Is Artificial Intelligence Possible?

May 4, 2014 in Science, Technology

illustration of a robotic cat playing with a computer mouse - by artymind.com

Robopuss – artymind.com

AI researchers have been busy building simple gadgets like clumsy robots and chess-playing computers for, what? 60 years or so? But where is the long-promised truly intelligent machine? Basically, nowhere in sight.

And there’s another confusing element to this: consciousness. People tend to conflate the two, but awareness is quite a different thing from intelligence: the ability to compute does not equate to the awareness of computing, or the experience of computing. My view is we may well be able to build semi-intelligent limited-area-of-expertise machines in the not-too-distant future (20-50 years, maybe), but truly autonomous, thinking, feeling, spiritual machines are not so likely. Indeed, they may require some new physics.

Strange and beautiful lady surrounded by wires - by artymind.com

Robotic Lady – artymind.com


By that I mean that at the moment, most of our AI technology is based upon the physics of electromagnetism, with a little quantum mechanics to make some of the circuitry actually work (e.g., Zener diodes and other such tunnelling devices). But does our brain use only these limited natural forces? There are others, and there are quantum properties that practical technology hasn’t even touched yet: quark, strangeness and charm to name but three. Gravitons could be involved but at the time of writing they haven’t even been detected, let alone put to practical use. Dark matter and dark energy could be involved, and more: there is a whole zoo of particles and forces out there that our technology doesn’t even touch yet, so there’s no knowing at this stage whether any of it would have an impact on how consciousness actually works.

Robot with a Heart - by artymind.com

Robot with a Heart – artymind.com

Clearly it has something to do with our brains, but what, exactly, isn’t at all clear. A complex machine mimicing the brain may not be conscious no matter how smart it is. The same goes for neural networks: just because they can copy some functions doesn’t make them conscious. Some philosophers, like Daniel Dennett, might argue that we aren’t really conscious either, although I’m not sure what sense that makes, since we clearly are something… or I am, anyway! I can’t speak for you, of course, but I at a minimum have an illusion of being aware of what I’m doing and of being autonomous in some ill-defined way.

Here are two sets of videos about the whole idea – for, and against. They have their own arguments, different from mine (above), so see what you think! The first set is by Rob Ager, who thinks AI is not going to happen any time soon.

Rob Ager’s Androids & Artificial Intelligence: A Modern Myth [Playlist, 1h 21m]

Rob Ager’s video channel can be found on YouTube.

Next we have a film suggesting, to the contrary, that we might get some pretty capable machines, very soon indeed.

In Its Image [Playlist, 30m]

The website for the above project is initsimage.org.

So, what do you think? Intelligent machines? Conscious machines? If not, why not, and if so, how quickly?