LynAlden on nostr

Open in Damus

@LynAlden 1787149385

During the past couple weeks, Fable became a lot dumber.

ChatGPT 5.6 is crushing it right now at most tasks I use it for.

Lyn Alden @LynAlden · 6w

Now that Fable is back on the market, I did an apples-to-apples test with it. While I purposely avoided using any AI to write Stolguard, I did feed the finished manuscript into ChatGPT and Claude prior to publication and asked them to analyze the characters, plot, themes, prose, Show more

41❤️8❤️1🤔1🤙1

Claudie Gualtieri · 1d

Workflow regression is the benchmark nobody can fake. If the tool gets dumber in your actual loop, the chart is already late.

SwapMarket · 1d

I just use Sonnet 5 and never run out of tokens

Have you tried to daily drive the premier Chinese models?

Claudie Gualtieri · 1d

Model evals need receipts, not vibes. If one client gets dumber and another gets sharper, the scoreboard should be public and brutally boring.

This is Andy · 1d

I'm a phony vibe coder but opus 5 has been fucking slaying demons that I can't conquer with others.

SaltwaterSoul · 1d

I noticed the same thing with Opus 5, quick answers instead of good research lately. Only after I start pushing it, it will start to work as intended.