Nosing through @tmbchr’s post on the WIRED post about humans augmenting AI, I have only 2 things to say:

1) humans being a helpful extension sounds like a nice idea… especially as presented in the article, but

2) Those “helpful” tasks stand a high chance of being crappy, drudge-filled exercises of boredom.

Picking out ‘what’s beautiful’ from a series of images (the example given) is one out of a gazillion learning tasks we can help our fledgling AIs with, and one which happens to be amusing / pleasant. I can think of countless others that aren’t / won’t be. As I keep saying, smarter machines will only leave smaller and smaller islands of finnicky complexity for their fleshier counterparts (that’s you and me, compadre) to deal with. Until they master those islands themselves and sink them completely. Hopefully by then we’ve found new uses for ourselves, cos those AIs sure aren’t going to have any for us.

But I do love the idea of people augmenting machines (see, I’m not all doom n gloom)… I just doubt its panning out as painted. Some other way, maybe…