So what exactly does the AI / LLM future look like when the data it’s trained from is all previously AI / LLM generated garbage? Seems like an accelerating circle of trash models built on trash input. πŸ€·πŸ»β€β™‚οΈ

Reply to this note

Please Login to reply.

Discussion

You may be ahead of the curve, and you're on nostr. The fringe πŸ˜†

True. Garbage in is garbage out. The biggest thing the models still need is human curated data for its specific use case. I think the training and such has gotten there but we have a long way to go for awesome datasets with superb unlock. I anticipate it in the tech sector more since protocols and standards are easier to train on than more arbitrary stuff.

That makes sense. Instead of the general. So are you experienced in AI modeling too or techsavvy dev slash fan etc πŸͺ­ (had to use this emoji first time , a fan lol πŸ˜†)

I work with big brains for a day job and we're always talking how we "do the AI" for cyber stuff.

Wait until AI on the blockchain. Its going solve all the problems trust me bro 🀣🀣. Here is our token.