AI Fundamentals Series: Understanding 'Intelligence' in AI #12
An introduction for Indie Developers
Issue #12 of the
Infinite Waves
Newsletter
It is worth, I think, considering what the ‘intelligence’ part of ‘Artificial Intelligence’ is, or means. In ‘AI Fundamentals Series: a basic Artificial Intelligence definition #2’, the first issue in this series, I laid out some basic definitions of AI, but this current issue is a bit of a different idea.
It is the approach of what we are doing when trying to build something ‘intelligent’.
What is intelligence?
In his book ‘Life 3.0’, the physicist Max Tegmark offers a broad definition of intelligence as the “ability to accomplish complex goals”.
This seems like a pretty good way to frame it, although his book centres more around Artificial General Intelligence (AGI), so we could even do away with the ‘complex’, as some simple AI systems, in comparison, might be doing tasks that could be defined as basic. As an example think of an AI system that just plays tic-tac-toe, compared to one that can play StarCraft II.
I’m thinking about this in two ways of why it might be important or interesting to consider.
The first is that it could help when designing and building AI systems of your own. Having this concept in mind, of creating something intelligent to ‘accomplish a goal’.
The second is what it says about intelligence itself. Holding up a mirror to human intelligence. The goal of AI research and practise is to build intelligent entities and systems, but to also understand them.
An example of this questioning is put forward in Peter Norvig and Stuart J. Russell’s book, Artificial Intelligence: A Modern Approach:
“AI addresses one of the ultimate puzzles. How is it possible for a slow, tiny brain, whether biological or electronic, to perceive, understand, predict, and manipulate a world far larger and more complicated than itself?”
This has been the occupation of humans for, at least, thousands of years. Philosophers in Ancient Greece considered what intelligence is, what it means, and what we can do with it. This was the foundation of what we know today to be science.
They go one step further recounting the philosopher Hubert Dreyfus:
"The story of artificial intelligence might well begin around 450 B.C." when Plato reported a dialogue in which Socrates asks Euthyphro, "I want to know what is characteristic of piety which makes all actions pious... that I may have it to turn to, and to use as a standard whereby to judge your actions and those of other men." In other words, Socrates was asking for an algorithm to distinguish piety from non-piety.
Steady progress made by philosophers and mathematicians working on domains such as logic (George Boole (1815-1864), for example, opening the door with his work on logical inference, which contributed to the functioning of modern computers).
This progress has culminated to a point at which we are now creating what can be deemed as ‘intelligent’ systems. Systems that, at least in part, can emulate or echo the sort of intelligence humans display.
Thinking about AI as an Indie Developer
This then provides a good platform as Indie Developers to start thinking about what’s going on under the hood of AI systems and how we can make them ourselves. Thinking of them as ‘intelligent’ entities, not just as another piece of software, but as something that can achieve a goal independent of direct human intellect and input, but obviously still indirectly derived from it.
Sources: