3 ms·
It seems like some quite strong claims are being made for the AI side of this here. Especially around things like adaptive understanding/learning. Using languag
by kolektiv 13y ago
It seems like some quite strong claims are being made for the AI side of this here. Especially around things like adaptive understanding/learning. Using language like "aware" and "autonomous" is either inaccurate or potentially amazing/terrifying.
That said, it's a fascinating project, and getting mass backing for these kinds of things is going somewhere I'd like to see the world heading.
(Although I share the feelings that toys are really supposed to be about imagination - I've met so many young people lately with poor problem solving/conceptual skills, and I don't think that's disconnected from the increasing prevalence of highly passive young age play activities over the last couple of decades. I've got nothing scientific with which to back that statement though!)
- hexagonc 13y agoIt doesn't take very advanced AI algorithms to satisfy the claim that the toy has adaptive understanding and learning. These are semi-marketing terms and could be covered pretty easily for laypeople and children. You have a set of predefined responses and use a weighted sampling algorithm to choose which one is executed, modified by user feedback or some other fitness test. Genetic and evolutionary algorithms already use a similar methodology.