When you read about Artificial Intelligence, there is always a huge debate on whether or not we are getting closer to true human level Artificial Intelligence. Lots of people claim that AI is completely possible, its just that we as a human race keep moving the goalpost, or changing the definition of AI. And so researchers keep making breakthroughs in AI, but it never seems to be enough.
It usually happens like this: There is some hard problem that scientists think is the related to the core of intelligence like computers being able to play chess. Computer scientists work on the problem for a long time. Someone solves it and now computers can play chess and beat humans. Lots of people tests it out and see that it can play chess, but notice it is still not intelligent. They then declare that chess wasn’t important for AI and instead if you solve some other problem X, then AI will be solved. One of the current next generation problems that researchers are working on is self-driving cars, and many people do believe that if we solve this, we will have true AI.
The goalpost for AI has never been moving. These problems that researchers focus on are chosen because they are small and more achievable than the grand vision of Artificial General Intelligence (AGI), Strong AI, or human level intelligence. The goal has always been to create intelligence in a computer that mimics human intelligence, but the problem is too hard. It will be easier for us to get humans to Mars, invent autonomous vehicles, super batteries, and many other tough problems. In fact, creating an AI is probably the hardest problem humanity will ever work on. True AI is up there with nuclear fusion and light speed travel. Unlike those other problems though, we know it is possible because we see it with the birth of every human baby.
AI and our understanding of the human mind is still in the stone age, but there is nothing wrong with that because we are getting new inventions that help improve our lives everyday.
So instead we work on these smaller problems with the hopes that we are chipping away at the grand problem of human level intelligence. Maybe they will help towards human level intelligence, but most likely they won’t. I still encourage anyone interested to do AI research because we never know where a path may take us…..