We seem to be experiencing an AI revolution, spurring advancement in many fields.
- Please explain, at a high level, the uses of artificial intelligence and why we might need it.
Artificial Intelligence Stack Exchange is a question and answer site for people interested in conceptual questions about life and challenges in a world where "cognitive" functions can be mimicked in purely digital environment. It only takes a minute to sign up.Sign up to join this community
The question, "Why do we need artificial intelligence?" is quite to the point. Technically, the answer is, "No reason." If we needed artificial intelligence, then we would have become extinct over the last 50,000 or more years we've been a species with human intelligence, so we want it.
We do want it, and there are benefits. Some claim there are risks too, which is logical. We should be open to all views. Some of them may be spot on, and others may be without merit.
One list of benefits already seen might include these, with most beneficial on top.
These are further potential benefits that may also emerge.
These are the risks from highest to lowest.
We don't really need artificial intelligence, but it is proving ever more useful. This is a function of what is known as utility--the capability of an algorithm to perform a task adequately. The new utility of AI has upsides and downsides.
Weak AI: Less capable than a human
At the lower end of the scale, there might be situation where a human would be better, but the work is so dangerous, or expensive for humans, that we use automatons instead. (Space exploration is a good example. The AI on a deep space probe or Mars rover.)
Semi-Strong AI: About as capable as a human
I'm defining "semi-strong" for this answer. Here we have AI or automation that can do tasks as well as humans, such as on an assembly line, but where automation is more efficient. (This type of automation spurred the Luddite movement, responding to the loss of human jobs due to automation.)
As Machine Learning continues to get more effective, the range of tasks that AI can perform as well as humans will surely grow. This might lead to unprecedented levels of persistent human unemployment, but some argue that automation will also create new opportunities for humans and mostly eliminate repetitive, less fulfilling work. (See also: Technological unemployment)
Strong AI: Exceeds human capability
Strong AI is the "holy grail", and some believe it will lead to a technological "singularity" in which smarter machines make ever smarter machines. (No one knows:) Nevertheless, Machine Learning has demonstrated greater than human capability in a number of tasks, and the range of such tasks will surely grow. (AlphaGo was a milestone because the game of Go is unsolvable and notoriously difficult for AI, prior to AlphaGo. Now it's unclear if unmodified humans will ever again be able to beat strong AI at these types of games.)
Although game AIs are not directly useful, except for recreational purposes, the methods used to create them can be extended to real-world problems. There are many forms of Machine Learning, not restricted to Neural Networks and MCTS, with Evolutionary Algorithms as another type, also recently demonstrating strong utility.
Strong AI is useful because it has greater utility than humans. It is desirable because it increases efficiency and expected return on investment.