In a recent Mind Matters podcast, computer engineer Robert J. Marks puts the AI hype into perspective by investigating what it can and cannot do. Marks is of the mind that AI can offer a myriad of benefits to the modern world, and notes that the technology has already made inroads into various spheres of life including banking, accounting, and facial recognition technology. The danger lies in believing that AI can replicate human creativity and understanding. While AI can do a lot, it can never understand itself the way human beings can. Marks thinks this is essential to keep in mind. We are not machines.
Here is a quote from the episode in which Marks sets some terms and definitions:
Artificial intelligence depends on your dictionary. If you go into academia and go into the different disciplines, you’ll find skirmishes as to what AI means. There’s kind of synonyms which are used for AI, such as computational intelligence, machine intelligence, and to those in the field, this means a lot. They do subcategories and things of that sort. I think for purposes of our discussion, we should probably treat it just like the media treats it, which is any gee, whizz thing that can be done by a computer. If a computer does something and it’s just miraculous and you go, “Gee, whizz,” then that’s probably AI and I like it for purposes of the discussion because what it does is it brings AI to the level of the computer, so when we talk about the capabilities of AI, we can talk about the capabilities of computers and if a computer can’t do something, then AI is doesn’t have the ability to do it either.AI: Is it Good or Bad for Society? | Mind Matters