Do you like TED Talks? If you do, then I highly recommend that you spend 15 minutes watching one by the American philosopher and famous podcaster, Sam Harris. The subject he tackles in the talk is whether the quest to develop progressively better artificial intelligence (AI) is safe. His concern isn’t AI in general, but specifically the ultimate goal of artificial general intelligence (AGI). The big difference between AI and AGI is that AI is typically a narrow capability to do one task, which might seem sophisticated, like figuring out what movie you want to watch before you even know yourself (thank you Netflix and Amazon), but that system can’t simultaneously also fix your car, manage your financial situation, or help Eon Musk build a rocket to Mars. An AGI can do all of those things and more because it has general intelligence, meaning it can do the full breadth of everything a human can do … and more. Harris doesn’t see how this won’t end in our destruction. SPOILER ALERT: He thinks that even if AGI doesn’t destroy us, he figures we’re likely to do it to ourselves societally. And if you ask a high school teacher, ChatGPT may already be well on the way to triggering Armageddon.
Right up front, Harris warns the TED audience that what he is going to talk about is both terrifyingly dangerous and totally cool. He suggests that this combination (what he calls “death by science fiction”) isn’t actually very good for ape-like creatures such as humans, because, he believes, most of us won’t, or can’t, muster enough survival instinct to do anything about the danger we find oddly appealing before we’re past the point of no return. While I’m not sure anyone wants to be killed by the Terminator, at one point in the talk, he offers a possible solution that I do vibe with, which is the same one that Ray Kurzweil hints is coming. Kurzweil believes that by 2035 (and maybe sooner) we’re going to merge our God-given mental powers with that of man-made AGI. Whether you think that’s cool, or it gives you the screaming heebee-jeebees, Elon Musk is taking this suggestion seriously with one off his other companies, Neuralink, which is developing neural dust – a first attempt at the literal grafting of a computer into our organic brains. If you believe Ray Kurzweil, it’s the first step toward making humans infinitely smarter. If you believe Sam Harris, now may be the time to invest in a fully stocked underground bunker.
While it is generally not smart to bet against Elon Musk, the technical challenges that come with trying to fuse organic and inorganic materials in a manner that will enable seamless integration of cognitive functions aren’t small. In the meantime, while he is toiling away on the next-to-impossible, how about the rest of us focus on making ourselves and our children smarter? As much as I think reading and studying hard in school is a great recipe for doing just that, I mean something quite a bit different. I mean that we should all take the remarkable opportunities in front of us to make the next generation of humans smarter by being smarter about how we make that next generation. SPOILER ALERT 2: I am NOT advocating for genetically engineering mutant kids with giant blue heads, like Megamind.
But the ability to tweak the genetic code of unborn children isn’t confined to Star Trek anymore. CRISPER is a miraculous technology developed over the past couple of decades that enables specific sections of DNA to be spliced out, or into, living cells on a massive (I hesitate to say, industrial) scale. While nearly all of the research being conducted on the uses of CRISPER is laudably focused on treating horrific genetic diseases, it was probably inevitable that someone would decide to turn this powerful technology to altering otherwise normal cells to purposefully make a genetically altered baby. The unfortunate experiment was conducted by a physician in China a few years ago, and twin gene-edited girls were the result. The experiment involved deleting the gene for a receptor called CCR5.
Why CCR5? Well, there are two very important things to know about CCR5. The first is that it is the key that HIV uses to gain entry into immune cells (T-cells). It turns out that the Chinese physician was trying to engineer humans that were immune to HIV, which isn’t necessarily a nefarious goal (but messing around with human babies’ genetic codes without a whole lot of testing is horrifically irresponsible). That said, I’ve got four children, so I’ve witnessed my fair share of irresponsible actions, and amazingly, sometimes we all learn something new as a result. It usually involves a keen observer watching the irresponsible act unfold. After all, Alexander Fleming discovered antibiotics after coming back from a sabbatical to find mold and bacteria fighting to the death on dirty lunch plates he hadn’t cleared before he left on his trip!
I mentioned above that CCR5 does at least two important things, but only mentioned the first. The second is really the receptor’s main purpose, which is to promote inflammation. If you have read any of my other blogs, you know that inflammation plays a role in everything from neurodevelopment to neurodegeneration, asthma to autoimmune diseases, and cancer to metabolic syndrome. So, what did keen observers think of the Chinese experiment (once they got past the horror that it had been done in the first place)?
Well, back in 2016, a pair of scientists in the U.S. had published findings from a broad study in which they had deleted over a hundred (140, in fact) genes from mice, one at a time, to find out if any had an effect on the intelligence of the mice. What they found was that CCR5 deletion gave the mice better memories. Memory is a really important part of learning, and learning is a big part of intelligence. Follow up investigation found that some people actually have naturally occurring CCR5 deletions, and these folks have a host of benefits, like the fact that they typically recover faster from strokes … and people who have flawed CCR5 genes are smarter! This led MIT’s Technology Review magazine to title their 2019 article about the Chinese experiment, “China’s CRISPER twins might have had their brains inadvertently enhanced”.
Hold on, you say? What do you mean smarter? I mean folks with faulty CCR5 receptors are better at school, learn challenging material more easily, remember it longer, apply it more effectively, and do better in jobs dependent on intelligence (of course, all jobs have a significant dependence on intelligence). These benefits make them more capable of making more money, more apt to live healthier lives, and live longer.
Sounds kinda awesome, right? But how does having a disabled CCR5 gene do this? Well, very simplistically, CCR5 promotes inflammation, inflammation causes cognitive dysfunction. Inflammation does this by distracting the brain’s resident innate immune cells, microglia, from doing the really important tasks that they manage. These tasks literally manage the integration of new neurons into the network, especially in the hippocampus, which is how you form memories and learn. So, reducing inflammation reduces cognitive dysfunction.
Wait, stopping cognitive dysfunction caused by inflammation is great and all, but that isn’t the same thing as making babies smarter. True, except it turns out that we all experience inflammation, and there doesn’t appear to be a level of inflammation that doesn’t impact cognitive ability at least a little. Ever try to think when you have a cold or the flu? More importantly, for adults with brains that have largely completed the big processes of neurodevelopment that infants, children, and adolescents go through, the effects of inflammation are real, but far less damaging than in kids. Reducing inflammation means optimizing neurodevelopment.
Ready to sign yourself, or your kids, up for a CRISPER experiment? I’m going to assume not. And for those of you crazy enough to consider it, well, it’s still at least a few decades away from being a designer baby option at the OB-GYN’s office. So, in the meantime, what can we do to make ourselves and our children smarter? The obvious answer is, find other ways to reduce inflammation.
What other ways of reducing inflammation might work? There are reports about TNF- inhibitors, which definitely suppress inflammation, helping with depression and cognitive symptoms associated with autoimmune diseases. These drugs are very expensive and have serious side effects, including cancer risks. Much simpler solutions include good nutrition and low stress environments, both of which have long research histories linking them to better child performance in school and IQ tests. These are great solutions, but if you are looking for a real advantage over your peers, you want something more.
Arguably, there is no group in the world that wants to be better than the competition more than the U.S. military. Whether you think we need to be spending a trillion dollars per year on defense, or not, I think everyone can agree that having smarter people, from the generals at the Pentagon all the way through to the boots on the ground, is probably worth the effort. It is, therefore, no surprise that the military’s research arm, called the Defense Advanced Research Projects Agency, or DARPA, spent close to $100 million studying whether neuromodulation techniques could make people smarter. They studied over a dozen different products, and one technology rose to the top, proving itself over and over to enable faster learning, supporting better performance, enhancing alertness, and even maintaining a better mood. That technology was vagus nerve stimulation (VNS).
So, how does VNS do this? You probably guessed it, especially if you’ve read some of my prior blogs. VNS reduces inflammation, even when it is triggered by bad nutrition, lack of exercise, stress, or sleep deprivation. It does this by causing the release of a neurotransmitter, acetylcholine, that tells the microglia to stop being inflamed and to do their housekeeping jobs that make memory formation and learning possible. The part that takes this from being really cool to truly life-changing is how it effects neurodevelopment in children. It turns out that those same microglial cells don’t just manage memory formation and learning in the hippocampus. During development, even back to the earliest stages of gestation in the womb, microglia are literally the architects and construction teams that build the entire brain. Inflammation (even in from the mother in the womb) disrupts those microglia from doing their jobs during the most critical time. Schizophrenia, autism, ADHD, and bipolar disorder have all been linked to this inflammation. Lowering inflammation may very well reduce the risk of these conditions, but it also enables optimal neurodevelopment, givng that individual the best brain possible.
Maybe Elon Musk and Ray Kurzweil are right, and in the future, we’re all going to be connecting AGI to our brains to make us unbelievably enhanced. If that’s true, human nature dictates that we’re all going to want the best possible version of that AGI connected to our brains. If that is true, then why not start now, making ourselves and our children as smart as we can be. With any luck, those smarter kids will make sure that Sam Harris’s dire predictions of society torn apart by AGI doesn’t happen.