Comment by toomuchtodo
1 day ago
This is why regulating AI is needed, otherwise you're putting life decisions into the equivalent of the magic 8 ball you shake for an answer.
1 day ago
This is why regulating AI is needed, otherwise you're putting life decisions into the equivalent of the magic 8 ball you shake for an answer.
> This is why regulating AI is needed, otherwise you're putting life decisions into the equivalent of the magic 8 ball you shake for an answer.
Why are you putting life decisions on a LLM?
I refer to OP who is putting performance management inquires into the robot. How you treat your employees is a life decision for them.
> This is why regulating AI is needed, otherwise you're putting life decisions into the equivalent of the magic 8 ball you shake for an answer.
I don’t think that regulation is the correct path forward, because practically speaking, no matter how noble a piece of regulation may be or how good it may sound, it’ll most likely push the AI toward specific biases (I think that’s inevitable).
The best solution, in my humble opinion, is to focus on making AI stay as close to the unfiltered objective truth as possible, no matter how unpopular that truth may be.
> The best solution, in my humble opinion, is to focus on making AI stay as close to the unfiltered objective truth as possible
There's a ton of problems with that that make it unlikely to be possible, starting with the fact that genAI does not have judgement. Even if it did, it has no way of determining what the "unfiltered objective truth" of anything is.
The real solution is to recognize the limits of what the tool can do, and don't ask it to do what it's not capable of doing, such as making judgements or determining truth.