|
RPG Imaginings@rpgimaginings.bsky.social |
I'm battling with a colleague right now who has a good heart but is ignoring the environmental implications in the "usefulness" discussion. We've long held that best practice is "training the right way," but I'm wavering on whether even training is wise. :/
1 replies 0 reposts 1 likes
|
Rick Schmelz
@rickschmelz.bsky.social
[ View ] |
IMO, current LLM/GenAI models and implementations carry too many cons with not nearly enough pros to make them worth the trouble. They require theft, they cost too much, they consume too much across many domains, and they're not good at their stated uses.
1 replies 0 reposts 1 likes