Patryk

I smell... potential.

if you're overly excited about Fable performance it just means your devflow/harness suuuucks

@EmreCoklar @gabrielste1n resonates, rooting for you both!

@gabrielste1n Congratulations! Having been following your journey since Bex, it's been inspiring to follow along. So much of what you said resonates deeply with me and I'm still on the pre-money side of the fence so I found your words inspiring to keep pushing forward. Also, @fullpatstack

$100K Annual Run Rate, Reflecting on the Loss of My Father, and the Curse of Entrepreneurship

you're at 95% weekly usage, monitoring 3 Fable agents and one of them decides to launch 10 subagents

@ImSh4yy Ultra eats through usage like Fable for me

It seems like GPT 5.6 Sol is using my 5 hour codex limit much faster than GPT 5.5 despite it being more token efficient?

@corbin_braun That’s very offensive sir

I am convinced that even if you had access to super intelligence, you still wouldn’t know how to make money. oh wait….

@corbin_braun True but do you actually consider paying API prices for it when it’s no longer covered by sub?

sooner or later I will stop posting about Fable 5 but if you know me, I don't usually over hype things just to hype them up. but to put simply how I felt when I first used ChatGPT 3.5 is happening now with Fable 5. leap years ahead.

@bindureddy Opus>Fable for design…? Please ..

All the best models per use-case changed in the last few weeks.... master agent - fable sub-agents - grok 4.5, 5.6. sol chat - 5.6 terra real-time - grok 4.5 ocr- flash 3.5 design - opus 4.8 images - gpt image-2 video - seedance 2.5 open-source - glm 5.2 A massive leap in intelligence

Do people watch entrepreneurs live streams?

@alexwestco 🔒

cleaning up my feed. only business.

this guy's pretty cool!

in my setup each piece of work delivered by Claude Code is reviewed by Codex. i now added tracking to understand over time how different models actually improve in real world (albeit my world). I am looking for stuff like: - "Opus4.8: 4 iterations, 2 p0 findings on avg" - "Fable5: 2 iterations, 0.5 p0 findings on avg" will post once i have enough data

Idk what this means but it feels right

@TheAhmadOsman Fable limits are reset. Do you hate them enough to not give it a go? 😬

I can say that I have had access to GPT-5.6 for the past hour or so and I like it so far.

Yeah buddy!

@tymotauro @KinasRemek o tak o

@fullpatstack @KinasRemek Jak wy to możecie na oko oceniać xD

Pierwszy test (u mnie standardowy) - GPT 5.6 SOL. Porównuję model do najlepszej obecnie wygenerowanej symulacji one-shot za pomocą Fugu Ultra. Oczywiście to pierwszy strzał ...

i went to the download url for chatgpt work and downloaded it and got this when i clicked on the dmg???? then i updated codex and it changed to chatgpt work?? CODEX WAS SUCH A GOATED BRAND WITH MOMENTUM AND AURA BUILDUP FOR MONTHS AND THEY JUST SHIT ON IT LIKE THIS trillion dollar company trillion employees trillion meetings unlimited inference INCREDIBLE fumble

Gpt5.6, grok4.5 and some Meta stuff

@tibo_maker openai valuation is crazy and can only be defended if they go beyond models - like anthropic is doing with enterprise software. the 'models' part of the business itself is getting commoditised. so the value is in infra (which xai has and invests in) and enterprise software

and with Grok 4.5 matching GPT-5.5, you could argue xAI is worth as much as OpenAI making SpaceXAI insanely cheap (just increased my position)

@DataChaz @krzyzanowskim Finally something that makes sense

this is how new devs are learning Kubernetes now. I'm f*cking done.