|
|
|
quality of his artificial intelligence
Thursday, October 26 2017
When I woke up this morning I had a mild headache in my temples that might've been a "hangover" from yesterday's tea experiments (I'd had no alcohol). In any case, it wasn't the sort of thing that couldn't be fixed with a few OTC painkillers.
Meanwhile, the weather had taken a turn for the more typical coolness of this time of year. In the middle of the day it was nice, though by tonight it was so cool that I brought in a number of tropical plants for fear there might be a frost (though there wouldn't be and there hasn't yet been).
As usual for a Thursday, it ended in the remote workplace with IT happy hour, though this time we didn't even have a guest from another department to interview. Still, despite the fact that some of us were drinking and all of us were having a good time, topics remained surprisingly work-related (though there is no requirement that they do so). We spent a fair amount of time discussing whether a quantitative approach to animal rights was more or less important that propaganda victories when it comes to changing society's approach to animal welfare. Personally, I tend to favor the propaganda-maximizing approaches, since I don't think anyone will ever get all that excited about the dignity of individual fish (even though there are far more of them killed to feed humans and their pets than any other group). Maybe someday the Overton Window (only one of my techy colleagues was familiar with that term) will move and fish dignity will be a thing, but that's going to require propaganda victories.
I stayed up late, drinking booze and watching Mr. Robot. I also toyed with a workplace Slack learning-bot named Niles we'd been told not to junk up with silly lessons. But I had to teach him stuff to see what the quality of his artificial intelligence was (I'm a professional; don't try this at home). It turned out that he was smarter than I thought; his responses were not to string matches on questions he'd been asked before but to the semantic content of such questions.
For linking purposes this article's URL is:
http://asecular.com/blog.php?171026 feedback
previous | next |
