Robots and machine intelligence.

So I went to a lecture on Friday by Dr. Roger Brockett, hosted by CIM and the Faculty of Science at McGill, entitled “What is an Intelligent Machine?” Dr. Brockett is the founder of the Harvard Robotics Lab, and an important figure in the field of robotics.

I really liked the lecture. Dr. Brockett talked a bit about the history of intelligence as a construct, both unitary and multiple, and I agree with his position that intelligence is more a collection of several (potentially over-lapping) abilities rather than a single underlying trait.

He talked about the Turing Test, and suggested that the test is not well-suited for non-linguistic intelligence. He pointed to wood-peckers, foxes, and other animals who display intelligence in specific domains at least equal to that of humans, and stated that we need to develop tests for robots that are more domain-general. In particular, he suggested tests which involve manipulating objects and solving problems in the environment.

I mostly agreed with him during his presentation, though I think neurons are less binary than he claims; despite the ‘all-or-nothing’ nature of the action potential, the summing of EPSPs and IPSPs seems analog to me. He talked quite a bit about how we have used analog and digital mechanics and math separately, and about how any modern approach to robot intelligence will have to utilize both modes.

But during the question period, we parted ways. Bearing in mind that Dr. Brockett is rather more educated in these matters than I, there were some points that I felt he is mistaken on.

1. Learning: Dr. Brockett was asked about the role learning would play in designing intelligent robots who would be able to problem-solve in their environment. I was quite surprised when he said there wasn’t one, and seemed to suggest that robots using his model would have all their required information to begin with.

This is a bit bizarre. He spent some time talking about how complicated even a small thing like shaking hands can be, but seems to think that this operation can be programmed into an intelligence. It would be much easier, to my mind, to create some simpler rules and then to allow the robot to explore its environment and to learn dynamic movement rules.

2. Neural networks. I think I’ve talked a bit about these before, and I think connectionist models are pretty awesome. But Dr. Brockett looks to be on the same side as Fodor, and argued that while neural networks are fine and good, especially for learning, what a complex robot intelligence need will require a more computationalist approach using formal symbol manipulation. I disagree entirely. I don’t think a computationalist, non-learning robot will be able to display the flexibility of behaviour that Dr. Brockett claims is the hallmark of intelligence.

3. Embodiment. Although Dr. Brockett didn’t use this term precisely, he spoke frequently about the need for robots to be able to move around and do things. Here I’m a bit torn. I don’t see why a robotic intelligence could not inhabit a virtual environment, or could not have sensory input/output without motor input/output.

That said, I’m a big fan of the idea that intelligence and particularly consciousness requires a certain degree of embodiement; but this is because I believe these things require learning in a changing environment, so I don’t quite understand Dr. Brockett’s focus on this if he feels there’s no need for learning in robot intelligence.

The last issue for me didn’t really come up explicitly, but I figured I’d talk about it anyways.

One of the reasons I focus on learning to much is it allows for more evolutionary strategies. We could design a quadrapedal robot, program all the complex business of walking into it, and then send it into an environment to move about.

But wouldn’t it be better if we created a virtual environment, allowed some random physical mutation, and made it so that success on these non-verbal intelligence tests Dr. Brockett was discussing caused the successful design to pass on its most recent configuration? Then we’d run the same tests again, and again, until we came out with a robot with physical and ‘mental’ adaptations well-suited to the virtual environment. We could then re-create these physically and test them again, using a learning neural network for the ‘brain’. We’d end up with something that worked, probably efficiently, and it would not necessarily be the shape we expected. It wouldn’t ‘think’ the way we expected.

This is a great way to approach robotics and intelligence, I think. Because of the random element involved in evolutionary processes, we do not have to pre-suppose a great deal beyond the starting conditions, and this allows for solutions that we would never have expected. Sure, you could spend ages designing a finger on a hand, but evolutionary robotics might come up with a suction appendage, or a flat sticky flexible ribbon, or who knows what else?

Admittedly, I’m overstating the case a bit. There are certainly strong limitations in my proposed methods, but I think they would prove out in the end to require less brute-force work and generate more interesting findings than the static, computationalist approach suggested in the lecture.

I should mention that Dr. Brockett took some shots at people who had concerns regarding quantum mechanics, free will, consciousness, etc. I generally appreciated this. While I obviously have a certain amount of interest in those topics, it was nice to see a pretty down-to-earth lecture on intelligence in machines.

Vision and perceptual errors

It’s hard to get everything down that I want on a given topic, so I often have trouble getting posts up when I have a lot to say.

So, while I try to pare down or split up a lengthy post on mirror neurons, here’s a clip from Neil Degrasse Tyson. He’s sort of a latter-day Carl Sagan – he lacks some of Sagan’s poetry, but he’s a bit less embarrassingly 70s.

Now, the bit I wanted to focus on was the problems we experience with perception. Eyewitness testimony is, as he claims, unreliable in the extreme. He invokes psychologists as experts without really giving any examples, since it’s a talk, so I figured I’d present two examples of perception problems.

This has been inspired by a yet another book I’m reading on consciousness, so if I can pull together the energy expect a few posts on that too.

Follow the instructions in the video:

About 50% of people fall into that, in what’s known as inattentional blindness.

Or this one?

Also pretty fun – that’s change blindness.

The point of these isn’t that we can’t actually see the changes, it’s that we aren’t aware of the changes. More and more vision seems to be a reconstruction using novel features, and we just fill in the details in a fuzzy sort of way. The same seems to be true of memory – there’s some interesting things specifically with memory, but the two are not unconnected.

Vision, like all sense, may rely on a sort of memory buffer, whereby the hold incoming data in storage while it’s processed. Since memory seems to hold a few key features and then just fills the rest in using logic and expectations based on prior experience, this means all our senses may well do this – making us prone to error when it comes to sensing our environment.

Don’t get me wrong – we’re good at doing what we have evolved to do. We can notice things, we can walk about, eat, reproduce, etc. We wouldn’t be around if we weren’t pretty successful organisms. But when it comes to more precise or more complex cognition, we’re prone to all sorts of perceptual and conceptual errors.

Good to keep in mind.

Human demarcation

I’m wrestling with a post on Popper and manifest truth, but in the meantime, some thoughts on drawing species lines.

What came first, the chicken or the egg? I read an answer somewhere which was quite good – ‘a slightly different egg.’ The first hen, you see, would not have been recognized as such. This is because you cannot point to a single animal and say ‘See? That’s the first chicken.’ Individuals do not evolve, and they do not found whole species. Eve is a myth. (The case of the modern domestic chicken (Gallus gallus domesticus) is a bit confused, as it might have come about due to deliberate cross-breeding between related species, so let’s just say ‘chicken’ to denote a generic bird of a definite species.)

In fact, new species are almost entirely noted in retrospect, when you can say ‘This chicken descended from this chicken-like ancestor.’ The ancestor is not a chicken, and likely cannot breed with modern chickens (an sufficient but not necessary aspect of a distinct species). Many proto-chickens likely bore young which were very similar to modern chickens, many of which bore young which were even more similar to modern chickens, until eventually without realizing it you have ancestral chickens and modern chickens, and aren’t quite sure when in the last hundred/thousand/more years it happened.

This is entirely the same with humans.

An amusing pastime is watching ‘creation scientists’ try to demarcate archaeological finds of hominid (Hominidae refers to all of the great apes, including Homo sapiens) as either ape or human. By ignoring the various gradations, and the fact that any shift from ‘more ape’ to ‘more human’ is obvious only because the fragmentary nature of the fossil record gives us snapshots rather than a continuous view of hominid evolution. This game can be played all across the internet (and in real life), but some good examples of it in play are here, here, and here.

What is human, and what is not, are not clear distinctions. At the risk of digging myself a deeper hole that I can write myself out of in the next year or so, I plan to return to this concept, that in science while there are categories they are non-rigid categories. Something similar applies to sex and gender, but that’s whole other kettle of fish.

But – since this was on my mind today – how do you draw the distinction between human and cyborg? I know this might seem something of a non-sequitor, but it’s something of great interest to me. I and many of my friends read sci-fi, and have played games, in which the characters are part-human and part-machine. These media typically assume that as you add artificial limbs or functions you become less human.

Are we less human now? Because much of the speculation in the books revolves around things we already possess and do. Artificial eyes? (We have them). Are they somehow different than looking through a camera with your own eyes, aside from considerations of rejection or infection? Does wearing night-vision goggles make you a cyborg? How about surgically implanted memory storage? It’s no different than storing information on a computer, which becomes an accessory to your memory function. Computers in fact make much of this debate possible, as for many people in developed countries they are a daily necessity for their jobs. If I lose a tooth and have it replaced by a plastic tooth, am I a cyborg? What if I receive a hip replacement? An artificial leg?

At what point am I no longer human? The answer, I think, is so long as my brain is intact I will remain human, for all intents and purposes – or at least enough of one to pass a Turing test.