we keep blaming the tech when the issue is actually
bad product thinking upstream. most ai failures stem from things like
vague prompts and zero evaluation frameworks . it is usually just a lack of
clear intent or setting the wrong expectations before we even open figma. anthropic is even doubling down on hiring more pms to fix this rather than just adding more compute. if you dont have defined outcomes, no amount of engineering will save the user experience. are we all just ignoring the need for better
user requirements in favor of chasing new models?
found this here:
https://uxdesign.cc/most-ai-problems-are-really-human-problems-this-is-how-you-can-fix-it-ab64843f845e?source=rss----138adf9c44c---4