tons of inconsistencies found, issues skimmed over or ignored, too optimistic assumptions, math that doesn't really add up if you look at it in context. And as far as I can tell, all of these issues are entirely valid. I now feel embarrassed I'd already sent it to a few people for review. This clearly needs more work.