[thaumasiotes]

> Flying birds of necessity optimize heavily for weight. This implies their bird-brains must be remarkably efficient for their size.

The first sentence is true. The second is a meaningless non sequitur.

You could equally claim that bird bones must be remarkably efficient for their size. That is true, but only in terms of mass used to prevent a given amount of bird from degenerating into a puddle. In terms of physical robustness, which is a primary purpose of all bones including bird bones, bird bones are straightforwardly worse than normal bones.

You can't just "be efficient". You have to be efficient in terms of a metric. And optimization for weight does not imply that you're efficient in terms of computational power, whether that be computation produced per unit of sugar consumed, computational capacity per unit of structural mass[1], or any other computation-related metric.

[1] "But it's optimized for weight!" Yes, but that just means it uses (ideally) the minimum weight necessary to support however much computation it wants to do. Using more of a denser material might produce much higher capacity-per-gram figures, but it would also weigh more.

And an important part of optimizing for weight is to reduce the needs that your massy infrastructure must satisfy; by this line of reasoning, you would expect birds to be exceptionally stupid, not smart.

[WalterBright]

Bird bones are shaped optimally to put the bone where the most stress is. I recall reading once that they were hollow.

> You have to be efficient in terms of a metric

Since the topic here is intelligence, that is the metric.

[thaumasiotes]

Intelligence is not a metric that the concept of efficiency can be applied to. To evaluate efficiency, you need a benefit and a cost. "Intelligence" is just one thing; the most you can do is evaluate it in terms of more or less.

What kind of intelligence efficiency do you want? Most efficient use of time? Glucose? Volume? Mass? Highest utilization rate? Are we looking at computational capacity? Computational quality? Economy in computational demands?

> Bird bones are shaped optimally to put the bone where the most stress is.

That is true, but it doesn't distinguish bird bones from the bones of any other species. However, bird bones are uniquely bad at suffering that stress.

> I recall reading once that they were hollow.

Also true. They also have very low density in what you might consider the non-hollow parts. That's why they function so poorly for maintaining physical integrity.

Sometimes two goals are in conflict. Birds need bones to maintain their physical integrity, but in order to fly, they have trouble supporting bones. The fact that their bones are subject to more constraints than those of other vertebrates doesn't mean that they magically have better bones than other vertebrates do. It means they have worse bones, because the quality of their bones had to be subordinated to their unique need to fly.