Anyone else witnessing a panic inside NLP orgs of big tech companies?

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by iandanforth on Thursday March 16, 2023 @05:37PM (#35189065)

"LLM’s can’t scope themselves to be strictly true or accurate"

This isn't true though the techniques to do so are 1. Not as yet widespread 2. Decrease the generality of the model and its perceived effectiveness.

[ Parent | on HN ]

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by shawntan on Thursday March 16, 2023 @06:37PM (#35189865)

I'm interested to hear what these techniques are. Decreasing the generality will help, but I fail to see how that scopes the output. At best that mitigates the errors to an extent.

[ Parent | on HN ]

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by 256lie on Thursday March 16, 2023 @07:51PM (#35190651)

conformal prediction

[ Parent | on HN ]

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by shawntan on Thursday March 16, 2023 @08:28PM (#35191002)

Predicting a set of answers to some confidence interval would still result in hallucinated answers.

[ Parent | on HN ]

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by HDThoreaun on Thursday March 16, 2023 @10:19PM (#35192075)

low probability answers get shot to a human and reviewed for model improvement.

[ Parent | on HN ]

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by fulafel on Friday March 17, 2023 @01:16AM (#35193429)

Requiring the answers to automatically verifiable, or having answers be inputs to a reliable query system?

[ Parent | on HN ]

Re:Anyone else witnessing a panic inside NLP orgs of big tec... by le-mark on Friday March 17, 2023 @07:37AM (#35195648)

I’d rather use the reliable query system!

[ Parent | on HN ]