i keep running into this while building fudge: agents are much more creative when you show them what “good” looks like b

i keep running into this while building fudge: agents are much more creative when you show them what “good” looks like before asking them to make anything. if i ask for “an editori…

i keep running into this while building fudge: agents are much more creative when you show them what “good” looks like before asking them to make anything. if i ask for “an editorial yearly recap for an e-commerce platform”, i’m probably getting something competent and forgettable. if i give the agent shopify’s winter editions site, a strange magazine layout, an unusually good product launch, and a couple of references from completely different categories, the ambition of the result changes.

Transcription

Sous-titres réalisés par la communauté d'Amara.org Sous-titres réalisés par l'Amara.org

Voir la source originale

À la une

ANTHROPIC 🔥: Claude limits will be permanently increased by 25% starting from September 14. > Applies for Pro, Max, T

post

threads

·

273 likes

Enter Pro

listing

·

#1 · 199 pts

This Tiny Robot Duck Learned to Walk With Reinforcement Learning Microduck is the tiny biped duck robot that actually

video

instagram

robotics

reinforcementlearning

embodiedai

·

112,7 k vues

PageIndex

listing

·

#1 · 177 pts

Dans la même veine

video

threads

·

35 likes

Over 11k+ sections, 3k+ fonts from 1.8k+ websites for your AI to design better

video

threads

·

17 likes

Can your AI vibecode this ?

video

threads

·

3 likes

Stop using text prompts use image prompts. Agents design better with visual references. Show and tell. explorer.withfud