On the inevitability of the evolution of intelligence
Published: August 27, 2010Tags: cogsci evoution intelligence seti transhumanism evopsych
A few months ago I happened to catch an interesting documentary on television. I don't remember what it was called, but it was narrated by Neil deGrasse Tyson and the focus of the show as the question: if the universe is so chock-full of intelligent life, as most scientists believe it ought to be, how come we have completely failed to detect any evidence of it, despite 25 years or so (the SETI Institute was started in 1984) of concerted effort to do so? (this situation sometimes referred to as the Fermi Paradox, although I don't recall the show using that term)
The show was largely structured around the well known Drake equation, which tries to estimate the number of intelligent civilisations within the Milky Way galaxy, other than our own, which we should in principle be able to make contact with. It does this by multiplying together estimates of a bunch of relevant terms, namely:
- the average rate of star formation per year in our galaxy,
- the fraction of those stars that have planets,
- the average number of planets that can potentially support life per star that has planets,
- the fraction of the above that actually go on to develop life at some point,
- the fraction of the above that actually go on to develop intelligent life,
- the fraction of civilizations that develop a technology that releases detectable signs of their existence into space,
- the length of time such civilizations release detectable signals into space.
The show's discussion of this term was mostly centred around the fact that for intelligence to evolve from unintelligent life takes quite a lot of time, and this time may not always be available. All kinds of events, ranging from asteroid impacts to strong tectonic activity, can very easily completely or almost completely wipe life off a planet (Earth itself has had 5 major extinction events so far, and some would argue, not unconvincingly, that it is currently going through a 6th, in the form of humans wiping out species at an alarming rate), and if these average duration between these events is shorter than the average time it takes unintelligent life to evolve intelligence, then that suggests that the jump from life to intelligent life will be very rare indeed.
There's nothing wrong with the above analysis, of course: sufficiently frequent extinction events are both a reality (I recently finished reading reading Bill Bryson's "A Brief History of Nearly Everything", which was quite an eye-opener on just how inhospitable Earth is to life on long enough timescales) and a real obstacle to the evolution of intelligence. But underlying all of the discussion on the show seemed to be an implicit assumption that this was all that was standing in the way: that if there was some particularly lucky life-bearing planet out there which was somehow shielded from asteroids and solar flares and supernovas, and had relatively stable, benign weather and tectonic activity, and basically was left completely unmolested by forces of great destruction, then it would be a matter of certainty that, eventually, intelligence would evolve. To be fair, I don't know if the producers of the show or Neil himself believe this, but certainly the show did nothing to explicitly dismiss this notion.
The problem with this is that it's completely wrong, and yet it is surprisingly often overlooked. I wouldn't have noticed this oversight myself if I hadn't previously read either Steven Pinker's "The Blank Slate" or his "How the Mind Works" (I forget which it was), which talks about this misconception in considerable detail (I can't remember whether or not it was in the context of the Drake equation). Although it's easy to fall into the trap of thinking that it is, evolution is of course absolutely not some kind of driven, inevitable progression from simple to complex organisms. It's about adaption to an environment so as to maximise reproductive success, and unless there is an unintelligent organism somewhere in an environment with the following conditions satisfied:
- intelligence is evolutionarily accessible to the organism (i.e. it already has necessary prerequisites, like a sufficiently complex nervous system, for a few mutations to lead to some kind of intelligence),
- evolving intelligence will give it the organism a significant reproductive advantage over its unintelligent companions,
- and there are no other evolutionary pathways open to the organism which will yield a better ratio of reproductive advantage to "cost" (in terms of energy requirements, etc.) than intelligence,
The problem that this situation poses for accurately estimating the 5th term of the Drake equation is that we actually have no idea why humans evolved intelligence. There are plenty of plausible hypotheses out there, but nothing for certain, and I don't think there is likely to be anything certain in the near future, given that we know extremely little about the lives of early humans and their ancestors (something else, incidentally, which Bill Bryson's book gives a good accessible account of) and that we really know extremely little about intelligence (to the extent that there isn't even a universally agreed upon, objective definition of what intelligence even is). If we have no idea how we became intelligent, we're not really in a position to speculate reliably about how likely other organisms are to become intelligent, given the chance. The 5th term of the Drake equation could, in fact, be arbitrarily close to zero: close enough to zero to completely counteract all the terms in the equation which are very probably quite large.
Of course, it's by no means a new criticism of the usefulness of the Drake equation to point out that the uncertainty surrounding our best estimates of each of its terms is so great that the final answer can vary by orders of magnitude, and even reach zero. However, as far as I know, the term relating to the likelihood of the evolution of intelligence is the only one which currently has no reasonable lower bound: you can push it as close to zero as you like and not really reach a point where you can compellingly say "come on, surely it has to be higher than that". Which means that no new discovery suggesting that one of the other terms is actually incredibly huge will be sufficient to guarantee a result of more than one. Which means, somewhat sadly, that perhaps we are much closer to being alone than a lot of people, myself included, have always thought.
On a related note, I recently read this BBC article, which discusses the opinion of one SETI astronomer that we should stop structuring the search for alien intelligence exclusively around the assumption that said intelligence will be biological in nature (which is an implicit assumption - and not the only one - of the Drake equation's structure) and instead start to consider the possibility that a lot of that intelligence will - in its own version of our own transhumanism movement - have become non-biological in nature; that we should be looking for civilisations of intelligent machines, which are likely to hang out in very different places to intelligent meatbags. I think this is a fairly persuasive argument. Eliminating the problem of the mind-blowing slowness of interstellar travel (which is essentially a necessity for a civilisation to be truly long lasting) by figuring out how to transplant our consciousness into machines is probably considerably easier than the alternative of getting around the slowness directly with some sort of sci-fi-esque wormhole stuff. At the very least, a lot of people who are experts in the relevant field believe that the former may be possible in principle, whereas, as far as I know, the latter is purely speculative.