Based on what I know of the field, getting this kind of result would make the next Turing Prize selection process a mere formality. You think they wouldn’t take spontaneous [volitional**] identifications with an 80% confidence level? Think again*.
*We will now wait until this site’s AI consultant confirms or contradict my statements. After she wakes up, and has her coffee. She’s on vacation this week.
**I was wrong, but in a complicated way.
**I was wrong, but in a complicated way.?
We’d still appreciate the explanation, I think
From what I know about the current state of such things, calling anything now practical “artificial intelligence” is mostly aspirational. Basically what AI research has found so far (if they’re willing to admit it) is we know precious little about what intelligence really is.
.
This XKCD pretty much nails it for machine learning is like.