What is General Artificial Intelligence?
Whatever the future holds, it will be good for AI.It would be nice if it was good for people too. _ Ilya Sutskever
Theoretically, the difference between AGI and AI is that AGI has human intelligence and perhaps even higher cognitive abilities.
But at this point, various experts have different opinions about this term. And the reason why we cannot have a clear definition is that we cannot make a clear definition about intelligence.
And we don't yet know what consciousness is and what gives rise to consciousness. If we are going to build an AGI, we need to answer these questions. If we don't have the answers, we will have created what we are trying to create partly by accident, or rather without fully understanding it, which means we have to worry about our control over it.
The limits of what is possible; can only be defined by going beyond them, that is, to the limits of the impossible. _ Arthur C. Clarke
Google DeepMind CEO Demis Hassabis defines AGI as human-level cognition. It predicts that we will have this level of AGI by 2033.
So why should we have an AGI? As always, there are different opinions on this subject. OpenAI chief scientist Ilya Sutskever says in a YouTube documentary prepared by The Guardian. "Because AI will solve all the problems we have today. It will solve employment, it will solve disease, it will solve poverty. But it will also create new problems. The fake news problem will be millions of times worse, cyber attacks will get worse, we will have fully automatic AI weapons "I think AI has the potential to create infinitely stable dictators."
And there are people who are extremely concerned about this. And one of those people is Geoffrey Hinton. Geoffrey Hinton is one of the names that paved the way for artificial intelligence with his work on artificial neural networks. And he expressed his concerns publicly.
In conclusion, what do we know or what do scientists know about AGI? What we know is that artificial intelligence is doable. It is possible to create artificial intelligence with various methods such as machine learning, deep learning, natural language processing. This is not a science fiction anymore.
So what's the unknown part? Is it possible to create AGI? Theoretically yes it is possible. As far as we know, there is no law of physics that prevents this from happening. So when can an AGI be done? Maybe tomorrow, maybe in 10 years.
So what will happen? My personal opinion is that if something can be done, people will do it. Ethical concerns or the harm that people will suffer are not taken into account, and a material and moral profit and loss criterion is made between how much benefit will be obtained and how much harm will be suffered for this. And as a result, human behavior is generally about taking risks if it does not pose a problem for itself.
What we create will not be God. This is a very eye-catching title, but no, not God. We will create something, maybe we will do it, but we will not have the most vital information about it, just as we cannot have it ourselves. Perhaps what we have created will tell us the answer. However, the laws of physics will apply to it, and even if it consists only of energy and mind, there are laws that it will be bound by.
AGI is one of the greatest ideas humanity has ever had. And I think we need to be careful not to get pessimistic. And perhaps if we scale up by building more democratic systems while maintaining open structures, good can be created for everyone. On the contrary, in closed structures, only speculation and confusion will come to the fore. This leads to a meaningless treatment of the idea. Both optimists and pessimists go to extremes and miss the main points. What matters is how this will change the direction humanity is heading.
