The Augmented Human Being

The Augmented Human Being

George Church [3.30.16]

There are now 2000 gene therapies where you’ll take a little piece of engineered DNA, put it inside of a viral coat so all the viral genes are gone, and you can put in, say, a human gene or you can have nonviral delivery, but the important thing is that you’re delivering it either inside of the human or you’re taking cells out of the human and putting the DNA in and then putting them back in. But you can do very powerful things like curing inherited diseases, curing infectious diseases.                                 

For example, you can edit out the receptor for the HIV virus and cure AIDS patients in a way that's not dependent upon vaccines and multidrug resistance, which has plagued the HIV AIDS story from the very beginning. You’re basically making a human being which is now augmented in a certain sense so that, unlike most humans, they are resistant to this major plague of mankind—HIV AIDS.              

There are now people walking around who are genetically modified: There are some that are resistant to AIDS because they have had their T cells, or more generally, their blood cells modified. There are children that have been cured of blindness by gene therapy. None of this is CRISPR, but it’s in the same vein. CRISPR is overtaking it very quickly and it’s drafting behind all the beautiful work that’s been done with delivery of DNA, delivery of genetic components to patients.

GEORGE CHURCH is a professor of genetics at Harvard Medical School and director of the Personal Genome Project. George Church's Edge Bio Page

THE AUGMENTED HUMAN BEING

I worry about a lot of things. I encourage people to worry about a lot of things, but worry in the sense of taking action, doing something about it and being cautious as you do something about it—doing safety engineering. Every field of engineering has a safety component, eventually. You have civil engineering, aerospace, and so forth; huge amounts of their budgets go to safety components, and biology is no exception. Certainly in pharmaceuticals a huge fraction of the budget for bringing a new drug to market is not the research and development that produces the first prototype drug; it’s all clinical trials—toxicity efficacy testing.                                 

Some of the technologies that we come up with are pretty transformative—disruptive, in a good sense. The more transformative they are, the more important it is to consider safety upfront—ideally, talk about it before it’s on the streets. The counterexample is the influenza "gain-of-function" research where they brought to influenza some pretty scary new capabilities and didn’t discuss it with anybody who could have modulated, blocked, or cautioned them in their research until they were ready to publish it, which is a little too late. It’s the sort of thing that should have been regulated before the first grant proposal was made, much less the first research or the first paper.

Some of the things that we want people to worry about, in an enabling way these days going forward, are a lot of new applications of a new technology: CRISPR. We helped invent that. January 2013 was when it was first published. The applications are even more shocking than those papers in January 2013. The papers basically said that you can do homologous recombination—precise editing—not random editing. If you want a G to switch to an A, you can do that very efficiently in human stem cells; that’s what we said.  

CRISPR was a phenomenology since 1987, but it didn’t turn into a technology until 2012-'13. It's a mechanism by which bacteria protect themselves from invading viruses by making a molecular machine that would recognize those viruses and cut their DNA, hopefully killing or at least disabling the virus. It seemed like that might be adaptable to turning it from a killing machine into an editing machine, but that transition didn’t occur, or it wasn’t public at least, until 2013. It's a molecular machine, like many enzymes, catalysts are a protein component and this one has a nucleic acid component—a relative of DNA, which is RNA. The two will then scan along your genome more or less randomly jumping around, sometimes revisiting the same site over and over that’s the wrong site. Eventually it finds the right site and when it does, it rearranges the double-stranded DNA to insert the RNA—making a triple-strand—then the protein cuts both strands. Now you’ve got a broken piece of DNA, which seems like it’s killing still not editing, but that break allows you to bring in yet another molecule, which is now your donor DNA, which now has the new sequence that you’d like to swap in. It can cause a deletion, or an insertion, or just a change of sequence with the same number of base pairs of Gs, As, Ts, and Cs; just changing the composition, the order.                                 

That’s editing. It’s just like you would edit a book or an article: you want to be precise. You don’t want to just change G to some random other thing; you want to change it to an A. And that’s the technology that we happened to demonstrate. First we moved it from bacteria into human—a huge jump—in January 2013. Then many other labs, including ours, applied it to many other organisms.                                 

Basically, there is no organism on the planet that I know of that somebody has tried that it doesn’t work in now, which is not true for every editing method. This editing method, you could say, is just a little more efficient and a lot cheaper. That makes it sound like it’s increment, but every now and then if it’s a sufficiently large increment, it’s transformative. Most people who are familiar with it are classifying it as transformative.                                 

When you look at the applications of it beyond engineering human stem cells, which we showed first thing, you can do gene therapy with greater precision; that’s the most obvious thing. A little less obvious is you can engineer agricultural species in such a way that many governments are now classifying it as not a genetically modified organism. This is a big deal, and it shouldn’t be a big deal; it should be a minor bureaucratic footnote.                                 

Because of people willfully ignoring scientific studies on safety, they draw this sharp line between genetically modified and not, especially for foods. Even the most ardent anti-GM are still pro-GM if it’s life and death; like genetically modified insulin, where you grow human insulin in bacteria. But we’ll come back to that. Those two things are the more obvious ones: human gene therapy that's more precise and efficient than ever, and agricultural.                                 

Less obvious and fewer groups working on it is gene drives, which can be used to eliminate any vector-borne disease—malaria, dengue, lime disease—as well as invasive species like rodents that are killing off precious, endangered species on hundreds of islands worldwide and mainlands. That’s gene drives.                                 

Then transplantation: going from pigs to humans. There are a million people in need of transplants, which are not limited just by incompatibility between people; there are just not enough people. Even if we were all compatible, there are not enough donors. Pigs offered that possibility, but there were two problems. One is the immune incompatibility, and the other is they had endogenous viruses. We have used CRISPR to solve both of those problems.                                 

Then there is ecosystem manipulation. In addition to gene drives, you can address the isolation of a species’ elements: territories shrinking, getting divided by roads and other human artifacts—farms, and so forth, so that they become inbred. When species are inbred, they become less robust, less fertile, and that can be found by another revolution we’ve been involved in, which is "next generation sequencing," or reading the genome. You can now insert, using CRISPR, the proper more fertile and more robust version of the genes, or generate greater diversity.                                 

Some of that diversity you can bring in not only from adjacent populations that are separated by manmade structures, but also diversity separated by time. You can bring in DNA from the ancient, extinct versions of these animals, near relatives, because this amazing next generation sequencing is so inexpensive and powerful that we can reach back up to 700,000 years into the past and get accurate sequences of long-extinct species, but with potentially very valuable lessons for modern ecosystems.              

New technologies do change our perception of ourselves. It used to be new discoveries, and it still is; it’s integrating. If you have a new technology like a telescope, it can cause a discovery about where our planet sits in the universe—whether it’s at the center or not—but more and more frequently in the present, we have new technologies.

Sometimes people ask me why everybody is so worked up about applying CRISPR to the germline of humans. They’re not worked up particularly about applying it to the germline of animals. We just got approval for genetically modified salmon, and plants have been genetically modified for many years now. Even though some people will eat it and some people won’t, the fact is it’s a multibillion dollar business.                                 

Why are humans special? You could say we have the Food and Drug Administration (in multiple countries) that makes sure every new medical technology, whether it’s a medical device or pharmaceutical has to be safe and effective. It does you no good to have a drug that’s safe but does nothing, nor having one that’s very effective but kills people.                                 

What is it that makes germline manipulation of humans special? It's what you were just getting at—our perception of ourselves. If we feel that we can change any aspect of ourselves, where do you begin and where do you stop? and who sets those rules?                                 

When you’re in a more primitive phase of the technology, you don’t have to ask that question because it seems so far off. We can only make minor changes: a little nip-and-tuck, cure a few vaccines; it doesn’t fundamentally change human nature. But if you ever did get a tool where you could fundamentally change human nature to anything you wanted—any hybrid with any animal properties that you like, hybridization with your inorganic machines that’s more intimate than it is now—that changes our view of ourselves. I guess that’s why people not only want more caution than ever before, which I would concur. They want maybe so much caution that it can never happen. There are many technologies that get banned at one point or another; it’s not unusual. Railroads were banned because trains were colliding with one another, sometimes in the middle of towns.                                

There was a hue and cry about in vitro fertilization. It was pejoratively labeled "test tube babies"—some scare quotes around test tube babies as if that was intrinsically yucky and unacceptable. But in ’78 when Elizabeth Brown was born as a beautiful, healthy baby, the first real attempt brought to term, suddenly it looked like it was 100 percent successful and within a few years there were millions. There are now over 5 million test tube babies. For a while it was kind of quiet, between ’78 and now.

It came up that maybe recombinant DNA plus test tube babies would be a big thing, but that still seems very far, could be centuries off. Sometimes there are delays in technology, like where is my jetpack? or my flying car? and other times, it comes faster than we think. Next generation sequencing, if you drew trend lines and if it went the way computers go, which is pretty fast, it should have taken a little less than a century to arrive. Instead, it took a little less than a decade. Same thing with CRISPR as a way of doing genome engineering—it came out of nowhere. There was no CRISPR technology in the beginning of 2013 and three years later it’s everywhere; it’s all over the place. And it’s quite plausible that we can do genome engineering.              

The reason that some of these technologies arrive so far ahead of expectations is a mixture of obsessive pursuit by engineers and serendipity, and with next generation sequencing, a little of each. My group has been involved in almost every different way of doing next gen sequencing, ranging from nanopore sequencing, to fluorescent sequencing, to electronic sequencing like Ion Torrent. Similarly, we’ve been involved in every way of doing genome engineering starting with recombinant DNA, with endogenous homologous recombination, so-called meganucleases, which one of my mentors as a graduate student, Bernard Dujon, discovered was not only the first way of cutting DNA (CRISPR is now the most recent), it was also the first gene drive that he discovered in the early 1980s.                                  

Bernard Dujon was one of my co-mentors during my PhD. He discovered the first meganuclease, the first very specific cleavage, where it would cleave once in a genome. After that, we were involved in the zinc fingers with Aaron Klug, who started a little company called Gendaq, which later became part of Sangamo, and then TAL proteins were even easier to program than meganucleases or zinc fingers. We were involved in that, including one of my talented post docs named Feng Zhang; he and I worked together on TALENs, and then he started up his own lab at MIT. Then we both started working on this new phenomenon called CRISPR. It had been in the literature since 1987, but it only recently started to dawn on groups, mainly biochemists like Jennifer Doudna and Emmanuelle Charpentier, how you would go about harnessing this as a technology.                                 

There are a number of people before Jennifer and Emmanuelle, Feng, and my own labs that had established that it was important in resisting viruses and bacteria; but turning it into a technology of gene editing was not obvious and, in fact, had a good chance of failure. For example, we had been trying to adapt another genome engineering strategy which we called MAGE, some other people called it recombineering, but that, after decades of work, only works well in E. coli. We have not been able to transfer it to other organisms. CRISPR, on the other hand, serendipitously works in every organism without major modifications, so it's just tweaking how the component molecules are produced.                                 

Feng Zhang and one of my graduate students Le Cong published a paper (Le was the first author) in January 2013, side-by-side with the one that I did with Prashant Mali (first author), who is now at UCSD, and Luhan Yang, who is one of the key players in the xenotransplantation work of applying CRISPR to getting organs from pigs. We deposited the molecular tools to do this in a nonprofit called Addgene, and that just caused it to spread even faster. People realized that for $50 they could get in the game, and of all the technologies that I've helped develop, this was probably the easiest one for people to adopt.                                 

Next gen sequencing, you had a big million dollar machine, even in the very early days to practice it. This, you just needed whatever you already had plus this little plasmid you get from this nonprofit, Addgene. It just spread and the rest is history.              

Getting back to why people should worry about it, there are these very powerful applications which is like somatic gene therapy in adults, which people should be worried about.              

There are now 2000 gene therapies where you’ll take a little piece of engineered DNA, put it inside of a viral coat so all the viral genes are gone, and you can put in, say, a human gene or you can have nonviral delivery, but the important thing is that you’re delivering it either inside of the human or you’re taking cells out of the human and putting the DNA in and then putting them back in. But you can do very powerful things like curing inherited diseases, curing infectious diseases.                                 

For example, you can edit out the receptor for the HIV virus and cure AIDS patients in a way that's not dependent upon vaccines and multidrug resistance, which has plagued the HIV AIDS story from the very beginning. You’re basically making a human being which is now augmented in a certain sense so that, unlike most humans, they are resistant to this major plague of mankind—HIV AIDS.              

There are now people walking around who are genetically modified. There are some that are resistant to AIDS because they have had their T cells, or more generally, their blood cells modified. There are children that have been cured of blindness by gene therapy. None of this is CRISPR, but it’s in the same vein. CRISPR is overtaking it very quickly and it’s drafting behind all the beautiful work that’s been done with delivery of DNA, delivery of genetic components to patients.

There are 2000 gene therapies of various sorts in clinical trials in Phase 1, 2, and 3. Typically Phase 1 is about toxicity, Phase 2 is about efficacy, and Phase 3 is when you say it's ready for market. One has been approved in Europe by the EMA, not in America yet. It’s called Glybera and it’s for a pancreatitis, which is a lipid disorder. It is the most expensive drug in history—over a million dollars per dose. In principle, it could be one dose per lifetime. So, it’s comparable to orphan drugs, which are rare disease treatments. We hope that we can bring down the cost, either by going for larger markets, so some of the diseases that are aimed at are not rare diseases, they're more common ones, these are infectious diseases.

The Orphan Drug Act is a way of encouraging the industry to go after these very rare diseases by permitting them to charge the correct amount to compensate for the drug development costs and profit and get reimbursed for it. There are handfuls of orphan drugs. In fact at this point, even though these are very rare—some of these of these diseases are one in 1000, one in 10,000, one in 100,000—they're nevertheless gaining profitability, which is on the order of 30 percent of the entire pharmaceutical industry.                                 

Nobody is sure whether this is sustainable. If we get to the point where we have custom drugs, every person has their own drug, and each of those drugs costs half a billion dollars to produce, then we’re broke. But if we get better at it and, in fact, if we have a class of drugs that work well, where you can use smaller cohorts—by definition, rare drugs are using small cohorts, sometimes they’re so small you never even make it to Phase 3.                                 

For example, some of these hereditary blindnesses, if you gather up everybody in the world you can find that is willing to participate and you cure them all, or most of them, you can’t move on to [the next] phase, because in the toxicity trial you proved efficacy, you cured them, and there is nobody left to cure. That’s a pretty high quality, first world problem.                                 

That’s all in adults and children. Some of these you have to do in children because the brain is wired up at an early age. If you cure blindness late in life, you can make it so the patients can see the photons—can see that there is something going on that they didn’t see before, but they can’t process it. They can’t say, "Oh, that’s a face or that’s a person," it’s just these blurry, or even sharp but uninterpretable images. If you do it early enough, then they can develop close to fully functional vision.              

You can imagine some developmental diseases, especially those that affect the brain, you might have to do it as a fetus. There is already fetal surgery that is done to catch things, but if you have to tinker with the molecules, then you might want to do it as a fetus, or you might want to do it even earlier. Then you get into this whole business of whether you can do the germline or not, and society doesn’t know where to draw the line.                                 

Is the line drawn germline or is it drawn at the level of medical significance. In other words, if it’s very medically significant it doesn't matter whether it’s in an adult or a child or a fetus or an embryo or a sperm or egg, if the positive impact is high enough and there is no alternative. But sometimes people like to draw lines with a buzzword rather than with medical. An example is golden rice, where the line, I would think, should be drawn between medical versus nonmedical foods.

Most GM foods, I have to agree with the GM critics, don’t get the average person anything that they can recognize. You go to the supermarket and there’s not that big a difference between the GM foods and the non-GM foods, in terms of price or quality or anything like that. So why not object to them, you know? Anything that you can label, you can start an ad campaign that’s effective, you can say potato chips without cholesterol. They never did have cholesterol, but now the labeling sells a few more bags of them. That makes sense.                                 

Golden rice was a tough call strategically for Greenpeace and some of their associates. They could have classified that as medical, like the way that you would classify GM insulin. It’s very powerful and a million lives are at stake every year due to vitamin A deficiency, and golden rice was basically ready for use in 2002, so it’s been thirteen years that it’s been ready. Every year that you delay it, that’s another million people dead. That's mass murder on a high scale. In fact, as I understand it there is an effort to bring them to trial at the Hague for crimes against humanity.                                 

Maybe that’s justified, maybe it isn’t. The fact is we have a pretty good way of addressing this vitamin A deficiency, and nothing else has worked during those intervening thirteen years. It’s hard to get them pills because that’s very expensive. These are people who could barely afford rice as their sole source of calories; how are they going to afford medicine?

Greenpeace is very well funded. They have many, many times more funding than the groups that are developing the golden rice and they can lobby the governments to say it’s not safe. They can demand higher and higher levels of safety testing. Then, when the safety testing starts to look good, they can go in and trash the plots of land that’s growing the golden rice through vandalism—as happened in the Philippines recently—and then say, "Well, where is the safety data?" It was destroyed by the vandals. There are many ways that you can block development or approval of something that’s quite clearly safe.                                 

The same thing can happen with the human germline. You could just treat it the same as other medical technologies: artificial limbs, new pharmaceuticals, and so forth; they have to be shown to be safe and effective. Or you could draw a line saying if it has the word germline modification in it, then no matter how safe it is, we’re going to take a less-safe method or nothing, rather than have this scare word involved. Once it’s proven to be safe and effective, it will be like in vitro fertilization, it will be very difficult to deprive people of the technology that is going to help their children.              

Typically in medicine, it’s not relative to nothing, it’s relative to various other technologies. In the case of genome editing or gene therapy, more broadly, the alternatives are things like genetic counseling.                                 

The parallel revolution of next generation sequencing provides an alternative for next generation genome editing, in this case, because as the price has plummeted from $3 billion to now less than $1000, complete with interpretation and genetic counseling, you can stop many of these diseases way before you need a million dollar gene therapy: You can prioritize who you’re going to date, if you have a particularly deleterious carrier status. Let’s say you have Tay-Sachs carrier status, which is a disease that when you have two copies—one from each parent—the child is very disabled and usually dies a painful death shortly after birth. It's something that most families that know that they’re at risk try to avoid. But there are many people who don’t know they’re at risk until they get the first child.                                 

The solution is for everybody to know their genome, and to not use it as a way of wasting money on hypochondria, but seeing if there is something very well characterized in medicine, something that’s highly predictive and actionable by deciding who you’re going to date and then marry. Or it can be done later with slightly more medical and psychosocial inconvenience at the stage of prenatal testing, including a revolution that’s coming in part because of next generation sequencing called noninvasive prenatal testing where you can test the fetus’ genes by getting a little bit of blood from the mother.                                 

The cost of genetic counseling is about $1000, which can save you a million dollars later for either orphan drugs or gene therapy. All these prices will drop. There is not a developed gene therapy that’s been approved or an orphan drug, and the anxiety of hunting down and figuring out what the genetic problem is after you’ve already got symptoms in a child, like developmental delay is not worth it, even if the prices were identical.

You need to temper the enthusiasm and concern for these new technologies with the alternative technologies, for example, genetic counseling.                                 

Oh, I forgot to mention aging reversal. This is a big project both in my lab and in one of our startup companies. This is not about wellness or drugs that affect diseases of aging, which are effects rather than causes; it’s trying to get at the causes of aging and reverse them. And there are a fair number of precedents for this in animals, but the idea is to get it transferred to humans.

Reversal of aging: Some examples of this are if you take blood from a young mouse and exchange it with an old mouse. The small molecules, macromolecules, and cells in the blood result in a variety of biomarkers of aging being reversed. You can affect the vasculature, the blood vessels, the nerves, skeletal and cardiac muscles, and there are measures of these that indicate that it’s not just prolonging a very aged state or going for longevity; you’re actually reversing it.                                 

This is a much better target, in any case, than prolonging longevity because, A, it takes years to decades to even prove that you have extended longevity. Also, if you’ve done it on somebody that’s quite old, the economic consequences are dire; that’s the part of your life where you spend huge amounts on medicine and don’t improve the quality of life tremendously. If you can reverse it to an age where you essentially don’t use any medicine, this will be much more cost effective.

The reason that an academic would help start so many companies is it is a way of accelerating the pace at which technology is transferred from an idea to an experiment, like that might get published in a paper that has no impact on society other than scientists that read it, from a paper to a company, from a company to the market.                                 

For example, the next generation sequencing technologies—quite a few of them now—are starting to have impact in things like noninvasive prenatal testing. Some have said that this is the fastest-growing medical diagnostic in history. It’s gone from basically nonexistent three or four years ago, to now millions of women that have been tested—mainly for trisomy—things that can cause fetal wastage or severe medical problems later in life.              

One of the issues with the transition from academic work to corporate work that’s needed to get it to have an impact on society are questions about scale and secrecy. Some of the practices in academia and small startups, and even large companies, are aligned—some of them are motivated to publish in the same journals, sometimes side-by-side, because a high quality, peer-reviewed article is a magnet for attracting some of the best scientists to come join you. If you do everything in total secrecy, the very best scientists don’t know you exist, they don’t care deeply. Publishing can happen in similar ways in companies and academia.                                 

Another aspect of secrecy is without patents there would be a lot more secrecy than there is, but the patent system is a way of tantalizing and encouraging companies to put their ideas and their inventions in the public domain. They give them a limited monopoly for twenty-some years, but only if they describe in enough detail that it is considered "enabling."                                   

Enabling technologies when they’re patented, everybody can then build on top of them. Some people claim that there still will be trade secrets but not nearly as much. If you keep it secret and somebody else patents it, they win because your secrets are not longer valuable to you, because they can sue you for infringing their patent, even if you did it first. Worldwide, it’s first to file, not first to invent, so the fact that you’ve been keeping it a secret all this time does you no good anymore in the United States. That’s a fairly recent development.                                 

You have to file and that was the whole point. Some people say that patents are evil somehow, but the alternative is worse—trade secrets, which is what would be rampant otherwise. Also, the patents don’t stop people from inventing; to some extent, they help us to invent because we can look at the patent and say, "Oh, that’s how they did it," and you can build on top of it—stand on the shoulders of giants.                                 

The only thing it prevents you from doing is doing the same thing that somebody else did and then making money without giving them a cut; that’s what’s banned. Some people say, "Well, why would I want to invent on top of somebody else’s technology because I’m just going to have to give it all to them?"                                 

You don’t have to give it all to them; you can cross-license, you can make a deal where they give you something, you give them something, or where you’re both enabled. It’s case-by-case and it depends on how much value you add. If you add enough value, you could get more than the original patent, which might have been very preliminary and limited.                                  

Revive and Restore is a spinoff of the Long Now Foundation. Ryan Phelan and Stewart Brand have been championing the merging of some of the most cutting-edge molecular methods like CRISPR and next generation sequencing with very significant needs in the ecological, environmental conservation movements.                                 

We have islands that have many of the world’s most diverse and beautiful species of birds, reptiles, amphibia, and so forth, and they’re endangered by invasive species and new diseases. Even a few degrees change in temperature can shift the ecosystems faster than the species can adapt. We could just let nature take its course, even though humans are influencing it, or we can use these powerful molecular methods to track, diagnose, and then to take action with gene drives or with preparing vaccines. (Gene drives being a kind of vaccination that spreads itself.) Or we can make synthetic viruses, so we can accelerate the processes in CRISPR vaccines to fight them, as we’re doing for the herpes virus in elephants.

Revive and Restore is largely about conserving important ecosystems, not just important in some abstract sense, but sometimes in a very practical, human-centered way.                                

For example, in the tundra of Canada, Siberia, Russia, and Alaska, the permafrost, which is supposedly permanent, is melting, and with it could be released as much as two times the carbon of all the forests all over the world, as if they all burned and released their carbon dioxide and methane into the air, causing global warming. Some of the environmental studies, for example from the Zimov group, have shown that if you can restore some of the keystone species that were there not too long ago, that have very recently left or gone extinct, you can lower the temperature by 15 to 20 degrees, which could greatly delay or maybe even reverse the process of permafrost thawing.              

Some of the questions that come up with Revive and Restore, of using cutting edge molecular technologies for ecosystem conservation and preservation, some of the same questions come up as come up with using these molecular technologies in medicine, which is who gets to choose? who decides? Are people not being heard or not being invited to sit to talk?                                 

In all these cases, both medical and environmental, there is, maybe more than ever before, an effort to engage all sorts of citizens, regular folks that know no science at least before they start the conversation, patient advocates, environmental advocates, scientists, ethicists, lawyers, politicians, and so forth—many of these meetings are presented as video conference on the web, webcasts real-time, and Twitter questions are encouraged. It doesn’t mean it’s perfect, but at least it’s very different from, say, the Manhattan Project where the public was not invited to weigh in on whether we should, for example, take a chance that atmospheric testing would ignite the atmosphere. That was one of the things that physicists tried to calculate on their own, without anybody meddling. If it had ignited the atmosphere, that would not have been highly protective to our nation or the world. Here, though, the idea is to get as many people thinking about the unintended consequences as possible, and to proceed cautiously.              

The question of who decides ultimately with these kind of transparent and open projects, where it’s not being done in secret like the Manhattan Project, is —society decides. We vote with our wallets, we vote with the free enterprise system, with our politics, the power of the pen, and in some cases, we may change our mind later. There's an emphasis on things that are reversible—those get higher priority.                                 

But eventually, we do irreversible things. Certainly, it’d be very hard at this point to surgically remove automobiles from our lives, even though they kill a million people per year. We couldn’t just delete them. It would be very hard to go back to previous agricultural methods because the early pre-agricultural society could not possibly sustain 7 billion people. In fact, probably even just a couple of generations back in crops could not sustain 7 billion people.                                 

Who decides whether particular species come back? Does it help human beings? It’s a very species-chauvinistic way of looking at it, but if the mammoth can lower the temperature of the permafrost by 15 to 20 degrees and we don’t have a particularly better way of doing it, let’s say, with motorized versions of mammoths, then we might do it. Hopefully, it will involve many countries making this decision maybe at the United Nations level.                                 

Unlike the vaccines where you have medical professionals going door to door, essentially village to village, gene drives and mosquitoes do it themselves, and you could spread it throughout sub-Saharan Africa. They don’t respect borders or wars or any other things that would hold back medical professionals. You probably want to get the buy-in of all the nations, not just the ones that are most desperately in need of the gene drives to eliminate malaria.              

It’s an extraordinarily exciting time for scientists, in particular, those involved in reading and writing genomes. It should be an exciting time for everybody and also a scary time for scientists and everybody, where an increasing number of decisions of politicians, CEOs, and regular citizens depend on some technical nuance and expertise. We can no longer say, "Oh, well, I’m just going to vote with my party." Don’t complain about being excluded from a discussion if you’re excluding yourself from that discussion. That’s the power of the community of intellectuals that are trying to reach out to everybody in the world. It’s not intended to be an exclusive club. It’s intended to be a conversation.