I am also assuming that intelligence is distributed in some normalized way, so that an increase in average species intelligence means that the species “high water mark” for intelligence also increases. So, if the species increases average intelligence, that increases the top end, which increases the likelihood of new species-survival-positive knowledge being added to the knowledge base of the society.

The kind of intelligence it takes to survive (and produce progeny) in the Amazon jungle is different in kind than the kind it takes to work (and have children) as barista at Starbucks. If I was forced to make a quantitative comparison, though, I’d say the jungle requires more intelligence.

The other thing I want to respond to is Karl’s “smarter is better only 50% of the time”. Ted challenged him on it, pointing out eyes and flight; he seems to have backed down. But I think there are two problems with the eyes-and-flight challenge.

First, it may be that species are unlikely to abandon complex mechanisms not because it tends to be “better” (fitter for more available niches) to have those mechanisms, but rather because when such a niche opens up, creatures that are *already simple* have a comparative advantage in filling the niche, in competition with creatures that are complex and have to lose their complex features. In other words, in coming up with a “smarter is better” statistic, it’s a distortion to look at the history of species. It may be rare that when a eat-fish-in-cold-water-hide-on-land-from-predators niche opens up, it’s rare that a bird (like a penguin) colonizes it, having to lose flight in the process, as opposed to a mammal (note that seals lost walking to fill the same niche, though). But that’s irrelevant. What’s interesting is that *something* fills it. If a niche for being dumb opens up, that’s an instance where being dumb is better — it’s irrelevant whether the creature that fills it evolved from something smart or something dumb.

The apples-to-apples way of thinking about it would not be weighted by species history, but by, say, portion of the planetary biomass. How much of the planetary biomass is complex, and is that proportion changing? That’s the “50%” rule. For any two siblings born of a single organism, the more complex one has a competitive advantage only 50% of the time — weighted by biomass, so that we don’t “ethnocentrically” bias the principle in favor of our own little neck of the woods, the niche most suited to complex animals.

(Biomass is a proxy — really at issue is genes. What’s competing, in Dawkins’ language, is the replicators, not the survival machines).

The second objection is implicit in Karl’s statement “the general trend toward greater complexity in earthly life is not the result of some direction to evolution; it’s the result of random drift taking place in the absence of mass extinctions”. Evolution does not go “up” — it goes “out”, expanding in a spherical space, if you will. But if you’re looking at the start of life (or the start after a mass extinction), there’s a floor — you can’t get less complex than zero-complexity, non-life. So that means it’s really expanding in a hemisphere. So the average may be getting more complex, and the top of the hemisphere is going up… but for any given entity at any given time in the middle of the hemisphere, “up” is no more “fit” a direction than “down”.

That depends. Is that follow-on effect inevitable? Or if not, how likely does it have to be?

I think it’s entirely plausible that the human species will destroy itself, or that it will persist for a good long time. Even if we do destroy ourselves, that wouldn’t prove that our intelligence was a harmful trait, because it would just be a single example of a bad outcome. Conversely, if we don’t destroy ourselves, that wouldn’t prove intelligence was a beneficial trait, because it’d be a single example of a good outcome.

If we could examine the multiverse, or if we had a computer capable of simulating the thought processes of every individual on earth, we could see just how often human intelligence leads to destruction. Then maybe we’d find that destruction resulted in 60% of cases, and conclude intelligence is a harmful trait.

Why do I feel that a multiverse-viewer or super-simulator is needed for humans, but not when evaluating the advantages of, say, faster flight in falcons? Because of the property you mentioned earlier: with humans,the actions of a single individual can disproportionately affect the species as a whole. I don’t believe that this property generally holds for other species.

In a way, the human species is like a chaotic system; small perturbations can lead to wildly different outcomes. Come to think of it, some might argue that such instability is fundamentally a bad thing, that even a small chance of self-destruction is harmful in an evolutionary context. In which case, we have enough information to draw our conclusion: the simple fact that humanity is capable of destroying itself means that intelligence is a harmful trait. No further data needed.

I prefer to think of it in somewhat different terms. I think that evolution, and its associated ideas of biological adaptation and fitness, lose most of their applicability when talking about a species as smart as humans. The forces at work are so different from the evolutionary forces acting on other species that we shouldn’t use the same terminology (except metaphorically). Biology has concepts and vocabulary distinct from chemistry, even though animals arise out of chemical reactions; in the same way, we need a different vocabulary to discuss human culture.

I think that’s why we aren’t really getting each other Ted–I see it as being the case that at some point the capability to form the social/technological knowledge pool arises from a capability that is genetic. I mean, the ability to develop a complex language depends on a series of evolutionary “switches” getting flipped. Similarly, the ability to develop technology for recording language depends on both the development of a language, and on a number of biological capabilities being present. Surely we can’t fairly evaluate whether or not a given biological change is beneficial or not, or specifically whether or not an increase in some biological capacity is beneficial, without looking at the follow-on effects of the biological changes.

The question of whether we’ll destroy ourselves with our technology or not is a political/cultural/societal question.

True, but I think it’s necessarily related to the question of whether or not the biological changes that gave us the capability to develop our technology are beneficial or not. If you want to limit the word “evolution” to refer only to first-line biological changes, and not to the second- and higher order effects that result from those changes, that’s fine, but that wasn’t what I was doing.

In any case, if we’re trying to evaluate the change in evolutionary fitness that results from increased biological baseline intellectual capacity, I don’t see how we can do it without reference to the details of the society it arises in, and the accumulated knowledge pool of that society.

Evolutionary success is independent of human values.

That’s probably as good a one sentence summary of Karl’s point as we’re going to get.

I agree that these are two very different things. Strictly speaking, however, only the second one is evolution, because only the second one deals with genetics. The question of whether we’ll destroy ourselves with our technology or not is a political/cultural/societal question. Societies do change over time, and sometimes it’s useful to draw analogies with the way species change over time, but these are two fundamentally different phenomena, even when we’re talking about the global society that contains the entire human species.

Sorry if I’m sounding like a broken record, but I feel the need to emphasize this point. Let’s assume that we want the human species to survive. Is this because we want to be successful in evolutionary terms? No. Sometimes evolutionary success coincides with goals that we find desirable, but that is not why the goals are desirable.

For example: Is it better to have two children that are happy and fulfilled, or to have ten children who are miserable? In evolutionary terms, the latter would be the more successful strategy. So what? Evolutionary success is independent of human values. (And one can’t claim that the ten miserable children are a bad choice because they’ll contribute to ecological collapse; even if we remove ecological considerations, most people would still say having two happy children is better.)

Another example: you said “We might slowly undo the work of evolution by using medcines to continue unfit genetic threads until we wreck our own gene pool.” Evolution does mean that traits that reduce reproductive fitness will tend to decrease in a population, but that doesn’t necessarily constitute good “work” that we are undoing. Would it be better to let people die of treatable diseases because their immune systems weren’t tough enough to save them on their own? Few would say so.

The survival of the human species is an immensely important question, but it’s not a matter of evolution any more than the difference between Republicans and Democrats is. And thinking that evolution is the justification for, say, nuclear disarmament is the same mistake as thinking it justifies laissez-faire capitalism. Policies like nuclear disarmament and laissez-faire capitalism have to be favored or opposed based on our own values, not an appeal to Nature.

About the notion of a simple 50/50 split in evolution/devolution, Ted’s right, and again we have to be careful with definitions.
In particular, the evolution vs. devolution dichotomy doesn’t hold if natural selection doesn’t have a direction. Also, the idea
that some species evolve into other forms tends to make us think that species that don’t aren’t evolving–eg. crocodiles, which are essentially the same beasts now
as they were 100 million years ago. Maybe crocodiles do evolve and at the same rate as everything else–they just keep
evolving into crocodiles.

Anyway, I just wanted to throw some cold water on the idea of us evolving into beings of light after the great rapture of the nerds.
Peter Watts takes these ideas even further–into questioning the utility of consciousness for intelligent tool-makers–in his upcoming
novel Blindsight. And he’s way more rigorous than I am.

The problem, though, with drawing that line hits when you go back to the question of intelligence as an evolutionary advantage–you can’t really talk about it fully while keeping the two separate, because it may well be the case that an increase in capacity is useless without the right social/cultural load, but could be a serious advantage in the right social/cultural environment. To put it sloppily, being smart enough that you could learn to sharpen obsidian has a very different suvival value depending on whether or not you are in a society that knows how to make simple stone tools. The intelligence only does you any good if you have enough of a cultural knowledge pool to let you apply that intelligence to something that affects survival.

Of course, we then also get into the interesting problem of the evolutionary advantage to the species as a whole of advancements in the collective knowledge pool–this is something very different than the model of evolution where changes at the individual level cause certain individuals to be more fit, eventually resulting in the increased fitness of the species. In the case of advancements in the collective pool, the applied intelligence of one member of a species–standing on the metaphorical shoulders of all the preceding contributors to the pool, of course–can radically affect the survival prospects for the entire species. Of course that can be positive or negative–Pasteur figures out germ theory (again, building on the work of others in the pool) versus the Manhattan Project.

Which, of course, brings us full circle to Schroeder’s original point–in some contexts, increased intelligence isn’t necessarily a survival trait. I think that applies whether you are looking at the individual, or the species.

We see one example of this when we look at that combination of native intelligence with the cultural/social pool: we have many ways of causing drastic damage to the species that are consequences of our applied intelligence It’s entirely possible that we might blow ourselves up with weapons that increased intelligence made possible. We might slowly undo the work of evolution by using medcines to continue unfit genetic threads until we wreck our own gene pool. We might overuse medicines and cause bacteria/virii, with their much shorter generation cycle, to evolve into resistant strains that could kill us off. We might pollute ourselves to death with the byproducts of technology. There are lots of potential disaster scenarios where the problem starts because we figured out how to do something, but didn’t see enough of the big picture–the old “smart enough to invent nukes, but not smart enough not to use them” kind of problem.

There is indeed a sense in which modern humans are intellectually very different from Cro-Magnons, in that we have millions of person-years of collective experience to draw from. Obviously that’s a profound advantage, but I don’t want to call it intelligence if we’re talking about evolution or natural selection. Otherwise we’d be liable to say that Europeans have evolved to be more intelligent over the last few thousand years while New Guineans have not, and do we want to say that?

I’m not saying that we can’t use the word “intelligence” to mean different things in different contexts, but we should be clear about our usages. For example, nowadays people can see and hear things happening thousands of miles away, and travel at incredible speeds for many hours without getting tired. Have we evolved super-senses and super-speed? Yes, in a way, we have. But in another way, our vision, hearing, and powers of locomotion haven’t improved at all. I think the same can be said about our intelligence.

You could probably raise a CroMagnon child in our society and have it be as intelligent as any other person–does that mean that he was as intelligent a priori, or that intelligence includes the nurture aspect as well?

That is a great point. I wonder to what extend we, as a society tend to conflate the sum of technological acheivement with collective intelligence, and what other kind ofs of questions we might be misthinking because of that conflation–I know I tend to use a mental shorthand that often equates technological advancement with intelligence when talking at a group level, and I think many other people do as well, but obiviously there is only a narrow set of questions where that equation is true.

That’s certainly possible. In one way, it’s like the evolution of any highly specialized trait. For example, peacocks’ fancy tails are believed to have evolved via sexual selection, in response to the preferences of peahens looking for signs of vigor among potential mates. There’s probably no advantage in a peacock having a tail of N+1 level of fanciness until it’s competing against other peacocks with tails of N level fanciness. It’s possible that ancient peahens accustomed to plain tails might not be attracted to a male with a modern, ultra-fancy tail; peahen preferences evolved in sync with peacock tails, and a modern tail might fall outside the range of what they found attractive. A lot of evolution is like an arms race, and a neutron bomb is of no use in a knife fight (as William Gibson once said in a different context).

When you get intelligent enough to threaten species survival, though, I don’t think evolution has much to say on the issue. For example, consider global warming. Is greater intelligence needed to solve this problem? Some would say no, because it’s a matter of political will, which is largely separate from species intelligence. We may develop the necessary political will to stop global warming, but it would probably be due to social and economic forces, not because smarter people are being born.

But let’s suppose that a genius is born who is smart enough to come up with a better solution for global warming, one that doesn’t require political cooperation; let’s even suppose s/he has the opportunity to implement this solution. Is s/he likely to have so many children and grandchildren that the intelligence of the general population will increase? I tend to doubt it. There’s no evidence that geniuses or innovators, even popular ones like Einstein or Edison, have more grandchildren than average.

This person might invent a new technology that could save the species, and that technology might in turn create other problems that would require yet another new technology to solve. But that would again be separate from species intelligence. Our technology is much higher than that of Cro-Magnons, but there’s no evidence that we are any more intelligent.

Also, the “nothing is free” thing becomes interesting when you look at the dramatic rise in autism and Asperger’s in the developed world, and the extreme (i.e. WAAAAY more than just “statistically significant”) rise in their incidence in places like Silicon Valley–wouldn’t it be neat if breeding the highly intelligent geek types together resulted in the start down a path to idiot savants, or something equally unpleasant to our view.

But what is probably more relevant, there may be no selective advantage to be gained by flying faster; superfast falcons probably won’t leave any more offspring than regular ones. There may not be enough fast-moving prey for superfast flight to make a difference.

This is compounded by the fact that, in developmental biology, nothing is free; faster flight would almost certain involve reduced abilities in another area. The resources needed to grow a telescopic eye might be better used in keeping a decent sense of smell. So every species winds up with a compromise of traits to suit its particular needs.

Similar factors are at work in the evolution of intelligence, and that is probably what Schroeder is trying to point out.

For example, he says, “smarter is better only 50% of the time. Sometimes, new organisms survive because they’re smarter than their ancestors; equally often, it’s the dumber ones who survive. This fact is reflected in the fossil record.” He seems to be saying that there have been just as many cases in which species devolved, i.e. lost a trait that they had previously acquired, as there are in which species further developed that trait. I think this is incorrect. Some species have lost their eyes, but only a tiny fraction compared to those who kept them. Some birds have lost the ability to fly, but I don’t think its anywhere near half of all bird species.

Of course, there are many species that never evolved eyes, or the ability to fly. So one could say that vision or flight aren’t necessarily survival traits, because lots of species get by fine without them. But the vast majority of those species never had them in the first place. So it’d be incorrect to conclude that descendants often lose traits that their ancestors had, which the statement “equally often, it’s the dumber ones who survive” implies.

(I’m using vision and flight as examples rather than intelligence, since those traits leave clearer fossil evidence.)

On your first point, though, I think you’re being a bit picky. If his point is that intelligence is not always a survival trait, I think that includes the notion of environment–that any time intelligence might develop, it may be beneficial or not, in that environment.

With his wording, he’s also guilty attributing conciousness to the process: “Intelligence is just one of a whole suite of traits an organism can choose from” is the thin edge of the ID argument, since “choose” implies direction to the process. The river does not “choose” to make the rocks in its path round, it’s just a result of the process of the water moving.