Cognitive Enhancement in a World of Superabundance
Cognitive Enhancement in a World of Superabundance
Addressing an objection to genetic enhancement for IQ
The general public is slowly becoming aware of the immense potential of artificial intelligence. However, few people believe that AI is basically the only thing that matters right now. It just so happens that the people who are convinced of the immense power of AI are also the type of people who are interested in human genetic enhancement. Due to this not-so-coincidental relationship, I often see the objection that boosting IQ with genetic enhancement is a waste of time. The thinking is that we are either doomed or we will live in a world of superabundance very soon.
It looks like the stigma associated with biorealism has delayed genetic enhancement research to the point that a strategy of using highly cognitively enhanced humans to address the AI alignment question is probably not feasible. However, it is possible we still have time, and genetic enhancement would be significantly more important than any near-term concerns like the culture war if AI doom is a realistic concern. Aside from existential risk, we could potentially end global poverty and accelerate economic development while we wait for the singularity. As always, the future is hard to predict. There are definitely many more trivial ways to spend our time other than accelerating genetic enhancement.
Another perspective is that a superintelligence will give us superabundance very soon. Imagine a utopia in which no one has to work, everyone is extremely wealthy, and all diseases have been cured. Imagine that there is an oracle that can provide us with all sorts of information about history, science, and the universe with a high degree of accuracy. Certainly our lives would be better. In such a world, would it be useless to be more intelligent?
Likely, the economic returns to high IQ would diminish in such a world. However, there are other benefits to high IQ that would still be relevant even in a world of superabundance and practically infinite knowledge. People with a higher level of cognitive ability tend to make better choices. While the economic impact of bad choices will be greatly diminished, there are still unfortunate downsides to low IQ outside of its economic impact.
In our current world, there are some decisions that almost every reasonable person would say is a bad idea and yet they are still made. For example, having a lower IQ increases a person’s risk of committing crimes. It also increases the likelihood of getting into a traffic accident. Even if we had a superintelligent oracle, it is not clear that it could easily facilitate some of the beneficial behaviors associated with high IQ, such as prosocial behavior and maintaining a marriage. A high level of cognitive ability also provides a richer appreciation for the world’s complexities. Unless we live in a world in which the AI coerces or convinces everyone into making the optimal decision at every moment, it would appear there are still benefits to being intelligent.
Perhaps we will have some means of cognitively enhancing already existing people, maybe with a brain-computer interface or in vivo gene editing. It is not certain whether such a drastic intervention would allow a person to maintain their identity. If I were given the option to massively change my brain with surgery, I would be concerned that I would not wake up. Not that I would merely be changed, but that I would cease to be a person and some other person would be taking my place. In a world of superabundance and extremely long lifespans, I think that I may refrain from such a procedure, and I imagine others would too.
In my view, genetic enhancement for health is more undermined by artificial superintelligence than the case for cognitive enhancement. It is not unreasonable to think that by the time the children born with enhancement in the coming years grow up, AI or humanity will have eliminated many of the late onset conditions that we are using PGT-P against now. For that reason, prioritizing cognitive enhancement to some extent seems defensible. Jonathan Anomaly’s proposal of using prediction markets to appropriately weight certain disorders by factoring in their probability of being cured seems wise.
All of this is rather speculative, but it seems reasonable to anticipate that AI will have a large impact on the economy and improve our life for the better. In the face of uncertainty, we need to decide how to spend our time and what choices to make. If you truly believe nothing matters other than AI, then it would be prudent to advocate that humanity invest heavily in getting AI aligned. But from such a perspective, nothing else really matters. That is a coherent view, but superintelligence is far from a certainty. It would be wise to continue to invest in this technology as it is perhaps one of the most important alternative routes to immensely improving the world.
Another person who grapples with the question of how AI and cognitive enhancement are related is Anatoly Karlin, who has recently explored this in his article “The Biosingularity is Near.”
I totally agree - in a post-scarcity society, status is entirely about positional goods, and IQ and smarts are positional. IQ is going to matter *much more,* not less, in a post-scarcity society.
I had an interesting debate with Bryan Caplan's sons about education-as-signaling a few months ago. I maintained that millions of parents wouldn't be engaging in bitter Red Queen's Races to get their precious Jaydens into the right preschools, then grinding furiously for 18 years to ultimately get them into Harvard, unless the benefits were actually there and actually significant.
They maintained there were studies showing IQ matters more than undergraduate institution for career success - ultimately we agreed there probably is something there, and what it is, is posterity. An Ivy degree is a positional status good, and what it gets you is access to a better and higher status pool of mates, and presumably some decent percentage of the time, access to that pool pays off in mate and grandkid quality.
Similarly, IQ in a post-scarcity society will be purely positional, and presumably increase the quality of your grandkids, by increasing the quality of mates in the pool available to your high-IQ kids.
Agree on all points (and thanks for the call out).
I would say that even though there is a good chance that it will be way too late to use any of the extra brainpower generated from human intelligence enhancement that begins today on the alignment problem, it is still as Yudkowsky might say a more dignified usage of limited research resources than most anything else you can do apart from AI.
Besides, as you also correctly point out, the future is hard to predict. While it would have been best to accelerate intelligence the day before yesterday than today, it's nonetheless better to do it today than tomorrow.