According to the Cochrane’s article Biological limits to information processing in the human brain (1995, may be outdated), the human brain is already near a local evolutionary maximum.
The above analysis points to an interesting conclusion: genetic engineering could not be used to make a significant (ten-fold) difference to our information processing ability, since it would require simultaneously improving:
Transmission speed (evolve a better insulator or possibly an organic polymer as a conductor)
Synapse processing speed (without affecting memory ability)
Pulse width (which arises from the same fundamental mechanism that preserves a potential across all cell membranes and so must not disturb cellular function)
Thermal dissipation and energy transport
In a similar manner, drug based enhancement may marginally improve the use of inefficiently arranged or used sub-components but can never realise significant enhancements for the same reasons. There is a role for drugs and genetic engineering, but it is solely to reach the ceiling, not to fundamentally improve it. [...] From our analysis we conclude that the brain of Homo Sapien is within 10 − 20% of its absolute maximum before we suffer anatomical and/or mental deficiencies and disabilities. We can also conclude that the gains from any future drug enhancements and/or genetic modification will be minimal.
Despite the limitations of drugs for cognitive enhancement, I think they may be the best bet for increasing intelligence in the near-term, at least with respect to AI timelines, since other tech would be developed over a much longer time-frame and much further in the future, and I fear we are already operating on the timeline where AGI comes much sooner than genetic engineering and BCI for significant IQ amplification become available. It is more conceivable that we develop a drug that reliably raises latent g and increases IQ by, say 10% in adults, and this relatively small boost has a significant effect when multiplied out by millions scientists taking this drug worldwide. This same argument is given by Bostrom:
Imagine a researcher invented an inexpensive drug which was completely safe and which improved all‐round cognitive performance by just 1%. The gain would hardly be noticeable in a single individual. But if the 10 million scientists in the world all benefited from the drug the inventor would increase the rate of scientific progress by roughly the same amount as adding 100,000 new scientists. Each year the invention would amount to an indirect contribution equal to 100,000 times what the average scientist contributes. Even an Einstein or a Darwin at the peak of their powers could not make such a great impact. Meanwhile others too could benefit from being able to think better, including engineers, school children, accountants, and politicians.
Regarding drug based enhancement, the most promising approaches are optimizing neural efficiency and signal to noise ratio. The best 3 pharmacological targets off the top of my heard are nicotinic a7 nAChr, dopamine D1 (the only PAM I’m aware of is ASP-4345), and adrenergic a2A receptors (think guanfacine). The former is likely why nicotine patches are currently one of the most robust cognitive enhancers available/known. (See gwern). Not using nicotine patches (at least acute use, like 1-3x a week, but there are also cognitive benefits of chronic use, even after cessation) may already be dropping the ball. Another very interesting target is dopamine D2, and there is a study that found a significant association between epigenetic modification/methylation of the DRD2 gene and IQ, which implies that DRD2 is related to the environmental malleability of IQ. Maybe increasing DRD2 methylation somehow → IQ increase?
Also, I want to push back on the idea of more neuroplasticity = more intelligence (in a healthy adult brain). Simply increasing neuroplasticity and LTP is not desirable, as inefficient connections would slow down informational flow and destabilize networks. Rather, improving the identification of inefficient connections for synaptic pruning might be a highly underrated strategy. Some possible pharmacology involves C1q compliment tagging, microglial modulation via CX3C1/CX3CR1, astrocytic engulfment, and possibly a highly selective NR2B PAM.
If it were not for Hebbian decay, dreaming, and the ability to have a temporal depth of storage, we would run out of brain power very early in our childhood.
I speculate that a synaptic pruning optimizer might make us significantly faster at learning and solving problems, but also increase the rate of forgetting unused information and skills. There’s this article describing a savant with near photographic memory: he can memorize entire books in almost perfect detail and recall all matters of fact, but he doesn’t actually understand them. His general functioning is very poor and he has trouble tying his own shoes. He may have a perfect memory, but he has lost the ability to form abstract thoughts and generalize. What I imagine would be in the opposite direction. This tradeoff wouldn’t be too bad in our modern age, as we already increasingly rely on our vast digital knowledge repositories.
You’re right, they are just interesting thoughts that mostly amount to kinda-promising ideas. Sorry for ignoring the main point of your post. I should have prefaced my initial comment with my intention to use you as a springboard to get some of these ideas out in hopes someone more on the ball would eventually see and take them or refine them. I’m currently a nobody with little free time. But I realize a more proper response would have been (and will be) to sit a bit longer with them, and exercise more agency by sharing these ideas more strategically.