1118: Thwee beginnings of Artificial Intelligence ...

 With ELIZA we probably got the first example of a computer behaving like a human being, showing intelligence. It was 1966.

It is the real start of Artificial Intelligence. The basic structure of the AI philosophy in those days was: you have objects with properties and you have a set of rules and you combine it all in a computer program.

  

You have chess pieces and their properties. You have the rules of the game. You have specific strategy rules. Put it together and you have a chess-playing computer.

   

This approach was closely related to the development of predicate logic. It got the name Symbolic Artificial Intelligence.

  

In artificial intelligence, symbolic artificial intelligence is the term for the collection of all methods in artificial intelligence research that are based on high-level symbolic representations of problems, logic, and search.

  

Symbolic AI was the dominant paradigm of AI research from the mid-1950s until the mid-1990s. Researchers in the 1960s and the 1970s were convinced 

     

that symbolic approaches would eventually succeed in creating a machine with artificial general intelligence and considered this the ultimate goal of their field.

   

There were high expectations of AI in those days, but it didn't fulfill the expectations, which led to a kind of AI winter.

  

A second boom (1969–1986) occurred with the rise of expert systems, their promise of capturing corporate expertise, and an enthusiastic corporate embrace.

   

I remember that period well. I spent hours on learning PROLOG, THE programming language of expert systems. I was involved in a project that had as its goal to develop a kind of database and expert system for design purposes.

  

But again, high expectations that could not be realized. Another, second, AI Winter (1988–2011) followed. I was there, you know.

    

The philosophical quintessence here is, how AI scientists thought about how the human brain works. They thought it was a system of symbols and algorithms to manipulate the symbols.

  

So, the goal became to develop a system of symbols to represent how we think. This was closely related to the idea that it is just a matter of the right software. The hardware could be our brain or a computer.

   

And here we are back in the 17th century with Leibniz. He also hoped to capture all human knowledge in symbols and the rules that come with this then could help us to find new answers.

   

The big question here is: is intelligence really just a system of symbols, rules, and the right algorithms to manage the symbols?

   

This question opens a new debate in the Artificial Intelligence world: the debate about strong versus weak Artificial Intelligence.

   

Strong AI, like a smart character from a sci-fi movie, like HALL for instance, could think, learn, and perform tasks just like humans. 

    

On the other hand, Weak AI, the kind we encounter daily, focuses on doing one job well, such as playing chess or Go, giving us directions, or even helping us pick the perfect playlist. 

   

The current situation is that we find weak AI in many areas, where they show sometimes really biased behavior. Strong AI is still a theoretical issue: it really takes a lot to get a real mind in the machine. It is not yet done except in SF movies.

    

Thank you for your attention again....