I remember when a friend repaired our family van and painted the repaired area he explained to me about the ‘depth’ of the painted surface. If you could hold a ruler up to the side of the car and read the ruler in the reflection, the number of inches that you could read indicated the depth of the painted area. Of course, the greater the depth, the better the paint job was.
If you can visualize a large mirror, you can realize that a good mirror has considerable depth, perhaps only measurable in many feet or meters. Artificial Intelligence is like that. There’s the human language level, where the language is parsed and given some meaning ‘on a computer level’, there’s the machine-level languages and the algorithms that drive them and allow them to take advantage of processing power, and then there’s the machines themselves, which have to have a model that not only reflects the possibilities of what the software is capable of and wants to do, but reflects what the various clients are capable of or will be capable of and will be headed towards being capable of.
Those clients express themselves through various shapes and forms and interfaces, and none of those are negligible. Some of them are robotic, some are, 2-dimensional interactive, some are three-D interactive, some are remote devices, some even have biological characteristics, and some haven’t quite taken shape yet.
There is a great depth to the many levels of this topic, not the least of which is the ethical issues that pervade and surround all that artificial intelligence can come to mean and be.