Uri Alon
1,645,104 views • 15:50

Chris Anderson: So our first speaker gave a TED Talk at TEDGlobal I think seven years ago. His name is Professor Uri Alon, at the Weizmann Institute of Science. Now, he and his colleagues there have come up with a powerful idea that addresses this key question: How on earth do we get back to work without creating a second surge of the infection? Uri Alon, welcome to TED.

Uri Alon: Thank you. Nice to be here again.

CA: It's great to see you again. So, I guess the key to your idea is this obsession with the reproduction number R, R-naught. If that number is less than one, then fewer than one person is infected by a typical person, and eventually, the epidemic fades away. People are worried that as we come back to work, R will shoot up above one again. You have a suggestion for how we might avoid that. What is that suggestion?

UA: Exactly. So, we are suggesting a strategy that's based on a weak spot based on the biology of the virus, which is a cycle of work and lockdown. It exploits the vulnerability of the virus in that, when a person gets infected, they're not infectious for about three days. So you don't infect others for the first three days, and after another two days, on average, you get symptoms. So we're proposing a strategy which is four days of work and then 10 days of lockdown, and the next two weeks, again: four days of work, 10 days of lockdown. And that way, if a person gets infected at work, they reach their peak infectiousness during lockdown, and that way, they avoid infecting many others. This restricts the viral transmission. Also, just working four days out of two weeks restricts the amount of time the virus gets to see many other people, and that's a very powerful effect. So everybody works on the same four days, kids go to school on the same four days, with all the measures of social distancing and masks, etc, and then there's a lockdown period.

CA: So if you take the worst-case scenario, where you come to work on a Monday morning at the start of your four days, and you're infected on the subway, say, on the way to work, the theory here is that even by the end of that four days, you're not really starting to infect your coworkers?

UA: That's correct. So you're infected on the subway, and so for the first three days or so, you're in your latent period, you don't infect your coworkers, you reach your peak infectiousness at home, there will be secondary infections at home, and people with symptoms can self-quarantine, and over the long run, you have a reproduction number less than one, so the epidemic, if you continue these cycles, will go away.

CA: I mean, is it frustrating at the thought that people are going to say, "Wait — I don't want to infect people at home, I'd rather infect people at work than at home." What's the response to that?

UA: Yes, absolutely. So we have to consider the alternatives. If you open up the economy and there's a second wave, you'll get all those infections anyway during the lockdown that happens, along with the devastating effects on the economy, etc. And so, in the long run, if you do a cyclic strategy like this but with a reproduction number that's less than one, you avoid, at least with these mathematical models and considerations, the much larger number of infections you'd get if there's a second wave.

CA: Right. You're serving the needs of your family by — sorry, go on.

UA: Even people who are infected don't infect everyone at home. The attack rates are 10 to 30 percent, according to several studies.

CA: Right. But the hope is that you're serving the needs of your family by engaging in a strategy where very few of your fellow workers are going to be infectious anyway, so that's the plan, but um —

UA: That's right.

CA: Tell me this, though — because four days out of 14, someone's going to say, "Well, great idea, but that implies, like, a 70 percent loss of productivity in the economy, so that can't possibly work." I think you think that the productivity loss need not be anything like that much.

UA: That's right, and of course, most people don't work weekends, so it's four days out of the 10 work days in the two weeks, and once you have a predictable schedule of four days at work, you can work longer hours, you can design shifts and get higher productivity by prioritizing in those four days much more than 40 percent of the workdays.

CA: Yes, so talk through how that could work. I mean, let's imagine, first of all, manufacturing, which is currently shut down. Is the implication here that a manufacturer could set up two, possibly even three, shifts of four days, maybe 35 hours or something of work over those four days and still get a lot of productivity, basically, having the lines almost running continuously that way?

UA: Exactly. So this is a staggered version of this idea, where you take the population, divide it into two groups or three groups. Let's say one group works four days and then 10 days of lockdown. Then the other group kicks in. This idea was proposed by colleagues at Bar-Ilan University. Then you get an added benefit that during workdays there's less density. If there's two groups, there's half the density and less transmission. And you can keep production lines working almost continuously like that using this staggered idea.

CA: And applying it to thinking about offices coming back — I mean, it seems to me that, as we've already seen, there's a lot of productivity that can happen when you're at home, so you could picture on this idea of people doing one set of things during the four days when they're, say, back at the office, doing the exposure to each other, sparking off each other, the discussions, the brainstorming, all that good stuff, while at home, they're then doing all the things that we've been doing the last few weeks, kind of working solo. How much have you thought about how that, whether it's possible, effectively, to divide work into different types and actually use a strategy like this to maintain almost full or even better productivity?

UA: I agree — for many sectors, people work at home very effectively, and we've heard from several industries that productivity actually went up during lockdown and people working at home. So if you have a schedule, a [cyclic exit strategy] you can restrict the amount, or you can plan the work where you need to be together in a very effective way with avoiding a lot of time lost, if the person's work can be more effective at home and more effective at work and get high productivity. I should say that some sectors really need to adjust, like hotels, tourism, dining. In several industries, this will require more thought and adjusting. But other industries are almost built for ideas like this. Maybe it's even something you can consider after the epidemic, because productivity can be at least as high.

CA: I mean, I read this and I started thinking about our own organization, TED, and how, in many ways, you could argue that could work really well. I mean, for one thing, there's this question about extroverts and introverts. Some introverts, if they were honest, might say that this pandemic has been manna from heaven for them. They've found work less stressful. They've been able to focus and so forth. With this sort of four days on, four days off type strategy, perhaps you can imagine a work world that's optimized for both introverts and extroverts?

UA: Absolutely. I mean, I feel it also. Me and my partner, with different personalities, we both teach in universities, and teaching through this has [helped me] become productive in certain ways. So I agree completely, and I think harnessing the creativity of people at workplaces, we're only at the beginning of what these kinds of mixtures can offer.

CA: But for people who are on the front line, again, if you're delivering goods and so forth and you can't do that virtually, is there any thought about how a four days on and then isolation strategy, how that off time could be used to nonetheless contribute to that person's work through some form of training? Or is it more just that people would work very intensely during four days, and maybe people still aren't quite earning their full pay in this scenario, but it's better than complete lockdown, and it's better than going back to work and seeing another surge?

UA: That's right. So on a society level, it's better than opening up and seeing another surge, which would require complete lockdown. For people like hospital shifts, some hospitals adopted this kind of program so we can protect shifts and avoid mixing. It also creates a lot of simplicity and clarity. So you understand when you're working, and you have some confidence because this is based on scientific modeling about the effectiveness of this plan. It's also equitable in the sense that everybody gets to go to work, not only certain sectors, it's transparent, etc.

[Cross talk]

CA: And this is something that is best implemented by individual companies? Or is it actually much better implemented a city at a time or even a nation at a time?

UA: We think it can work [in levels]. So at certain companies, it's very natural to adopt, or at hospitals, schools, etc. It can also work at the level of a town or a region, and then we would advise trying it out for something like a month, seeing whether cases rise. In that case, you can dial down the number of workdays. Or, if cases are declining quickly, you can add workdays and therefore adapt to the climate and the location where a person is. So it's quite adaptable.

CA: But by aligning work schedules with schools, for example, that suddenly allows parents to go back to work on the days that their kids are at school, and you'd have to try —

UA: Absolutely.

CA: I mean, is the best instantiation of this that countries literally divide households into different A and B categories, or something like that, so that that kind of alignment could happen?

UA: Exactly. So you can align different households, Group A and Group B, and then the children go to school, the parents go to work in a synchronized way, and the other group, let's say, the alternating weeks. A certain amount of people need to work all the time. Maybe teachers are, like, essential workers and need to work throughout. Just like during lockdown situations, a certain fraction of the population still works throughout. But a region that does this should be protected, in a sense, because it has a replication number of less than one, so imported infections also can't spread very much.

CA: And here is the aforementioned David Biello. David.

David Biello: Yes. Hello, everybody. Uri, as you can imagine, there are lot of questions from the audience, and we have a first one kind of about those workers who have been marked as essential. Can you comment on how this would impact the health care professionals and others who may not have time or the flexibility to quarantine in the way you suggest.

UA: That's great. I want to say that there's essential workers, there's people with low income, that just can't adhere to lockdown because they have to make a living. And studies show that mobility [among] people in the low-income sectors is larger during lockdown. And also, in developing countries, people just have to go out of the house. You can't enforce lockdown. So this four-10 kind of strategy can actually make lockdown easier to bear for people who can still make a living during those days, or at least make their own choices about what fraction to work and what fraction to stay in lockdown. Some countries can't get R less than one even with lockdown, because of this adherence problem, because of informal sectors, etc. We believe that a four-10 cycle might make it easier to do lockdown and maybe get our infection level less than one. That affects billions of people in the world. I hope I answered your question.

DB: I think so, and we have another question, I believe, if that can be queued up, which is: Any chance you can do the math and quantify the increased risk of this four-10 cycle?

UA: So the increased risk, we're saying in our scientific paper, we did all the sensitivity analyses, etc, and the question is, it's comparing increased risk comparing to what? So, to the economy ... It's possible there will be a second wave. I mean, I hope there won't be, but it certainly is possible, and in that case, it's clear that a second wave and another lockdown will have worse consequences on health than a cycle of four-10. And so it's really a question of what you're comparing to.

DB: Sure. Well, thank you so much for sharing this idea, Uri.

CA: Indeed. David, stay on. But just before you go: Have any governments expressed interest in exploring this? Do you see people considering actually implementing this as national policy?

UA: Yes, we're in touch with several European countries and countries in South America and Israel, of course. Austria has adopted a similar program for their school system, which is five school days every two weeks. And several companies and hospitals, etc. And so we're very interested to see how this will play out.

CA: Well, I love the basic start point of starting by looking at the enemy's weakness. And you've got this four-day period where it's not necessarily that dangerous after an infection, if you could figure out a way to work then. I assume that testing would actually enhance this idea as well a lot, right? To test people before they come back —

UA: It's not predicated on testing. You don't have to have testing for this idea, so that can apply to regions without a lot of testing. If you do have testing, it could help you use testing in a more effective way by concentrating testing on people at the end of their 10 lockdown days, just as they're about to go to work; that could make each test more impactful in terms of reducing their reproduction number.

CA: Indeed, instead of having to test the whole population every three or four days, it's just once every two weeks. That's a much more imaginable goal.

UA: Sure.

CA: Yeah. Well, Uri Alon, thank you so much for spending this time.