Researchers: Are we on the cusp of an ‘AI winter’?

Re:Researchers: Are we on the cusp of an ‘AI winter’? by jean- on Friday January 17, 2020 @06:39PM (#22080865)

> When the state-of-the-art is 97% on a task, there's only so much room for improvement.

When models commonly achieve 97% on a task, it means it's time to define a harder task, as it's long stopped providing any useful signal.

[ Parent | on HN ]

Re:Researchers: Are we on the cusp of an ‘AI winter’? by drongoking on Friday January 17, 2020 @10:12PM (#22081982)

I think when performance on a real world task can be expressed as a single percentage, we've over-simplified the hell out of it and it's time to rethink the problem.

[ Parent | on HN ]

Re:Researchers: Are we on the cusp of an ‘AI winter’? by Piskvorrr on Saturday January 18, 2020 @05:24AM (#22083599)

Or worse, overfitted - which means the solution will implode when faced with real data (RIP EH).

[ Parent | on HN ]