2 Natural intelligence
Before we go on, we should perhaps remind ourselves of what ‘intelligence’ is.
Write down a brief summary of the key characteristics of ‘intelligence’ as the concept is understood in conventional Al work.
Symbolic Al is based on a certain conception of human intelligence, on a general understanding of the workings of our own minds. As sentient beings, we know that we bring special abilities to the world: we use language and logic to plan, communicate and carry out complex tasks; we reason about the present and the future; and we respond flexibly to new situations or unexpected developments.
To sum this up, these are some of the presumed basic characteristics of human ‘intelligence’ in the broadest sense, characteristics that conventional Al seeks to replicate on machines:
- rational thought (using symbols to reason with, as in logic and mathematics)
- language (using symbols to communicate in speech and writing)
- the ability to make plans, design and foresee
- the ability to learn specialised knowledge and skills.
But although we might not be inclined to call ants or geese ‘intelligent’ in any of the above senses, something more than simple randomness is clearly taking place in the behaviour described in the case studies. So perhaps it may be necessary to try for a rather more inclusive understanding of the idea of intelligence.
To what extent do you think a chimpanzee could be called ‘intelligent’? What about a dog? An insect? Do you think there could be any alternative conceptions of ‘intelligence’, or is human intelligence the only sort possible?
Chimpanzees might not seem to be capable of the kind of symbol manipulation that we think of as typical of the human mind. However, the evidence that they possess an intelligence in many ways like our own is compelling. They live in complex social groups in which status, rank, sharing, cooperation, deception and manipulation all play a part. They make and use tools, can solve complex problems and can be taught a grasp of certain limited forms of language. It’s hard to deny that they have sophisticated minds. Dogs, as we know, can learn complex tasks, display emotions, perform feats of recognition and problem solving, and learn to respond to language – of a simple kind. But we would probably baulk at the idea that dogs’ minds function logically or that they contain symbolically encoded knowledge. As you’ve seen, insects such as ants display highly organised mass behaviour, and some insects even seem to have basic learning capacities. But no one in their right mind would be inclined to call an individual insect – with its tiny, almost non-existent brain – intelligent, or to suggest it uses logic or symbols.
Clearly, the issue is a complicated one. On the one hand, few if any creatures seem to possess anything resembling the human capacity for rationality, language and knowledge. On the other hand, we see all around the natural world feats of problem solving, construction and organisation that are often startling. And the idea that humanity stands completely apart from the rest of creation, is unique and unrivalled in its status and abilities, does not sit well with the twentieth-century mind. When Shakespeare wrote in 1603:
What a piece of work is a man! How noble in reason! how infinite in faculty! in form, in moving, how express and admirable! in action how like an angel! in apprehension how like a god! the beauty of the world! the paragon of animals!
he was expressing the standard view of the time. This belief was expressed with blunt clarity in the words of the fifth-century church father, St Augustine:
if, when we say, Thou shalt not kill, we do not understand this of the plants, since they have no sensation, nor of the irrational animals that fly, swim, walk, or creep, since they are dissociated from us by their want of reason, and are therefore by the just appointment of the Creator subjected to us to kill or keep alive for our own uses ...
I think few of us would be comfortable with this now. Since Darwin, we are much more inclined to see close links between ourselves and the rest of the natural world, to view our abilities and those of other creatures as all part of a continuum. And in directly comparing our minds to those of animals (usually to the animals’ detriment), we may be making a deeper mistake. Many ethologists [Ethology is the comparative study of the behaviour of creatures, including humans, living in their natural environment.] argue that such comparison leads inevitably to anthropomorphism – seeing intelligence only in behaviour that resembles our own, dismissing everything else as ‘instinct’ or some such word. But the truth is that every species is specially adapted to the specific problems that it confronts. The abilities of, say, a pigeon and a lion are simply not comparable: they inhabit quite different mental worlds. Ludwig Wittgenstein once wrote, cryptically, ‘If a lion could speak, we would not understand him’. But you can see what he meant. How could we possibly understand a lion’s mind? Its purposes and ours barely intersect at all.
Perhaps the best way to investigate this further is to consider in detail some very broad qualities that we might associate with intelligent behaviour. In my answer to Exercise 1, I suggested that the case studies demonstrated behaviour that is purposeful and systematic, and leads to ordered results. In the next three subsections, I will look a little more deeply at each of these qualities and consider what they might tell us about intelligent behaviour.