What is your stance on the convergence of AI with humans? And is this convergence the next logical step in our evolution?
Currently, I think the way that people are looking at convergence has to do with humans just collaborating with AI. Fundamentally, the idea of AI started with the idea of an autonomous system that would make decisions on its own, and leave people out because humans are sort of messy. Now people say that we are not really ready to accept fully autonomous systems, and people want autonomous systems that collaborate with humans. There is a lot of human-computer interaction work right now, which tries to see the computer more as a support to humans than a decision maker. However, collaboration is not exactly integration. Collaboration is a step in the direction of allowing humans to make better decisions with the aid of a computer, but humans will still have the final say. This is the direction in which the AI community is going currently.
There is a slightly different approach, which has to do with putting chips into humans, and that is more of a true convergence of humans and computers. I don’t know how much success this has had, but there has been work done in this area for a while. For instance, implanting small chips into peoples’ brains or arms and so on, so that you can just unlock your door with a chip instead of a key. However, this has been quite controversial because of privacy and security issues. Suppose a chip is implanted in your hand or finger so you can open the safe of a bank. Then one day somebody chops off your finger, and now has access to the safe. When you think about chips from the angle of security, there is definitely a big issue because the part of the body that has this chip can be removed and used by someone with different intentions. There is also the issue of privacy when it comes to chips. The issue comes from the fact that if you have a chip, every one of your movements is recorded, so you are being tracked every second of your life. And, of course, when you are tracked you don’t own your own data. Somebody else has access to your data. This is obviously an intrusion of individual privacy. Now, some people don’t really care about this specific privacy issue, but most people do.
On another note, when you think about implants, like pacemakers, people accept them because they save lives. Now people are talking about implanting chips so that blind people can see. Personally, I think that people will accept convergence with tech because if the benefits outweigh the potential risks, why not? I mean, of course, convergence is limited because you are not truly replacing your brain with a computer– which I think is crazy–but convergence can give you help with some functions.
What do you think is the biggest misconception about AI?
I think because people don’t really know what this technology really does, people tend to overestimate and underestimate the abilities of AI. Either people say, “Oh no this is stupid it will never happen,” or, “Oh the computers are going to take over the world and killer robots!” And all of those sorts of things. People always tend to go to the extremes when it comes to AI because they don’t understand enough about the technology and reality of AI, and the current reality is in the middle of these extremes.