If you open and read any popular science magazine, you will notice that almost a third, and sometimes half of the articles are in one way or another related to artificial intelligence, neural networks and cloud computing. The rest will be related to new developments in various fields of science. It would seem, what's the big deal? The fact is that every day the amount of research and achievements grows faster and faster, but we still underestimate the progress. We are so underestimated that we cannot even imagine how much the world will change in just a couple of decades. And with a high probability, these changes will occur precisely thanks to artificial intelligence. But first things first.
Progress
This is the most important part to understand the rest. Humanity is now so close to the biggest change in our lives that it can be compared to the very birth of humanity. And it is extremely likely that it will happen in our lifetime. But how can we prove to you that this is really so? Very simple.
Our consciousness is inert, and we perceive history linearly. As we try to imagine the future 30 years from now, we look back at similar periods of time. And this is a big mistake - we look at the path traveled and it seems to us that the next segment will be exactly the same. But this is not the case. You need to look not at the path, but at the speed or even acceleration - then we can more or less accurately assess the prospects. There is a very good analogy to show all this.
If you take a random dude from 1750, when even there was no electricity, and transfer it to our time, then most likely what he sees will literally break his brain - our world will be so shocking for him. And this is even before we show him computers, the Hubble telescope and nuclear weapons.
But what if this guy decides to repeat our trick? Most likely, he will try to transfer a person from 1500 to his time. But 250 years won't be enough - the difference between 1500 and 1750 isn't that great. He learns a lot of new things, will be surprised by the development, but the world will remain practically the same for him. For everything to work, our hero will have to take a person from about 12,000 BC. Just at the time when the First Agrarian Revolution led to the possibility of creating cities and developing civilization. Then an almost devastating effect will happen for an alien from the past - majestic buildings, huge ships and the shock of the knowledge that we have accumulated over 13,000 years.
Promotional video:
A more developed society is accelerating faster and faster. Progress is described by an exponential curve, not a linear one.
As you can imagine, in order to have fun in the same way for a person from 12,000 BC, he will have to go after a test subject more than 100,000 years ago - for a test subject who will be surprised at the fire and the existence of language. Now think how long it will take you to move into the future to die of shocking progress? Ray Kurzweil (one of the most famous and accurate futurists) believes that already in 2030 the man of our time will be as shocked as that man from 1750.
That is, you probably simply will not recognize your own world in 2030, it will be so different. To believe in this, just stop being stubborn, looking back at your experience and the picture of the world that has developed during your life.
According to Kurzweil, the equivalent of progress for the entire 20th century was made between 2000 and 2014.
Artificial Intelligence
How does the development of AI relate to all of the above? In the most direct way. AI already controls the world, or rather, its individual aspects. From Google Translate to kicking ass at the best Dota 2 players (OpenAI) to coordinating air travel and curing diseases (Watson). We often do not even realize that we are using AI and are dependent on it right now. But there is one "but": all the AIs that we have at the moment are weak AIs. Or AI, specialized and sharpened for one specific area in which they are superior to humans.
The second category of AI that we are trying to create is strong AI - human-level artificial intelligence. This is extremely difficult at the moment, but we are getting closer and closer. The best minds of mankind, researchers and scientists do not even ask the question: "Can we create such a thing?" They ask themselves the question: "When?" The third category is the Artificial Superintelligence.
The last category is the most interesting and exciting. We are only at the beginning of the path to creating superpowered intelligence, but given the exponential growth of technological development, one day we will come to it. Supercomputers have now reached and even surpassed the computational capabilities of the human brain. And we continue to speed them up, despite pessimistic beliefs that Moore's Law no longer works. We have already been able to copy the work of the worm's brain, we will soon copy the brain of a mouse, and then more complex creatures.
When we create a full-fledged strong AI, the world will change dramatically. We will literally create something that will bring us even closer to Superpowered AI. AI devices will not be limited by speed, size, memory or reliability. They will be better at everything than our imperfect human brain. This means that they will not be limited in their development. And this is the path to the Supermind. The trick is that its development will be the same - exponential. It will take him decades to reach the level of a 4-year-old. Then in a couple of years he will catch up with the fifth grader. And in an hour he will give out the Theory of the Great Unification. In a minute, it will become a Supermind, which is hundreds of thousands of times smarter than a person.
We will create a being, the most powerful on the planet, who will surpass us in everything. So incomprehensible that we cannot even understand it. We will be to him what an ant is to you. And the existence of humanity will depend on this Supermind. Everything we dream of: immortality, control of the weather, knowledge of all the laws of the Universe - all this will be possible for him. Just like the destruction of life on Earth. When we create it, there will be only one question: will it be kind?
Based on materials from Tim Urban
Evgeny Entropov