Solve AI, and Use It to Solve Everything Else.
A couple of years ago, during my master’s degree, I found myself face to face with some of the most complex problems humanity was facing and, unfortunately, is still facing today: climate change, social inequality, and environmental degradation. These problems were presented to me as extremely complex and crucial to solve —which they definitely are. Weirdly, though, the proposed solutions did not feel right to me. A couple of months later, when I was required to write my master’s thesis I simply had no idea what to write about. I had no solutions that I believed in, to the problems that I was supposed to tackle, and ended up never writing my master’s thesis. Maybe, in some strange way, this blog is my search for these answers after all.
So how do you solve climate change, social inequality, and environmental degradation? I still don’t have the answers, but I’m quite sure now that we will find the answers by becoming more intelligent using intelligent systems. Today, we simply aren’t smart enough. Higher intelligence seems to be the solution to solving all our complex problems.
The two people who have convinced me of this are two of the most reliable and trustworthy researchers in AI that I’ve come across, Andrej Karpathy and Demis Hassabis.
Andrej’s original statement goes like that: “Ignore all problems, solve AI, and use it to solve everything else. I don’t think it’s a good idea to go after it [complex problems] specifically. I don’t think that humans will be able to come up with the answer. I think the correct thing is to ignore those problems, and you solve AI, and then you use that to solve everything else. And I think there is a chance that this will work, I think it’s a very high chance. That’s the way I’m betting at least.”
A slightly different version, which is probably what he really meant, goes something like that: let’s do our best managing our current problems with the tools we got, at the same time, focus on solving AI, and then, use AI to solve the problems that we still don’t know how to solve.
In January, I wrote about why AI is possible to build and if we want to take this path we must rethink our priorities and allocate more of our current resources towards it. There is no guarantee that this will work, but it seems plausible that if we are able to solve our complex problems one day, it will be by taking this path. The most important thing is that we get organized and remain determined enough to keep going until we solve AI.
I understand that this approach to solving complex problems might seem incomprehensible or even unimaginable. That’s ok. The whole point of building AI is to build systems that can understand, and learn strategies and solutions for themselves. We don’t have to comprehend or imagine the solutions, we just have to understand that AI is possible and put our resources towards it to build it. AI will then figure out the solutions for us.
As Demis believes, if we do this right, the invention of powerful AI will be as transformative as the invention of fire and electricity. Climate change, social inequality, and environmental degradation, we are coming for you!