I love the Very Short Introduction book series. Some are worse, some are better : but they often provide a concise overview of a complex subject and an in-depth reading list for those who want to learn more.
Boden’s Artificial Intelligence is a particularly good example of A Very Short Introduction. Margaret A. Boden is a recognized expert in the field of artificial intelligence. I’ve read a few things on the subject and her name keeps coming up again and again. When approaching this book, I was a little concerned that it would be too advanced for me. I needn’t have worried. Boden writes lucidly and logically about a very complex and controversial subject matter. Rather than writing a history of AI, Boden divides her book according to various research areas in the field, including:
- Language processing
- Artificial neural networks (ANNs)
- Robots and artificial life (known as a-life)
A particularly interesting area that I had not read much about before is robotics. Instead of necessarily aiming to imitate the human brain, many scientists are basing their robot ideas on insects. This is because insects often act when responding reflexes rather than acting as a deliberation. For example, female crickets recognise only one speed and frequency of cricket song – that of their own species of cricket. Whenever they hear this cricket song, they are neurally wired to move in a certain direction. A biological automaton, if you’d like. I’m sure Descartes would love this…
Boden provides a remarkably balanced view on the problem of consciousness and underlines the controversial nature of the statements made so far. She also gives some space to the view that real consciousness is impossible without metabolism – and therefore “conscious” AI would necessarily mean AI that is “alive” in some meaningful, possibly biological way.
But what I really appreciated in Boden’s view is the space she offered to AI ethics regarding applications of AI that many people choose not to question. She talks, for example, about the robots which are created to behave as companions for the elderly. What does it mean to live in a society that thinks it is okay to offer such robot companionship to the elderly rather than human care?
“Whether it is ethical to offer such quasi-companionship to emotionally needy people is questionable. Certainly some human computer-interactive systems appear to provide pleasure, and even lasting contentment to people whose lives seem otherwise empty. But is that enough?”
That is not to say that robots can’t do humans any good – just look at this robot fluffy seal that helps dementia sufferers…
But it is clear that they must be use as an enhancement to normal human care and not as a simple substitute for it.