different experiences and contexts can cause vastly different features to be emphasized. And thus, for example, the fact that we can readily recognize pictures of animals in cave paintings made by Stone Age humans depends greatly on the fact that our visual system still picks out the same specific features.

But what about more abstract art?

Although one has the feeling that this involves more human input, it rapidly becomes extremely difficult to tell what has been created on purpose. And so, for example, if one sees a splash of paint it is almost impossible to know without detailed cultural background and context whether it is intended to be purposeful art.

So what does all this mean about extraterrestrial intelligence?

My main conclusion is rather similar to my conclusion about artificial intelligence in Chapter 10: that the basic issue is not finding systems that perform sophisticated enough computations, but rather finding ones whose details happen to be similar enough to us as humans that we recognize what they do as showing intelligence.

And there is perhaps some analogy to recognizing the capability for sophisticated computation in the first place. For while this is undoubtedly very common say in cellular automata, the most immediate suggestions of it are in class 4 systems like rule 110 that in effect happen to do their computations in a way that looks at least somewhat similar to the way we as humans are used to doing them.

So should we expect that somehow recognizable extraterrestrial intelligence will occur at a level of a few percent—like class 4 systems?

There is clearly more to the phenomenon of intelligence than this. But if we require something that follows too many of the details of us as humans there is already evidence that it does not exist. For if such intelligence had ever arisen in the past, then extrapolating from our own history we would expect that some of it would long ago have colonized our galaxy—at least with signals, if not with physical objects.

But I suspect that if we generalize even quite modestly our definition of intelligence then there will be examples that can be found—at least with sufficiently powerful methods of perception and analysis. Yet it seems likely that they will behave in some ways that are