There's two separate issues at play here. On the left hand, it's true that ChatGPT, while impressive, on its current incarnation sometimes returns correct respo

There's two separate issues at play here. On the left hand, it's true that ChatGPT, while impressive, on its current incarnation sometimes returns correct responses, and some other times it returns seemingly correct hallucinations. Unless the accuracy and certainty is not over a certain threshold, say 95% for example, I don't think it would be safe to use it for critical use cases, like acting as a lawyer, as it very well might hallucinate laws or prior cases, etc. On the right hand, lawyers see the writing on the wall and see AI as a threat to their really lucrative business, and they'll use any means at their disposal to outlaw it or slow the adoption of AI technologies to replace lawyers. Hell, I'm a software engineer and I see the writing on the wall too, and I see AI as a threat as well. I also acknowledge the limitless opportunities. I'm at equal parts excited and terrified by what's coming.