Yesterday I gave an invited lecture to a local high school on Artificial Intelligence. After the lecture, three students came to ask me if it was possible to make a robot that would love a human.
That was not the only question, of course. I was asked about robot emotions, robot animals, flying robots, etc. One student asked my favorite SF books, which made me quite happy. The most “technical” question came from a student who asked me to detail how AI is used to detect spam mail.
These students reminded me why I started to study CS, artificial intelligence, and genetic algorithms in the first place. Robotics was for my generation what the Space Race was for the previous one. My first PhD proposal was on trying to teach robots to learn from the example of other robots, using ants and children as inspiration.
Since entering academia (even though I’m still strictly on the kid’s side of that pool), I’ve been bombarded from all sides with “grown up” talk. Scientific rigour, chinese rooms, mathematical correctness, publish or perish, statistical models, tenure committees, patents, liabilities, reviews, peer reviews, proper academic writing style, critical thinking, no free lunch, practical applications, etc etc.
Most of these concepts are important, some critically so. Still none of it was what put me where I am today. And none of it will bring those kids into the fold.
Recently I’ve been feeling disheartened with research in general. That might be because it has been a long time since I let myself consider my SF daydreams as legitimate research ideas. The other day I was thinking about a Shadowrun-lile PAN controlled by intelligent agents. I immediatelly filled it in as a “personal project”. Why not as a research subject? Because it didnt fit the mold I had made for “worthy of proper scientific investigation”.
That is bullshit. And up until now, it had influenced the way that I thought about my research work. As the high school kids asked me about flying robots and loving robots, I asked myself – why did I just spend 15 minutes in my talk explaining the difference between Weak and Strong AI? Why didn’t I just talk about this instead? When I told them about my favorite GA anedocte, he literally braced himself against a table in excitement.
Next time I give a High School talk, the title won’t be “Introduction to AI research”, it will be “Why I fuckin love AI”.
Maybe my next academic research plan should have the same title.