OpenAI fights back
Sol 6.1 is OpenAI's response to Opus 5.5, and it's like Astra, but less dumb, and way way cheaper.
Share on Bluesky: OpenAI fights back (opens in a new tab)Share on LinkedIn: OpenAI fights back (opens in a new tab)
Recent from Theo — t3.gg
OpenAI should be scared of this one
Anthropic surprised us with a Sonnet 5.5 drop today and...I kinda like it? The wildest part though is I don't think you should use it.
Share on Bluesky: OpenAI should be scared of this one (opens in a new tab)Share on LinkedIn: OpenAI should be scared of this one (opens in a new tab)
So much for "Pacing" the Frontier
We need to have a conversation around what "Pacing" looks like as both OpenAI and Anthropic keep pushing out models after their agreement to slow down frontier development
Share on Bluesky: So much for "Pacing" the Frontier (opens in a new tab)Share on LinkedIn: So much for "Pacing" the Frontier (opens in a new tab)
Getting the most out of Opus 5.5
Anthropic put out a blog post with some tips and tricks that they have for the Opus 5.5 release, so lets break them down
Share on Bluesky: Getting the most out of Opus 5.5 (opens in a new tab)Share on LinkedIn: Getting the most out of Opus 5.5 (opens in a new tab)
My new favorite model.
After Opus 5 wrecked my codebases, Opus 5.5 got a full day of real work done on about 20% of my weekly Claude limit, and I'm finally not scared to merge what it writes.
Share on Bluesky: My new favorite model. (opens in a new tab)Share on LinkedIn: My new favorite model. (opens in a new tab)
Elon promised this one would be good...
Elon claimed Grok 4.7 would be more token-efficient, and it's less by 30 to 80%, with very few forgiving qualities
Share on Bluesky: Elon promised this one would be good... (opens in a new tab)Share on LinkedIn: Elon promised this one would be good... (opens in a new tab)
So I was using Fable wrong...
It's time to start prompting better and make Fable 5.1 as good as we can, so you don't have drops in performance
Share on Bluesky: So I was using Fable wrong... (opens in a new tab)Share on LinkedIn: So I was using Fable wrong... (opens in a new tab)
Jev is incredible
Jev is a brand new Typesafe model that's a fast classifier with tons of perks and safety, but it doesn't replace reasoning models like Astra/Fable.
Share on Bluesky: Jev is incredible (opens in a new tab)Share on LinkedIn: Jev is incredible (opens in a new tab)
Please stop using stupid models
The Best AI Models aren't actually better because they're smarter, they're better because they do the dumb things right more often.
Share on Bluesky: Please stop using stupid models (opens in a new tab)Share on LinkedIn: Please stop using stupid models (opens in a new tab)