Learn something new every day More Info... by email
While there are many different Artificial Intelligence (AI) techniques that have been developed, with new methods being created, a few forms of artificial intelligence have become increasingly popular. Some of the most common techniques include the use of neural networks and the development of expert systems. These different artificial intelligence techniques can be used to develop different forms of AI, usually based on the amount of “thinking” the program can actually do, and these are known as either “strong AI” or “weak AI.”
AI techniques are methods that can be used to develop and create computer programs commonly viewed as forms of artificial intelligence. In general, artificial intelligence refers to a program that is able to mimic or re-create the thought processes demonstrated by the human brain. This usually involves solving problems, making observations or receiving input for use in analysis or problem solving, and the ability to categorize and identify different objects and the properties of those objects.
There are many different artificial intelligence techniques that can be utilized by an AI programmer, though two of the most common are neural networks and expert systems. Neural networks are computer programs designed around the cognitive processes used by the human brain. Essentially, a neural network consists of layers of categorization and methods by which objects can be identified and categorized. This is similar to the idea of schema in human cognition, which allows people to identify objects based on properties of those objects. New information presented to the neural network can then be analyzed and identified based on previously inputted criteria, allowing the system to “learn” new categories and identify known or unknown objects.
Expert systems are AI techniques built around logic and “if/then” statements. This usually involves a great deal of information that is “taught” to the computer system, which then makes the system an expert in a particular field. When new input is introduced, such as a request for processing financial reports, the expert system can analyze the information using these if/then statements to limit the output response.
These various artificial intelligence techniques can be used to develop systems that are considered either “strong AI” or “weak AI.” Strong AI systems are those that most fully seek to emulate human thought and cognitive capabilities through a wide range of functions. These systems can analyze new information and provide output that potentially goes beyond the limitations of input data. Artificial intelligence techniques that develop weak AI systems are narrower in focus, and seek to replicate only a single function or aspect of human intelligence.
@browncoat - I don't know, I think that people are able to respond to machines even without overt displays of emotion. Even now, there are all kinds of robots in Japan especially which people tend to get attached to, and they are designed for that purpose.
And isn't emotion just another aspect of intelligence? Or at least something else that could be programmed into an advanced machine? I mean, what is fear except wanting to be away from something that we don't like.
Love might be a bit more difficult to quantify, but I'm sure it could be done.
Artificial intelligence software can already program "fear" into a machine. No, it's not going to have sweaty hands and a pounding heart, but if you took all the physical things away from fear, what you would be left with is an aversion to doing something and that can easily be put into a machine.
@pleonasm - There are quite a few stories that deal with the ethics of creating artificial intelligence. The film that is actually called "AI" for example.
Even many of the stories where the AI goes rouge and tries to kill all the humans are generally touching on the fact that we might just be playing god by creating such a being.
I think, though, if it is possible to create an AI that we will do it eventually.
Hopefully with the help of all the speculation there has been over it and with generations who have been raised with media that asks ethical questions like this, we will end up treating AI like children rather than like slaves.
For now there
are some computers that can do some wonderful things being developed in artificial intelligence programs around the world, but I wouldn't call them true AIs, not in the science fiction sense.
But then, will we ever really think of it as an AI until it can be emotional? It's love and fear that we really identify with when it comes to judging "humanity". Until it can show those, it will just be seen as a collection of glorified calculators.
I have always liked the idea of artificial intelligence and particularly like it when it is treated in a good way in science fiction.
Usually the standby is to make the AI "evil" or at least to make it angry at humans or convinced it has to destroy them in order to make the world safe. In those cases, the story is more about the humans and the AI is just a convenient way to shine a light on them.
But there are some science fiction stories which focus on more interesting questions, like what is life, what is intelligence and so forth.
In those the AI is alien but it's also vulnerable. It's not just a massive conglomeration of
machines that wants to destroy humanity.
I know we are nowhere near either scenario at the moment when it comes to AI. They have managed to create machines that can learn though and I think that's a crucial step. Whenever I check into that project that was getting computers to talk to each other over the internet and get people to teach them language, I get chills. It's an exciting time to be alive.
One of our editors will review your suggestion and make changes if warranted. Note that depending on the number of suggestions we receive, this can take anywhere from a few hours to a few days. Thank you for helping to improve wiseGEEK!