Chevron icon It indicates an expandable section or menu, or sometimes previous / next navigation options. HOMEPAGE

We're about to become more intelligent than at any other point in human history

albert einstein
AP

We know of humans that we consider super-intelligent, but we don't yet know how to engineer that intelligence — or exceed it — in people.

Advertisement

But some researchers think that advances in genomic science and machine learning are going to open up new avenues of possibility in that realm, potentially leading to individuals whose cognitive abilities leave the greatest minds of history in the dust.

That future of superintelligent humans may be upon us sooner than we think.

Consider individuals that we consider the smartest of all time, those like Carl Friedrich Gauss or John von Neumann, says Stephen Hsu, a physicist who is the vice president for research and graduate studies at Michigan State University and an advisor to the genomics researchers at BGI. Hsu is a member of BGI's Cognitive Genomics Lab, a research group that's trying to unlock the genetic codes that account for complex traits like height, susceptibility to conditions like obesity, and — perhaps most controversially — intelligence.

Most researchers believe that intelligence is influenced by genes and environment, and when it comes to genes, we think a large number of genetic variants all make very small contributions.

Advertisement

It wasn't just genes that made Gauss or von Neumann or Albert Einstein the geniuses they were. Still, if we could analyze the many genes involved, those individuals most likely had a far greater genetic predisposition towards intelligence than the average individual.

Once we learn how hundreds (or more) of individual genetic traits interact to code for intelligence, we might be able to select for those traits. Some researchers think we might even be able to use new genetic editing tools to flip all those switches in an "optimal for intelligence" direction. If that's the case, even the greatest intellects in history could potentially be given far more to work with.

"Gauss or von Neumann, they didn’t have any huge or debilitating problems, but they were still carrying many [genetic] variants of the suboptimal type [for intelligence]," says Hsu. "Surely some of those variants could be flipped without causing them to get [a genetic disease]."

If we can do that, we might have individuals that are whole levels of intelligence above any human that's ever existed, able to figure out ways to exceed whatever mental limits we once thought humans had.

Advertisement

For now of course, this is all speculative. Other researchers think intelligence is far too complex to ever be modified genetically. The interactions between environment and genes may be too complicated for any tweak to make a difference. And for now, genetic editing technology would be incapable of making the hundreds of changes presumably necessary to make a superintelligent human.

But those genetic tools are changing rapidly, and Hsu also thinks that we might simultaneously augment our own intelligence with computers or AI technology. Here's how he sees the potential emergence of superintelligent humans.

The road to superintelligence

The initial key is understanding the genetics of intelligence in the first place. We first mapped a human genome in 2003, but in a way, that only showed us how complicated a system we were dealing with. We've been able to identify certain genes that always cause certain diseases or traits, but the codes for complex traits remain elusive.

genome
Shutterstock

That's what scientists are trying to change now and many, Hsu included, think that this could just be a numbers game. Major genome sequencing projects right now like the Precision Medicine Initiative, the UK Biobank, and those underway at places like Craig Venter's Human Longevity, Inc., are trying to map millions of genomes. All that data might help illuminate the causes of complex diseases — along with the genetic variants for other complex traits like intelligence.

Advertisement

This is controversial. "People don't want to accept that human cognitive ability is a trait like others," Hsu says. But if it does turn out that we can identify even a large proportion of the genetic traits responsible for intelligence, there's no reason to think we wouldn't be able to "push" humanity further down that road, provided we got past the distaste that idea gives many of us.

Hsu doesn't think we'd immediately start editing human embryos, even if we do find the genetic keys for intelligence in say, the next 10 years. Instead he sees it as a process that will start with embryo selection and in vitro fertilization.

As he points out, technicians and doctors already engage in some degree of embryo selection when choosing embryos for in vitro fertilization. In some countries they may select embryos of a certain sex, in others they might just pick ones that appear healthy. But if genome sequencing reveals more details about the embryos and their potential intelligence (or athletic ability or other characteristics), it could be possible to select for those characteristics too.

This is an idea that's incredibly controversial to some — there are questions of ethics, and fairness, as well as what will almost certainly be a troubling disparity in who would have access to such technology — but it's not equally controversial everywhere.

Advertisement

"Different countries have different standards for pre-implantation selection," bioethicist Josephine Johnston explained at a World Science Festival discussion on genome engineering on June 3. While some countries strongly oppose selecting embryos for "better" babies, there are different policies in different places. Johnston explained that England has laws against that type of selection; in the US that selection violates professional guidelines; and other countries have no opposition to that selection at all.

But if people somewhere start doing that sort of selection, others will be tempted to follow suit. "Maybe even before it becomes a reality there will be rumors that rich people are doing this," says Hsu.

Some places might even offer embryo selection as part of government healthcare policies, he says. After all, you'd first select to avoid dangerous, debilitating, and expensive diseases, which could reduce pressure on healthcare systems.

A whole new plane of intelligence

If you take selection for granted, the next question that comes up is genome editing.

Advertisement
Garry Kasparov deep blue IBM chess artificial intelligence machine learning
World chess champion Garry Kasparov studies the board shortly before game two of the match against the IBM supercomputer Deep Blue. Reuters

Provided the variants for intelligence can be figured out and the genetic editing tools sufficiently refined (no small matter, in either case), Hsu thinks there's an incredible potential for improvement. He's written for the science magazine Nautilus that his calculations from the work at BGI indicate a potential for "very roughly, about 100 standard deviations of improvement, corresponding to an IQ of over 1,000."

That's a level of intelligence beyond what we can comprehend right now.

At the same time, we're not just working on improving biological intelligence. We're also working on artificial intelligence and machine learning. Smarter humans might be better able to solve those problems, helping us create smarter machines. Smart machines capable of processing big data are already essential for efforts to understand millions of human genomes. These things work together.

Some visionaries, like Elon Musk and Stephen Hawking, have worried that smart machines and AI might leave us too far behind — "house cats," as Musk recently put it.

Advertisement

But smarter AI could also be incorporated into the human brain, creating a machine-supplemented mind that's far more capable than our own (we already rely on computers for memory, search, and computation; imagine if those capabilities could be built into our brains). "A third, digital layer that could work well and symbiotically" with our minds is how Musk has described this.

Beyond that, smart machines might help us change the way we think now. Hsu explains that when computers started to get better at chess than human Grandmasters, we learned that certain avenues of play that we'd thought were useless could actually be incredibly effective — we just hadn't been able to see their utility. AI might make a quantum physics breakthrough that we hadn't thought of on our own, but the smartest of us then might be able to make use of that knowledge ourselves.

As we talk about moving forward toward these currently unattainable levels of thought and analysis, the point is that we have no idea what the limits of human intelligence are.

Hsu wrote for Nautilus that in 1983, Noam Chomsky answered a question about whether we've hit the barriers of human intelligence by saying:

Advertisement

You could give an argument that something like this has happened in quite a few fields ... I think it has happened in physics and mathematics, for example ... In talking to students at MIT, I notice that many of the very brightest ones, who would have gone into physics twenty years ago, are now going into biology. I think part of the reason for this shift is that there are discoveries to be made in biology that are within the range of an intelligent human being. This may not be true in other areas.

Yet we've thought we've hit the barriers of human intellect in the past, only to be proven wrong.

With genetically engineered or computer-supplemented minds, who is to say that the same won't happen again soon?

Advertisement
Close icon Two crossed lines that form an 'X'. It indicates a way to close an interaction, or dismiss a notification.

Jump to

  1. Main content
  2. Search
  3. Account