Content-type: text/html Downes.ca ~ Stephen's Web ~ A Cautionary AI Tale: Why IBM’s Dazzling Watson Supercomputer Made a Lousy Tutor

Stephen Downes

Knowledge, Learning, Community

This story may well gain a lot of traction. But one wonders why we should draw any lessons from an AI initiative launched in 2011 and abandoned five years later. That's the story this article tells us, though, and we're supposed to conclude on that basis that AI is pretty limited. But as Ethan Mollick writes, "The current best estimates of the rate of improvement in Large Language Models show capabilities doubling ever 5 to 14 months." You can't base your predictions on what AI can or cannot do on what it's doing today, let alone ten years ago. No small number of populist writers will try to appear to readers' fears and emotions about AI by assuring them that they're still special and privileged. But AI is, in it's own way, a Copernican revolution.

Today: 0 Total: 935 [Direct link] [Share]


Stephen Downes Stephen Downes, Casselman, Canada
stephen@downes.ca

Copyright 2024
Last Updated: Dec 11, 2024 07:13 a.m.

Canadian Flag Creative Commons License.

Force:yes