I have a theory that intelligence is a compensatory mechanism for lack of fitness. For instance, whales and dolphins are smarter than fish, because they are not very well adapted to living under water (they can't breathe under water). Humans are pathetically unfit, so our only chance was to be smarter than our predators. Not stronger, nof faster--just smarter. Brain vs. brawn.
I'm talking about evolutionary fitness: how well we fit our environment.
Climactic change in Africa forced our ancestors to fend for themselves in the savannah, where they couldn't just climb a tree to escape a predator. They were pushed to the very edge of extinction until they compensated by increasing their brain size. It was a huge gamble. For a billion years no other species was successful in harnessing intelligence as a survival strategy. We just didn't have any other choice and we got extremely lucky to slip through a very narrow bottleneck.
@BartoszMilewski I think it probably heads the other way.
Intelligence is a pre-requisite for social behavior, cooperation is a superior strategy, more energy goes into brain power over strength over time.
In other words intelligence gets optimized for, it's not a coping mechanism.
@ryan
Explain ants
@BartoszMilewski cheeky answer: ants are smart, collectively
real answer: we were discussing mammals and insects are probably a different conversation for reasons way beyond my understanding.
@ryan
I was comparing mammals to fish. But instects are immensly fitter than vertebrates.
The point about ants is that collectivism doesn't require intelligence. And there's a mammalian example of this: naked mole rats.
@ryan If this were true, the Earth would be teeming with intelligent life at all levels.
Oct 9, 2026, 5:10 PM@BartoszMilewski humans evolved being supremely fit, in the long distance/endurance sense. Humans being probably the best species at this.
@stuartgale
Correct. But that's a hunting strategy, which doesn't help against predators. In fact humans went almost extinct about 100 thousand years ago, before we bounced back.
Oct 9, 2026, 5:49 PM@BartoszMilewski Sure it's a hunting strategy, but it's one that few other species can do. Humans are efficient movers because of being bipedal and our ability to sweat to cool down. The problem then is the (initial, lack of) definition of 'fitness'.
It'll be a combination of smart (ie constructing novel tools) and endurance that's the game changer.
@stuartgale
I should have been more precise: I meant evolutionary fitness. Fitting the environment.
@BartoszMilewski there might be something to this for Homo Sapiens because we use ballistics. But other humans were pretty intelligent too. They did so some tool use but I imagine social behavior and sexual selection is more to do with getting to a very high level of intelligence. You need to be able to model other animals and what they are thinking.
Oct 9, 2026, 5:18 PM@BartoszMilewski Isn't every trait like this? Modelling evolution like gradient descent in gene space, every gene feels a selection pressure based on the others + the loss function coming from the environment. The genes that will change the most are the ones not stuck in a steep local minimum.
Does this make sense? 🤔
@mc Yes, there is a Monte Carlo game constantly playing in which the minima keep shifting. After billions of years all the very deep and stable minima are occupied by life forms of extreme fitness. Humans got kicked out of our local minimum, from a forest to a savannah. We were supposed to quietly go extinct for lack of evolutionary fitness.
Oct 11, 2026, 11:48 AM@BartoszMilewski instead... invents category theory like a boss 😏
Oct 11, 2026, 11:51 AM@BartoszMilewski when writing my reply above I was wondering if genes are playing a game or are parameters of a learning process. I guess it's a matter of how we model the situation: the *distribution of genes* across a population is 'learning' (because this parameter can reasonably be said to vary continuously if the population is quite large) but the single genes in an organism can only vary discretely and thus fall back into a game-like dynamics.
Humans can run in heat for hours without overheating. we can moderately jog/walk for indefinite periods of time.
Eusocial behavior + general purpose body plan can go far.
@mlhaufe
That's a hunting strategy that doesn't work against predators
Isn't our primary predator other humans?
@mlhaufe Yes, but not on an evolutionary scale
@BartoszMilewski why do you separate "intelligence" from fitness? How does your hypothesis apply to creatures such as Corvids?
@BartoszMilewski I think the premise here puts this notion of "fitness" as both primary and somehow separate (I suppose you mean physical fitness). I am not sure what the reason for that would be. Perhaps other organisms are more fit to compensate for their lack of intelligence.
@alexanderbakst
I mean evolutionary fitness. Darwinian.
@BartoszMilewski Does the theory work with crows? Their place+body plan seems pretty well adapted as far as I can tell.
@drScott Bird intelligence is interesting. They are dinosaurs that were thrown into a very demanding environment, so their evolutionary fitness is precarious. Much worse than that of flying insects. But because of weight limitations, they had to miniaturize their brains. I'd say they pack more intelligence per gram of brain tissue than land animals (not to mention aquatic animals).
are humans unfit ?
afaik, when it comes to running, the human animal is superb
the human hand and gut allow the human to eat a wide range of foods
sounds pretty fit to me
@BartoszMilewski how would jumping spiders fit into this theory?
@sojournTime
Hard to say. The evolutionary niche for spiders is their ability to hunt using webs. For some reason jumping spiders stopped spinning webs (maybe external changes, like winds, rains? maybe they were outcompeted in the webiverse?). So suddenly they were in an unfamiliar territory, where their evolutionary fitness was challenged.
Oct 10, 2026, 2:55 PM@BartoszMilewski Perhaps the notion of intelligence is a hoax? To claim superiority over those that are considered less intelligent. I'm not sure that can be generalized to everybody. The question is whether the amount of work spent on "understanding" is a measure of intelligence?
@esap
I would agree that IQ is sort of a hoax.
@BartoszMilewski These are surely just alternative strategies. Neither is compensation for the other.
@dpiponi You're objecting to the word "compensate"?
@BartoszMilewski Yes, it's loaded.
@dpiponi True. Evolution doesn't have goals and it doesn't make decisions. You used the word "strategies," which is also loaded. Any time we talk about evolution, we anthropomorphise. I don't have a good non-teleological language to talk about it.