840

There has been altogether too much talk about whether A.I. chatbots are capable of thinking like humans, and too little talk about how much humans actually think like A.I. chatbots.

The unfortunate truth is that for the vast majority of humans, the vast majority of the time, we more or less operate like the machines (including you, the brave reader, and me, the wise writer). We get almost all of our knowledge not by actually understanding the world, but by basically just repeating what other people have said. The more something is repeated, the more true it is. It's why propaganda is so successful, and it's why some people have recently put so much money and effort into buying up social media sites. Not so they can actually educate people, but so they can get certain things repeated more often, to train us like they train A.I. chatbots. If something is repeated often enough, most people simply believe it, and start repeating it themselves. It's also why you can predict someone's ideas very well by simply knowing where and when they lived. We seem to mostly just absorb ideas passively in a kind of statistical approach, much like self-learning machines do.

The only way to counter this is for humans to be more like humans, and less like machines. Which means we have to use the one thing we have that machines don't: our consciousness. We have to be conscious not only of our ideas, but where we got those ideas from, and whether or not we actually understand them, and actually know them. This, I suppose, is the role of the philosopher, but ideally we should all be a little bit philosophers. Unfortunately it is a lot of work, so we can't be bothered most of the time. As George Bernard Shaw put it: "few people think more than two or three times a year; I have made an international reputation for myself by thinking once or twice a week."