News for hackers. Stuff that matters.

A comment on Machine learning won't solve natural language understanding. Parent comment

Re:Machine learning won't solve natural language understanding by SavantIdiot on Monday August 09, 2021 @09:19PM (#28124257)
Will the AI be able to ask for clarification? Translators do. In fact, even people speaking the same language need to ask for clarification sometimes. It is literally impossible to discern some sentences without follow up, why would someone think an AI can do it one shot?
[ Parent | on HN ]
Re:Machine learning won't solve natural language understanding by agravier on Monday August 09, 2021 @09:22PM (#28124268)
Why wouldn't an AI be able to ask for clarification in a live setting? I don't see your point.
[ Parent | on HN ]
Re:Machine learning won't solve natural language understanding by joe_the_user on Monday August 09, 2021 @10:22PM (#28124582)
Well, the approach of solving a corpus (advocated by GP) doesn't leave room for the possibility of asking a follow-up question so you can't train for it that way.
[ Parent | on HN ]
Re:Machine learning won't solve natural language understanding by visarga on Tuesday August 10, 2021 @02:18AM (#28125640)
GPT-3 was trained on a massive text collection so it can even ask for clarifying questions if the prompt context allows for it. Why? Because there are plenty of such examples in the training data.
[ Parent | on HN ]
Re:Machine learning won't solve natural language understanding by drdeca on Wednesday August 11, 2021 @02:30AM (#28138749)
Sure, but this is based on something kind of like “whether someone would be likely to ask for clarification here”, not based on whether GPT3 needs clarification.

Though, I’m confident that it should be possible to make something which, when unsure, asks for clarification.

In fact, I believe something at least rather similar is already done in, uh, I think it is called “delegative reinforcement learning”, where an agent takes actions, but when it is highly uncertain, it can instead choose to have an expert (who is assumed to be competent at the task, or at least has a tolerably low risk of very bad outcomes) to step in instead.

(Or, wait, maybe delegative reinforcement learning is still just used theoretically? Not sure.)

[ Parent | on HN ]
Re:Machine learning won't solve natural language understanding by rubatuga on Tuesday August 10, 2021 @02:09AM (#28125597)
It could list all possible solutions based on the response to the followup question it would have asked.
[ Parent | on HN ]
Re:Machine learning won't solve natural language understanding by mcswell on Tuesday August 10, 2021 @06:18PM (#28135421)
Translators may do so when they're acting as interpreters. I don't know for certain, but I doubt that most translators who are translating written documents do so.
[ Parent | on HN ]