Pedro Arantes

@arantespp

Building stuffs using AI

Born 1993

Pedro Arantes
@arantespp
4h
Fable is best at coding and GPT Astra is best at everything else?
Auggie
@aug5thmusic
18h
GPT-6 Astra has the best result yet on the Bach Benchmark. Its chorale contains no voice-leading errors, and its harmonic palette is sophisticated enough to include a Neapolitan sixth chord. More importantly, it is the first model to ever write passing tones on this benchmark, a
00:00
Pedro Arantes
@arantespp
Sep 2
Stop asking for software development estimatives. My answer is always, "Claude takes 10 minutes". Start asking if is really necessary build the thing.
Pedro Arantes
@arantespp
Sep 2
AI agents have taken over software development, but I haven't seen that happen in other fields yet. It's only a matter of time before AI takes over marketing, HR, legal, accounting, and so on.
1
Pedro Arantes
@arantespp
Sep 2
I asked Fable 5.1 to improve the SOAT homepage, but it wasn't as stunning as I expected. I think the problem might be myself as designer. soat.ttoss.dev
1
Pedro Arantes
@arantespp
Sep 2
.@tobi trained a small model, which was the first time I saw a company fine-tune one after mostly just hearing about LLM news. I think most companies won't succeed at fine-tuning specific models. The only exception is if a company has highly proprietary data that no one else