Below is footage of IBM’s supercomputer Watson taking on human challengers in Jeopardy. It is amazing, and I predict this will mark a huge turning point for humanity; it will be when the believability of artificial intelligence becomes mainstream.
I have to confess that I have always been bored by discussion about AI and singularity from Robin Hanson and others. I couldn’t get interested in debates about whether AI would be like this or that, or whether it would… um… see, it’s never even interesting enough for me to read far enough past “AI will be…” to even tell what it is they’re discussing about it.
This doesn’t say anything about the topic or Robin’s presentation of it, but simply reflects my own shortcomings in caring about something which seems, to put a Hansonian point on it, so far away. I’ve always heard that AI will likely happen and that it’s just a matter of time. I had no reason to doubt this, but still, it didn’t feel true. This, I know, is not an mindset to be proud of, but I think I’m probably close to the median on this. As in most things, I’d venture I’m far closer to the median than Robin anyway.
But now, with the spectacle of Watson, and after seeing this short clip of him, it changes things. AI and singularity suddenly feel near enough to care about. It feels believable, and so it feels suddenly much more important. Today for the first time ever I could really picture, in a near rather than far sense, all of my skills being replaced by a computer and myself arriving at near zero marginal productivity. I imagine it’s similar to what it would be like to grow up in a pre-flight age, and to be told that human flight is a scientific possibility. Sure you may factually accept the science behind the claims, but until you see it at least almost happen, it’s not nearly real.
I probably should have cared about this sooner, and there is a lesson here about thinking about and caring about things your gut tells you are far away but the facts tell are closer than they feel. I guess that’s why it’s important to force yourself to read Robin Hanson even if your gut tells you the topic feels boring and implausibly far. Maybe this means I need to start caring about cryogenics.
This is certainly going to change how people think about AI, and I’m beginning to wonder if Ken Jennings versus Watson will be the most watched television event of all time.