IdeainventionGenerative AI such as ChatGPT and GPT-4 by OpenAI are increasingly being used by lawyers and law firms to undertake various legal tasks. I've previously discussed five sensible ways to use generative AI for performing legal work and identified prudent strategies for law practices that are adopting the use of generative AI.

One ongoing and overarching mystery about generative AI is that it functions so surprisingly well at analyzing text and interacting in seemingly human-like ways. The fact that a mathematical and computational pattern-matching capability can produce nearly fluent essays and converse with us smoothly is not yet fully explained. For details about the inner workings of generative AI and how research scientists are stymied by these impressive results, see my discussion here.

|

The Logic Is The Issue

I'd like to clarify that it is indubitably feasible to trace the actual computations involved by generative AI, but trying to make logical sense of the complex internal mechanisms is the issue at hand.