I'm going to try to return to my journaling in the form of unfocused technical
rambling. I have vowed, over the past many months, not to write any journal
entries composed entirely of complaints about my life, my job, or the world in
general. This has had the unsurprising effect of me ceasing to write
altogether. As the age-old mantra goes, “If you have nothing good to
say, don't say anything at all.”
I'm also going to start with this post by tagging everything I write
appropriately, and hiding gory levels of detail behind tersely descriptive
cuts, so that those of you who aren't interested can easily filter it out as
you scroll down your friends list, and we'll all remain on speaking terms.
So, human computation is really interesting, and although many parts of it are
quite straightforward in concept, there are a bunch of interesting nuances. If you watch
this Google
tech talk, there's a question at the end from an audience member about the relationship between doing boring, repetetive work
and playing addictive games. And if you take those ideas to their logical extension, we could all be playing human computation games while teaching machines how to do the unpleasant or difficult (from the perspective of machine learning) tasks, and arduous labor being done by humans would become a thing of the past. The only people required to do “real work” then, would be those converting human computational tasks into fun games for others, and since this is a creative art, they would probably enjoy it anyway.
As an aside, one thing that
Paul pointed out to me
recently is that it makes sense to watch some online content, such as Google
Tech Talk videos, faster than realtime. You can watch someone's presentation
at 1.5 x for instance, with pitch-shift compensation, and it will sometimes
make up for the fact that the person punctuates their speech with
“um,” and, “uh,” and other thoughtful pauses. He also
points out that the people in these videos mostly just seem more excited about what they're
presenting when you play them back at higher than normal speeds, as if they just can't expound this knowledge upon you quickly enough. If
anything, it will make you pay better attention. You
should try it (but possibly not with the above video on human computation,
because Luis Von Ahn already talks pretty quickly). You can do this in QuickTime in the A/V Controls window. Finally, the solution to
getting more time out of life.