What is Artificial General Intelligence?
Artificial general intelligence refers to artificial intelligence that has the ability to understand the world just as well as a human and the capability to perform a wide range of tasks at the same level as a human can.
Artificial general intelligence is also known as strong AI or deep AI. It refers to the ability of AI to think, understand, and solve a variety of problems just like humans do.
What are the three types of Artificial General Intelligence?
Artificial intelligence is classified by comparing its abilities to that of a human. Here are the types of artificial intelligence.
- Artificial narrow intelligence (ANI) has a narrow range of abilities. It is subpar to human intelligence.
- Artificial general intelligence (AGI) is as capable as human intelligence. They are on the same level.
- Artificial super intelligence (ASI) refers to artificial intelligence that has greater capabilities than a human.
At this point, we cannot say for certain what artificial general intelligence or artificial super intelligence will be like. All the artificial intelligence that we have today is still artificial narrow intelligence. That means that we have no idea whether actual artificial general intelligence will be anything like the artificial general intelligence represented in science fiction.
What is the difference between artificial intelligence and artificial general intelligence?
The artificial intelligence that we use right now is nowhere close to artificial general intelligence. It’s far more advanced and sophisticated than anything we’ve ever had before, but it’s far from artificial general intelligence.
Today’s artificial intelligence can perform some tasks extremely well, but the operative word is ‘some’. Its range of abilities is narrow and focused. It can perform some tasks much quicker than humans, but again, it is limited to only a few very focused tasks.
To illustrate this, the K computer, built by Fujitsu is one of the fastest supercomputers in the world, in an attempt to reach artificial general intelligence. Even this computer took 40 minutes to simulate just one second of neural activity.
Artificial general intelligence involves AI that can think like a human and perform a wide range of tasks just like humans. It is far wider in its scope than the current artificial intelligence is.
Is artificial general intelligence possible?
Many leading researchers have conflicting opinions on this. Some have different thoughts on when we will be able to achieve it. Others argue that we may never achieve it.
Some say that we could achieve it in the next decade itself. Some researchers say that it would take at least a century.
The head of the University of Oxford’s computer science department, Michael Woolridge said, "neither I nor anyone else would know how to measure progress" towards AGI.
On the other hand, Google’s director of engineering, Ray Kurzweil, says that by 2029 there will be AGI that could even pass the Turing Test. A lot of researchers claim that human intelligence is fixed and limited, while artificial intelligence is continuously growing, improving, and expanding, which is, after all, the core concept of machine learning, an area of artificial intelligence in which the technology learns and becomes more intelligent as it is exposed to data and concepts and thus learns patterns.
Matthew O’Brien of the Georgia Institute of Technology stated that “ The long-sought goal of a ‘general AI’ is not on the horizon. We simply do not know how to make a general adaptable intelligence, and it’s unclear how much more progress is needed to get to that point.”
There are other people that don’t deny the possibility of artificial general intelligence but believe that it will be totally different from human intelligence. They believe that it will not be superior to human intelligence. According to them, similar to animal intelligence, artificial intelligence will have separate abilities. Those with this belief also assert that artificial general intelligence is not something that should be feared; rather, it would lead to much higher levels of product innovation since machines would be able to solve complicated problems that humans would not be able to handle. They maintain that there are already early implications of this phenomenon in areas like healthcare where artificial intelligence is used as a diagnostic tool and in other ways as well.
There is yet another perspective according to which artificial general intelligence is possible, but it won’t be realized until the very distant future. A lot of the experts who subscribe to this opinion believe that while artificial intelligence could become a reality at some point very far off in the future, it is not even remotely close to becoming a reality in our lifetimes.
Rodney Brooks, an MIT roboticist and co-founder of iRobot does not believe that artificial general intelligence could be achieved until at least the year 2300.