AI and genetic engineering are transforming human possibilities – and are set to fundamentally rewrite what it means to be human
The phrase ‘being human’ means different things to different people. To many, it implies our ability to relate to others and care for others. To an anthropologist it implies reversing the phrase to ‘human being’ – or more technically, to membership of the species Homo sapiens.
But the most important characteristic of being human – the thing that distinguishes us from all other species – is our brain, and what we can do with it. Relating to and caring for others is not unique to humans: the ability to create sophisticated tools is. Although chimpanzees, crows and a few other species can create some simple tools, like shaping a stick to probe for food, Homo sapiens have taken toolmaking to an extraordinary level. In fact, that capacity is so extraordinary that two of our creations, genetic engineering and artificial intelligence (AI), are set to change evolution itself.
By fully exploiting the potential of our uniquely human capabilities, we are set to change what being human means. We are on the verge of creating the next human species.
The impact of AI
Both AI and genetic engineering are already changing our lives. AI is our ability to do ‘smart’ things in a computer, the implication being that this tool is somehow emulating or recreating what happens in the human brain. It is already at work around us – most obviously in search engines, computer games, Siri, Alexa, cars, ad selection and speech recognition. Some not-so-obvious applications include finding new patterns in big data research, solving complex mathematical equations, creating (and defeating) encryption methodologies, and designing the next generation of weapons.
Computers have always exceeded the capability of human brains in some tasks, like doing calculations. But AI now also exceeds human abilities in tasks that were once considered the domain of human intelligence, like playing chess, Go, and other games; solving encryption puzzles; or cognitive tasks that require extensive training for humans, such as interpreting certain medical X-rays or pathology slides. The potential for AI is continually growing and it seems certain that it will replace a significant number of human jobs. Some predict that most of today’s jobs will be performed by a computer by 2050.
But generally, AI’s achievements, while amazing, remain somewhat narrow. We have not yet achieved what is called artificial general intelligence, or AGI. This is the point where a computer’s intelligence is equal to and indistinguishable from human intelligence.
It is the point toward which, in the view of many, AI is heading. There is considerable debate as to how long it will take to reach AGI – no-one believes it to be close – and even more about whether that will be a good thing or an existential threat to humans.
My view is that none of the possible eventualities for AI are existential threats any more than the industrial revolution of the 1800s was – a transformation which some at the time also predicted would be calamitous. We will adapt. Computers will never reach AGI, because we will quickly realize that we don’t really want them to. Human intelligence is too flawed, with biases and other errors, and too intertwined with evolutionary functions like emotions, sleep and hormonal regulation. Why would we try to emulate all of that? Nor will we reach agreement on what AGI actually would be – we still don’t have a definition for human intelligence.
But the power of AI will help drive human progress in understanding and using another era-defining tool, genetic engineering – to the point where it will eventually alter human evolution.
Like AI, genetic engineering is an artificial tool that emulates something which occurs naturally: in this case, evolution. In nature, evolution is the result of random mutations in the DNA of living organisms. It is the process that led to new species of any kind in the past – including humans. Simply, it is the result of a random genetic change that makes an individual more likely to have offspring. Over time, that beneficial change becomes propagated within an isolated group of humans and, along with other mutations, allows that group to diverge enough from all others to become a new species. Genetic engineering is neither natural nor random. It is a deliberate attempt by Homo sapiens to alter the genome of an individual – in any species, plant or animal, including our own.
We’ve been doing genetic engineering for decades, mostly on bacteria, plants and other animals. In most cases, we did it to alter a single gene to do something useful for us. We altered bacteria to produce human insulin so that we wouldn’t need to slaughter cows and pigs to use their pancreatic insulin to treat diabetes. The first plant that we altered for commercial use was tomatoes, so that they spoiled more slowly. To date, we’ve genetically engineered thousands of plants: these genetically modified organisms (GMOs) are an established part of our food supply. The first genetically altered animal to enter our food supply, a salmon that grows faster than normal, is now being sold in some US and Canadian grocery stores.
For humans, we began experimenting with genetic engineering in the 1990s. Of course, that is a really big deal, so we’ve been treading very cautiously. There are two types of cells in our bodies: somatic cells and germline cells. Germline cells are those involved in reproduction, including the sperm in males and the eggs in females – alterations of the genes in these cells get passed on to children and to future generations. Today, it is illegal, everywhere in the world, to attempt genetic engineering on germline cells.
But it is a different picture for somatic cells, which make up all the other cells in the body. Altering these cells affects only the individual: changes are not passed on. Nevertheless, in the US and most other countries, even attempting genetic engineering on human somatic cells is strictly regulated. It is limited to attempts to cure only the most serious diseases and abnormalities caused by genetic defects, including cancer (cancer cells are mutated normal somatic cells) and disorders resulting from a mutation in a single gene, such as sickle cell anaemia, cystic fibrosis and haemophilia. There are over 10,000 known single-gene disorders. Most other diseases, especially chronic diseases with a genetic basis, involve multiple genes such as certain types of heart disease and Alzheimer’s.
The reason that genetic engineering is so strictly regulated is that it is dangerous and not fully understood. In the earliest attempts, there were highly publicized deaths from complications of the procedure. This set back progress considerably. One of the biggest problems is the risk of ‘off-target mutation’. In attempting to alter a specific gene in a particular way, in addition to making that alteration, the same procedure might inadvertently alter the patient’s genome somewhere else. Usually, that is undetected and harmless – but it could be very harmful depending on where it occurs. In fact, in a couple of highly publicized cases involving genetic engineering cures for a white blood cell genetic disease, several patients went on to develop cancers because of off-target mutations.
So why do we allow human genetic engineering at all? Because we recognize that the genetic abnormalities being treated are so severe. Like any procedure in medicine, there comes a point where the hope of a cure outweighs the risk of the procedure, and that certainly is true in all the human genetic engineering clinical trials performed to date. And, like all medical procedures, they improve with time. The benefits become greater, and the risks become smaller. That has been the case in genetic engineering.
The most recent breakthrough in genetic engineering is a technology called CRISPR. In 2012, Jennifer Doudna at the University of California, Berkeley, and Emmanuelle Charpentier at the Laboratory for Molecular Infection Medicine in Sweden, published a paper demonstrating how a naturally occurring set of chemicals in bacteria could be adapted to perform genetic engineering on any organism. They received the Nobel Prize for this discovery.
Before CRISPR, the tools used to do genetic engineering were complicated to learn and use, and took years to perfect for any given procedure. CRISPR has changed all of that. Today, CRISPR is being used worldwide to modify thousands of plants and animals, and increasingly in human clinical trials. It is easy to learn and deploy, inexpensive, and it is much less risky than previous technologies.
Nonetheless, it is not risk-free, and off-target mutations remain one of its side effects.
So what does this have to do with the next human species? Germline genetic engineering. It is coming sometime – probably by the end of this century. CRISPR has advanced our progress in genetic engineering by orders of magnitude and has become big business, with dozens of companies now researching improvements. There are already dozens of variations on the original CRISPR tool with unique capabilities for different types of genetic alterations. These, and the developments that follow, will allow lower-risk procedures, larger segments of gene alterations, multiple simultaneous gene alterations, and many more improvements.
It is not yet safe to attempt germline genetic engineering. But at the pace improvements are being made, I predict clinical trials will begin in some country within decades. There will be many compelling reasons to do so. For example, two parents with sickle cell anaemia cannot have a normal child without it. It will begin as a part of in vitro fertilization. (Remember, germline cells include the sperm in males and the egg in females.) When the sperm and egg unite to form an implantable fertilized egg, or zygote, any alteration to that zygote is a form of germline genetic engineering that will not only affect the child born from that implanted zygote but all future generations of that child.
In fact, there are three children alive today in China who are already the result of such a procedure using CRISPR – illegally. The perpetrator of that procedure is in jail as a result. He jumped the gun, but it won’t be long before some country determines that enough progress has been made to attempt the first legal germline procedures.
When that happens, millions of procedures will be done within decades. AI will certainly be involved in looking at the six billion nucleotides in our genome after any genetic engineering attempt to search for errors. This will support rapid progress.
And from that point, it is likely, in my view, that a fundamentally new human species will
A new human species
This will, clearly, be unlike the way any new human species have emerged in the past. Rather than physical separation providing the mechanism for a new human species to evolve naturally, we will instead create a chemical incompatibility between some genetically engineered humans and those without such a procedure, making reproduction between the two groups impossible.
This will mark the moment at which a new, separately-evolving group of humans emerges – a new species, distinct from Homo sapiens. This species, which I call Homo nouveau, will look like us, act like us, live among us, and have the same cultural variations as us. But they won’t interbreed with us. They will evolve separately. And this will likely happen within the next two centuries.
The frontiers of human possibility are expanding at speed. We are heading towards a profound shift in what it means to be human.