A big problem is that LLMs are only useful when the facts don’t matter, but people are using them even the facts matter.
So much of the world requires as close to zero mistakes as possible. LLMs have mistakes baked into the underlying foundation of the technology, and there hasn’t been a way to remove that flaw. We are already hearing many stories of AI creating security issues, deleting databases.
LLMs are a toy unless used by someone who can do the job without the LLM in the first place.
A big problem is that LLMs are only useful when the facts don’t matter, but people are using them even the facts matter.
So much of the world requires as close to zero mistakes as possible. LLMs have mistakes baked into the underlying foundation of the technology, and there hasn’t been a way to remove that flaw. We are already hearing many stories of AI creating security issues, deleting databases.
LLMs are a toy unless used by someone who can do the job without the LLM in the first place.