This Kevin Kelly dude has written a summary that I find fully compatible with the biology. Read the whole thing — — it’s long, but it starts with a short summary that is easily digested.

Here are the orthodox, and flawed, premises of a lot of AI speculation.

1. Artificial intelligence is already getting smarter than us, at an exponential rate.
2. We’ll make AIs into a general purpose intelligence, like our own.
3. We can make human intelligence in silicon.
4. Intelligence can be expanded without limit.
5. Once we have exploding superintelligence it can solve most of our problems.

That’s an accurate summary of the typical tech dudebro. Read a Ray Kurzweil book; check out the YouTube chatter about AI; look at where venture capital money is going; read some SF or watch a movie about AI. These really are the default assumptions that allow people to think AI is a terrible threat that is simultaneously going to lead to the Singularity and SkyNet. I think (hope) that most real AI researchers aren’t sunk into this nonsense, and are probably more aware of the genuine concerns and limitations of the field, just as most biologists roll their eyes at the magic molecular biology we see portrayed on TV.

And here are Kelly’s summary rebuttals:

1. Intelligence is not a single dimension, so “smarter than humans” is a meaningless concept.
2. Humans do not have general purpose minds, and neither will AIs.
3. Emulation of human thinking in other media will be constrained by cost.
4. Dimensions of intelligence are not infinite.
5. Intelligences are only one factor in progress.

My own comments:

1. The whole concept of IQ is a crime against humanity. It may have once been an interesting, tentative hypothesis (although even in the beginning it was a tool to demean people who weren’t exactly like English/American psychometricians), but it has long outlived its utility and now is only a blunt instrument to hammer people into a simple linear mold. It’s also even more popular with racists nowadays.

2. The funny thing about this point is that the same people who think IQ is the bee’s knees also think that a huge inventory of attitudes and abilities and potential is hard-coded into us. Their idea of humanity is inflexible and the opposite of general purpose.

3. Yeah, why? Why would we want a computer that can fall in love, get angry, crave chocolate donuts, have hobbies? We’d have to intentionally shape the computer mind to have similar predilections to the minds of apes with sloppy chemistry. This might be an interesting but entirely non-trivial exercise for computer scientists, but how are you going to get it to pay for itself?

4. One species on earth has human-like intelligence, and it took 4 billion years (or 500 million, if you’d rather start the clock at the emergence of complex multicellular life) of evolution to get here. Even in our lineage the increase hasn’t been linear, but in short, infrequent steps. Either intelligence beyond a certain point confers no particular advantage, or increasing intelligence is more difficult and has a lot of tradeoffs.

5. Ah, the ideal of the Vulcan Spock. A lot of people — including a painfully large fraction of the atheist population — have this idea that the best role model is someone emotionless and robot-like, with a calculator-like intelligence. If only we could all weigh all the variables, we’d all come up with the same answer, because values and emotions are never part of the equation.

It’s a longish article at 5,000 words, but in comparison to that 40,000 word abomination on AI from WaitButWhy it’s a reasonable read and most importantly and in contrast, it’s actually right.

Don’t you just love it when people like @FerranSuay wrap themselves in the Prestigious Robes of Science and Evolution, and then make a series of statements that show they understand neither, and fail at logic to boot? A professor of psychobiology has written an essay in which he equates a refusal to make natural selection omnipotent with creationism. It’s a familiar and wrong tirade. I should have been keeping track of how often I get accused of being a creationist because I find evolutionary psychology poorly founded and full of sloppy research, because if I had, I’d have a really big number.

In academic environments it is very difficult to find someone who will openly and explicitly deny the principles of evolutionary theory. Professors and researchers from any scientific discipline will endorse, more or less accurately, the principles of natural selection, and everyone has a rough idea about what genes, chromosomes, and DNA are. Certainly, nobody will deny that we walk on two legs or have a hand with an opposable thumb because evolutionary pressures have shaped our anatomy in this way. And very few academics refuse to acknowledge that human brains underwent a unique frontal development, which clearly distinguishes them from those of other primates, and even those of our closest relatives, the great apes. This is accepted as an obvious consequence of the evolutionary process that has shaped life on Earth today.

But the situation is very different when we apply the same principles to the study of human behaviour. In this area, there are scientists prepared to deny any genetic influence whatsoever. Some will say instead that behaviour is wholly the product of social and environmental variables. Others will try to consistently minimize the explanatory power of genetics. But how can a species rid itself of the laws that govern the rest of life on the planet?

See the highlighted sentence? Name them. Go ahead. You should be specific in your claims if you’re in those science robes, you know. I don’t know anyone who fits that description, unless he’s thinking of some fringe New Age wackaloon like David Avocado Wolfe.

Only a few minutes of thought reveals all this to be extraordinarily unscientific.

Yes, I agree, that essay is extraordinarily unscientific. It’s not going to get better.

Are we to believe that evolutionary pressures, which have configured the anatomy of the body and the brain, cannot also be used to explain and understand the whys and wherefores of human behaviour? Everyone agrees that we have opposable thumbs because those of our ancestors born with this mutation possessed certain reproductive advantages and left more living descendants on Earth. As this trait continued to provide benefits to subsequent generations, it became so dominant it is now the norm for the vast majority of humans. The same can be applied to the standing position, and to the size and the particular anatomical configuration of the human brain. This is all uncontroversial.

We can credit all kinds of things to evolution, but this fellow has three major problems: 1) he thinks all of evolution is explained by natural selection, 2) he assumes that every single feature of the human form is adaptive, and 3) he has this overly simplistic notion that opposable thumbs are a product of a “mutation”. Every one of those points is false.

Do we need to go on after he reveals that all of his premises are wrong? Of course we do, for the spectacle of someone digging themselves a very deep hole.

Why should the same logic not apply to human behaviour? Let’s take physical aggression, for example—the tendency to impose on others through coercion. Didn’t aggressive individuals enjoy (some) reproductive advantages? Didn’t the most aggressive males climb the hierarchy of social groups thereby enhancing their ability to attract resources and mates? Didn’t that privilege the transmission of aggressive genes to the next generation? The statistics on violent crime reveal a very clear over-representation of the male sex. Without needing to study the numbers, anyone with eyes in their head can conclude that human males are generally considerably more physically aggressive than females.

That explains nothing. It’s a lazy, sloppy attempt to justify a patriarchal status quo without looking for any evidence.

The logic is wrong. If physical aggression is an advantage for men, why not also for women? Wouldn’t aggressive women enjoy some reproductive advantages? Don’t women experience hierarchical social groups? Why is this being framed as a male thing with arguments that should apply to all sexes?

And if you want to argue that submissive behavior is advantageous for women (somehow I suspect he would), wouldn’t it also be the case that submissive behavior would be advantageous for men? He has trapped himself in an argument that can work in any direction you want.

Let’s look at reality, too. Does this professor expect to climb the rungs of the hierarchy at the University of València with physical aggression? That would be truly remarkable. Universities are not purely intellectual meritocracies, but still — using violent crime to work your way up the ranks probably wouldn’t work. Social skills are far more important. Attempting to coerce one’s colleagues with a good punch-up or skillful use of a club will not get you far.

It’s nice of him to announce that you don’t need numbers, since he doesn’t have any.

However, unlike the shape of our hands, the standing position, or the anatomy of the brain, this trait is not a universally accepted product of evolution. Instead, it is a response to social conditioning, such as patriarchal education, the nefarious influence of the media, or the excessive availability of violent video games. In this scenario, miraculously, evolutionary pressures have no part to play, and the socio-environmental, psychosocial, or psycho-socio-environmental variables (we can keep on juxtaposing terms until we find a sufficiently abstruse formulation) are the sole determinants of behaviour.

There he goes again.

Look. This shouldn’t be so difficult to understand. You did not evolve to be well-adapted to your academic niche. Evolution gave us a plastic brain capable of learning and adapting — in an immediate, developmental sense, not an evolutionary sense — to diverse and complex circumstances. You are capable of both bashing in a competitor’s skull with a rock, or publishing papers to demonstrate your cognitive superiority (this guy ain’t doin’ so well on that front). We can simultaneously see that human minds have a genetic predisposition to process, understand, and use symbols, and a learned ability to speak Spanish, English, or Russian.

These are not hard concepts to reconcile. A genetic/biological substrate that has many predispositions and capabilities, plus a general ability to learn and modify behavior on the basis of experience is the obvious, universal, scientific understanding of the human mind. It is simply perverse to throw away the last part of that description and believe in strict genetic determinism of behavior.

But when psychologists deal with one of the most complex phenomena we know about, human behaviour, they must discard the methods that have proved useful, and the knowledge derived from them, and embrace a new faith; one that says that the cause of behaviour are to be found only in social and environmental variables. This is unscientific and intellectually dishonest—it is creationism by another name. Only it is “hidden,” because its advocates will not openly resile from evolutionist positions and, instead, drape their irrational beliefs in the prestigious robes of science.

Since genetics is relatively inflexible, it is only reasonable to assume that the cause of variation in behavior may be a product of learned experience. That there may also be some genetic biases between individuals is not off the table, but dang, guy, you’ve got to do a heck of a lot of work to show that.

Because science expects you to drop the pretentious Robes of Science and buckle down to work.