@edsu
Another indication that LLM tech is reaching/has reached it's limit in capability is that the efforts to make it work "better" are using exponentially more compute - instead of just asking one question and getting an answer, we have to generate (say) 6 questions, get 6 answers, then check the 6 answers and choose one. So instead of 1 question, 18 questions plus an evaluation/comparison. Yikes.
And it is still not great at finding the right answer.

@edsu
I found Prof. Zhao, tenured prof at U Chicago, described the current state of LLM research very well.
If you want to point out a snippet to stakeholders who do not have a lot of time, I recommend:
30:51 -> 34:58.

@edsu it was brilliant - I told the organisers in the feedback form that this single presentation made the trip to Chicago worthwhile. I’ve been sharing the link to slides & recording far and wide!

PS. thanks @mjgiarlo for dropping it in the $work Slack so those of us who weren't there could benefit as well.