I would guess one change would be a higher metabolic demand on glucose + oxygen, beyond that which the body can supply. Like a CPU that requires more peak power than its motherboard VRMs + PSU are capable of drawing.
The brain already has a large metabolic cost, even dominating the total metabolic needs of the body during early development (https://www.pnas.org/content/111/36/13010).
I've long held a suspicion that the main reason IQ has a normal distribution is not anything to do with brain architecture per se (i.e. it's not a polygenetic trait that builds some brains out of better or worse genes than others) but rather that the brain is limited in its ability to become more complex by a proportional need for metabolic energy; and that the metabolic efficiency of human bodies is a polygenetic trait, such that the body systems required for metabolism are built from better or worse genes, that will thus get energy to the brain more or less efficiently. (This would explain why the brains of higher-IQ people don't look any different under histological analysis—there's nothing genetically or epigenetically different in them, in terms of what proteins are being expressed. Brains are brains. The differences that determine brain complexity would be elsewhere, in their bodies!)
This also, in my thinking, explains the Flynn effect: anything that we as a civilization do to get rid of an obstacle in the way of our metabolism—e.g. decreasing parasite load, stopping exposure to environmental toxins like lead or pollution, fortifying foods with vitamins, etc.—should bring the average human living within civilization ever closer to "peak performance" of the human body's metabolic system, and thus give the brain more "headroom" [hah!] to become more complex.
Of course, the Flynn effect says that this only happens to new generations (who grow up with such advances in place); not to older people (who don't grow up with such advances, but are exposed to them later in life.) I would suppose we just have some epigenetic triggers that "give up" on brain complexification after a certain point in life, probably assuming that whatever equilibrium the brain has reached between growth and apoptosis-through-energy-starvation by that point, is the final limit.
Alternately, as proposed here (https://en.wikipedia.org/wiki/Synaptic_pruning#Energy_saving...), the body might do well-enough to feed the brain when that's the body's only job; but not well-enough to feed the brain when both the brain and the sexual organs (and all descendent demands, e.g. pregnancy) are fighting over metabolic energy. So the "throttle" on the brain's complexification becomes "choked off" during puberty, such that the resultant metabolic energy can be reserved for reproduction. (Under that hypothesis, preventing puberty might result in higher-IQ people. It apparently worked in rats!)