0:00:00.558,0:00:04.139 Chris Anderson: So our first speaker[br]gave a TED Talk at TEDGlobal 0:00:04.163,0:00:06.388 I think seven years ago. 0:00:06.412,0:00:09.353 His name is Professor Uri Alon, 0:00:09.377,0:00:12.034 at the Weizmann Institute of Science. 0:00:12.058,0:00:16.397 Now, he and his colleagues there[br]have come up with a powerful idea 0:00:16.421,0:00:18.415 that addresses this key question: 0:00:18.439,0:00:21.641 How on earth do we get back to work 0:00:21.665,0:00:25.458 without creating a second surge[br]of the infection? 0:00:26.658,0:00:29.353 Uri Alon, welcome to TED. 0:00:30.179,0:00:33.652 Uri Alon: Thank you.[br]Nice to be here again. 0:00:33.676,0:00:35.866 CA: It's great to see you again. 0:00:35.890,0:00:39.596 So, I guess the key to your idea 0:00:39.620,0:00:43.755 is this obsession with[br]the reproduction number R, R-naught. 0:00:43.779,0:00:46.288 If that number is less than one, 0:00:46.312,0:00:49.870 then fewer than one person[br]is infected by a typical person, 0:00:49.894,0:00:53.633 and eventually, the epidemic fades away. 0:00:53.657,0:00:55.928 People are worried that[br]as we come back to work, 0:00:55.952,0:00:58.885 R will shoot up above one again. 0:00:58.909,0:01:01.330 You have a suggestion[br]for how we might avoid that. 0:01:01.354,0:01:03.249 What is that suggestion? 0:01:05.115,0:01:06.267 UA: Exactly. 0:01:06.291,0:01:08.088 So, we are suggesting a strategy 0:01:08.112,0:01:12.550 that's based on a weak spot[br]based on the biology of the virus, 0:01:12.574,0:01:15.504 which is a cycle of work and lockdown. 0:01:15.926,0:01:21.026 It exploits the vulnerability of the virus[br]in that, when a person gets infected, 0:01:21.050,0:01:24.111 they're not infectious[br]for about three days. 0:01:24.135,0:01:26.592 So you don't infect others[br]for the first three days, 0:01:26.616,0:01:29.316 and after another two days,[br]on average, you get symptoms. 0:01:30.297,0:01:33.815 So we're proposing a strategy[br]which is four days of work 0:01:33.839,0:01:36.056 and then 10 days of lockdown, 0:01:36.080,0:01:39.374 and the next two weeks, again:[br]four days of work, 10 days of lockdown. 0:01:39.398,0:01:42.731 And that way, if a person[br]gets infected at work, 0:01:42.755,0:01:46.409 they reach their peak infectiousness[br]during lockdown, and that way, 0:01:46.433,0:01:48.737 they avoid infecting many others. 0:01:49.374,0:01:52.488 This restricts the viral transmission. 0:01:52.512,0:01:55.066 Also, just working four days[br]out of two weeks 0:01:55.090,0:01:58.399 restricts the amount of time[br]the virus gets to see many other people, 0:01:58.423,0:02:00.066 and that's a very powerful effect. 0:02:00.090,0:02:02.128 So everybody works on the same four days, 0:02:02.152,0:02:04.336 kids go to school on the same four days, 0:02:04.360,0:02:09.047 with all the measures[br]of social distancing and masks, etc, 0:02:09.071,0:02:11.115 and then there's a lockdown period. 0:02:12.401,0:02:14.893 CA: So if you take[br]the worst-case scenario, 0:02:14.917,0:02:18.479 where you come to work on a Monday morning[br]at the start of your four days, 0:02:18.503,0:02:22.155 and you're infected on the subway,[br]say, on the way to work, 0:02:22.179,0:02:25.645 the theory here is that even[br]by the end of that four days, 0:02:25.669,0:02:29.174 you're not really starting[br]to infect your coworkers? 0:02:30.339,0:02:31.493 UA: That's correct. 0:02:31.517,0:02:33.105 So you're infected on the subway, 0:02:33.129,0:02:36.369 and so for the first three days or so,[br]you're in your latent period, 0:02:36.393,0:02:37.948 you don't infect your coworkers, 0:02:37.972,0:02:40.733 you reach your peak[br]infectiousness at home, 0:02:40.757,0:02:43.086 there will be secondary[br]infections at home, 0:02:43.110,0:02:46.686 and people with symptoms[br]can self-quarantine, 0:02:46.710,0:02:50.383 and over the long run, you have[br]a reproduction number less than one, 0:02:50.407,0:02:53.121 so the epidemic,[br]if you continue these cycles, 0:02:53.145,0:02:55.304 will go away. 0:02:57.420,0:03:01.174 CA: I mean, is it frustrating 0:03:01.198,0:03:03.298 at the thought that people[br]are going to say, 0:03:03.322,0:03:05.539 "Wait -- I don't want[br]to infect people at home, 0:03:05.563,0:03:07.779 I'd rather infect people[br]at work than at home." 0:03:07.803,0:03:09.987 What's the response to that? 0:03:10.011,0:03:11.189 UA: Yes, absolutely. 0:03:11.213,0:03:13.134 So we have to consider the alternatives. 0:03:13.158,0:03:15.654 If you open up the economy[br]and there's a second wave, 0:03:15.678,0:03:19.271 you'll get all those infections anyway[br]during the lockdown that happens, 0:03:19.295,0:03:22.466 along with the devastating effects[br]on the economy, etc. 0:03:22.490,0:03:24.484 And so, in the long run, 0:03:24.508,0:03:26.313 if you do a cyclic strategy like this 0:03:26.337,0:03:29.655 but with a reproduction[br]number that's less than one, 0:03:29.679,0:03:34.100 you avoid, at least with these[br]mathematical models and considerations, 0:03:34.124,0:03:38.087 the much larger number of infections[br]you'd get if there's a second wave. 0:03:38.989,0:03:43.874 CA: Right. You're serving the needs[br]of your family by -- sorry, go on. 0:03:44.429,0:03:47.800 UA: Even people who are infected[br]don't infect everyone at home. 0:03:47.824,0:03:51.743 The attack rates are 10 to 30 percent,[br]according to several studies. 0:03:52.869,0:03:54.024 CA: Right. 0:03:54.048,0:03:56.899 But the hope is that you're[br]serving the needs of your family 0:03:56.923,0:03:59.937 by engaging in a strategy[br]where very few of your fellow workers 0:03:59.961,0:04:01.847 are going to be infectious anyway, 0:04:01.871,0:04:04.217 so that's the plan, but um -- 0:04:04.241,0:04:06.189 UA: That's right. 0:04:06.832,0:04:09.537 CA: Tell me this, though --[br]because four days out of 14, 0:04:09.561,0:04:11.646 someone's going to say, "Well, great idea, 0:04:11.670,0:04:15.023 but that implies, like,[br]a 70 percent loss of productivity 0:04:15.047,0:04:16.238 in the economy, 0:04:16.262,0:04:18.091 so that can't possibly work." 0:04:18.115,0:04:20.194 I think you think that[br]the productivity loss 0:04:20.218,0:04:22.268 need not be anything like that much. 0:04:23.059,0:04:24.313 UA: That's right, 0:04:24.337,0:04:26.582 and of course, most people[br]don't work weekends, 0:04:26.606,0:04:29.396 so it's four days out of[br]the 10 work days in the two weeks, 0:04:29.420,0:04:32.516 and once you have[br]a predictable schedule 0:04:32.540,0:04:33.696 of four days at work, 0:04:33.720,0:04:35.431 you can work longer hours, 0:04:35.455,0:04:37.936 you can design shifts[br]and get higher productivity 0:04:37.960,0:04:39.636 by prioritizing in those four days 0:04:39.660,0:04:42.482 much more than 40 percent of the workdays. 0:04:44.119,0:04:46.239 CA: Yes, so talk through[br]how that could work. 0:04:46.263,0:04:49.984 I mean, let's imagine,[br]first of all, manufacturing, 0:04:50.008,0:04:52.018 which is currently shut down. 0:04:52.042,0:04:56.721 Is the implication here[br]that a manufacturer could set up 0:04:56.745,0:04:59.918 two, possibly even three[br]shifts of four days, 0:04:59.942,0:05:04.601 maybe 35 hours or something of work[br]over those four days 0:05:04.625,0:05:09.054 and still get a lot of productivity, 0:05:09.078,0:05:12.607 basically, having the lines[br]almost running continuously that way? 0:05:13.418,0:05:14.569 UA: Exactly. 0:05:14.593,0:05:18.041 So this is a staggered[br]version of this idea, 0:05:18.065,0:05:22.099 where you take the population,[br]divide it into two groups or three groups. 0:05:22.123,0:05:26.267 Let's say one group works four days[br]and then 10 days of lockdown. 0:05:26.291,0:05:28.568 Then the other group kicks in. 0:05:28.592,0:05:32.173 This idea was proposed by colleagues[br]at Bar-Ilan University. 0:05:32.197,0:05:36.395 Then you get added benefit[br]that during workdays there's less density. 0:05:36.419,0:05:37.583 There's two groups. 0:05:37.607,0:05:39.810 There's half the density[br]and less transmission. 0:05:39.834,0:05:44.611 And you can keep production lines[br]working almost continuously like that 0:05:44.635,0:05:46.477 using this staggered idea. 0:05:49.199,0:05:53.789 CA: And applying it to thinking[br]about offices coming back -- 0:05:53.813,0:05:56.523 I mean, it seems to me that,[br]as we've already seen, 0:05:56.547,0:06:00.359 there's a lot of productivity[br]that can happen when you're at home, 0:06:00.383,0:06:03.917 so you could picture on this idea[br]of people doing one set of things 0:06:03.941,0:06:06.738 during the four days when they're,[br]say, back at the office, 0:06:06.762,0:06:11.337 doing the exposure to each other,[br]sparking off each other, 0:06:11.361,0:06:14.944 the discussions, the brainstorming,[br]all that good stuff, 0:06:14.968,0:06:18.134 while at home, they're then[br]doing all the things 0:06:18.158,0:06:20.077 that we've been doing[br]the last few weeks, 0:06:20.101,0:06:22.130 kind of working solo. 0:06:22.154,0:06:24.288 How much have you thought about that, 0:06:24.312,0:06:28.701 whether it's possible, effectively,[br]to divide work into different types 0:06:28.725,0:06:30.978 and actually use a strategy like this 0:06:31.002,0:06:34.247 to maintain almost full[br]or even better productivity? 0:06:34.898,0:06:38.274 UA: I agree -- for many sectors,[br]people work at home very effectively, 0:06:38.298,0:06:40.194 and we've heard from several industries 0:06:40.218,0:06:43.192 that productivity actually[br]went up during lockdown 0:06:43.216,0:06:44.604 and people working at home. 0:06:44.628,0:06:47.253 So if you have a schedule, cyclic ____???[br][cyclic exit strategy] 0:06:47.277,0:06:48.945 you can restrict the amount, 0:06:48.969,0:06:52.200 or you can plan the work[br]where you need to be together 0:06:52.224,0:06:54.235 in a very effective way 0:06:54.259,0:06:56.971 avoiding a lot of time lost, 0:06:56.995,0:06:59.404 if the person's work[br]can be more effective at home 0:06:59.428,0:07:01.951 and more effective at work[br]and get high productivity. 0:07:01.975,0:07:04.593 I should say that some sectors[br]really need to adjust, 0:07:04.617,0:07:06.419 like hotels, tourism, dining. 0:07:06.443,0:07:10.156 In several industries, this will require[br]more thought and adjusting. 0:07:10.180,0:07:12.951 But other industries are almost[br]built for ideas like this. 0:07:12.975,0:07:16.922 Maybe it's even something[br]you can consider after the epidemic, 0:07:16.946,0:07:19.550 because productivity can be[br]at least as high. 0:07:20.785,0:07:25.411 CA: I mean, I read this and I started[br]thinking about our own organization, TED, 0:07:25.435,0:07:29.890 and how, in many ways, you could argue[br]that could work really well. 0:07:29.914,0:07:31.084 I mean, for one thing, 0:07:31.108,0:07:33.667 there's this question[br]about extroverts and introverts. 0:07:33.691,0:07:35.497 Some introverts, if they were honest, 0:07:35.521,0:07:38.865 might say that this pandemic[br]has been manna from heaven for them. 0:07:38.889,0:07:41.800 They've found work less stressful. 0:07:41.824,0:07:44.483 They've been able to focus and so forth. 0:07:44.507,0:07:48.617 With this sort of four days on,[br]four days off type strategy, 0:07:48.641,0:07:51.095 perhaps you can imagine a work world 0:07:51.119,0:07:54.898 that's optimized for both[br]introverts and extroverts? 0:07:56.451,0:07:57.609 UA: Absolutely. 0:07:57.633,0:07:58.900 I mean, I feel it also. 0:07:58.924,0:08:01.433 Me and my partner,[br]with different personalities, 0:08:01.457,0:08:02.939 we both teach in universities, 0:08:02.963,0:08:04.604 and teaching through this [?] 0:08:04.628,0:08:07.460 has [helped me] become[br]productive in certain ways. 0:08:07.986,0:08:09.376 So I agree completely, 0:08:09.400,0:08:13.340 and I think harnessing the creativity[br]of people at workplaces, 0:08:13.364,0:08:17.295 we're only at the beginning[br]of what these kinds of mixtures can offer. 0:08:18.758,0:08:21.386 CA: But for people who are[br]on the front line, 0:08:21.410,0:08:26.824 again, if you're delivering[br]goods and so forth 0:08:26.848,0:08:29.007 and you can't do that virtually, 0:08:29.031,0:08:30.457 is there any thought about 0:08:30.481,0:08:34.166 how a four days on[br]and then isolation strategy, 0:08:34.190,0:08:36.758 how that off time could be used 0:08:36.782,0:08:39.013 to nonetheless contribute[br]to that person's work 0:08:39.037,0:08:41.702 through some form of training? 0:08:41.726,0:08:48.385 Or is it more just that people would work[br]very intensely through the four days, 0:08:48.409,0:08:54.096 and maybe people still aren't quite[br]earning their full pay in this scenario, 0:08:54.120,0:08:56.032 but it's better than complete lockdown, 0:08:56.056,0:08:59.425 and it's better than going back to work[br]and seeing another surge? 0:09:01.013,0:09:02.164 UA: That's right. 0:09:02.188,0:09:03.386 So on a society level, 0:09:03.410,0:09:05.957 it's better than opening up[br]and seeing another surge, 0:09:05.981,0:09:07.825 which would require complete lockdown. 0:09:07.849,0:09:09.399 For people like hospital shifts, 0:09:09.423,0:09:12.817 some hospitals adopted[br]this kind of program 0:09:12.841,0:09:16.112 so we can protect shifts and avoid mixing. 0:09:16.136,0:09:18.456 It also creates a lot[br]of simplicity and clarity. 0:09:18.480,0:09:20.347 So you understand when you're working, 0:09:20.371,0:09:25.095 and you have some confidence because[br]this is based on scientific modeling 0:09:25.119,0:09:28.420 about the effectiveness of this plan. 0:09:28.444,0:09:31.852 It's also equitable in the sense[br]that everybody gets to go to work, 0:09:31.876,0:09:34.046 [but only?] [in only?][not only?][br]certain sectors. 0:09:34.070,0:09:35.531 It's transparent, etc. 0:09:35.555,0:09:38.463 [Cross talk] 0:09:40.452,0:09:42.948 CA: And is it something[br]that is best implemented 0:09:42.972,0:09:45.317 by individual companies? 0:09:45.341,0:09:48.417 Or is it actually much better[br]implemented a city at a time 0:09:48.441,0:09:50.500 or even a nation at a time? 0:09:51.857,0:09:54.298 UA: We think it can work [?]. 0:09:54.322,0:09:58.015 So at certain companies,[br]it's very natural to adopt, 0:09:58.039,0:09:59.914 or at hospitals, schools, etc. 0:09:59.938,0:10:02.792 It can also work at the level[br]of a town or a region, 0:10:02.816,0:10:07.295 and then we would advise[br]trying it out for something like a month, 0:10:07.319,0:10:09.233 seeing whether cases rise. 0:10:09.257,0:10:12.994 In that case, you can dial down[br]the number of workdays. 0:10:13.018,0:10:15.931 Or, if cases are declining quickly,[br]you can add workdays 0:10:15.955,0:10:21.046 and therefore adapt to the climate[br]and the location where a person is. 0:10:21.070,0:10:22.820 So it's quite adaptable. 0:10:24.161,0:10:27.047 CA: But by aligning work schedules[br]with schools, for example, 0:10:27.071,0:10:29.812 that suddenly allows parents[br]to go back to work 0:10:29.836,0:10:34.417 on the days that their kids are at school,[br]and you'd have to try -- 0:10:34.441,0:10:35.603 UA: Absolutely. 0:10:35.627,0:10:38.227 CA: I mean, is the best[br]instantiation of this 0:10:38.251,0:10:41.114 that countries literally divide households 0:10:41.138,0:10:45.283 into different A and B categories,[br]or something like that, 0:10:45.307,0:10:48.196 so that that kind[br]of alignment could happen? 0:10:49.173,0:10:50.334 UA: Exactly. 0:10:50.358,0:10:53.509 So you can align different[br]households, Group A and Group B, 0:10:53.533,0:10:56.328 and then the children go to school,[br]the parents go to work 0:10:56.352,0:10:57.739 in a synchronized way, 0:10:57.763,0:11:01.153 and the other group,[br]let's say, the alternating weeks. 0:11:01.177,0:11:03.676 A certain amount of people[br]need to work all the time. 0:11:03.700,0:11:07.782 Maybe teachers are, like, essential[br]workers and need to work throughout. 0:11:07.806,0:11:09.627 Just like during lockdown situations, 0:11:09.651,0:11:12.490 a certain fraction of the population[br]still works throughout. 0:11:12.514,0:11:15.943 But a region that does this[br]should be protected, in a sense, 0:11:15.967,0:11:18.480 because it has a replication[br]number of less than one, 0:11:18.504,0:11:22.151 so imported infections[br]also can't spread very much. 0:11:23.359,0:11:26.645 CA: And here is the aforementioned[br]David Biello. David. 0:11:26.669,0:11:28.533 David Biello: Yes. Hello, everybody. 0:11:28.557,0:11:31.303 Uri, as you can imagine,[br]there are lot of questions 0:11:31.327,0:11:33.213 from the audience, 0:11:33.237,0:11:35.058 and we have a first one 0:11:35.082,0:11:39.999 kind of about those workers[br]who have been marked as essential. 0:11:40.023,0:11:45.429 Can you comment on how this would impact[br]the health care professionals and others 0:11:45.453,0:11:51.049 who may not have time[br]or the flexibility to quarantine 0:11:51.073,0:11:52.388 in the way you suggest. 0:11:52.850,0:11:54.001 UA: That's great. 0:11:54.025,0:11:56.981 I want to say that[br]there's essential workers, 0:11:57.005,0:12:01.352 there's people with low income,[br]that just can't adhere to lockdown 0:12:01.376,0:12:03.053 because they have to make a living. 0:12:03.077,0:12:08.722 And studies show that mobility[br][among] people in the low-income sectors 0:12:08.746,0:12:10.254 is larger during lockdown. 0:12:10.278,0:12:14.024 And also, in developing countries,[br]people just have to go out of the house. 0:12:14.048,0:12:15.567 You can't enforce lockdown. 0:12:15.591,0:12:21.280 So this four-10 kind of strategy can[br]actually make lockdown easier to bear 0:12:21.304,0:12:24.668 for people who can still[br]make a living during those days, 0:12:24.692,0:12:26.884 or at least make their own choices 0:12:26.908,0:12:30.337 about what fraction to work[br]and what fraction to stay in lockdown. 0:12:30.361,0:12:33.208 Some countries can't get[br]R less than one even with lockdown, 0:12:33.232,0:12:36.534 because of this adherence problem[br][in formal sectors] [because of informal sectors,] etc. 0:12:36.558,0:12:40.908 We believe that a four-10 cycle[br]might make it easier to do lockdown 0:12:40.932,0:12:44.092 and maybe get our infection[br]level less than one. 0:12:44.116,0:12:47.897 That affects billions[br]of people in the world. 0:12:48.463,0:12:50.249 I hope I answered your question. 0:12:50.702,0:12:52.109 DB: I think so, 0:12:52.133,0:12:55.181 and we have another question, I believe, 0:12:55.205,0:12:58.197 if that can be queued up, 0:12:58.221,0:12:59.505 which is: 0:13:00.403,0:13:02.281 Any chance you can do the math 0:13:02.305,0:13:07.428 and quantify the increased risk[br]of this four-10 cycle? 0:13:08.503,0:13:11.894 UA: So the increased risk,[br]we're saying in our scientific paper, 0:13:11.918,0:13:14.908 we did all the sensitivity analyses, etc, 0:13:14.932,0:13:18.852 and the question is, it's comparing[br]increased risk comparing to what? 0:13:18.876,0:13:20.691 So, to the economy: 0:13:21.202,0:13:23.400 it's possible there will be a second wave. 0:13:23.424,0:13:26.273 I mean, I hope there won't be,[br]but it certainly is possible, 0:13:26.297,0:13:30.832 and in that case, it's clear[br]that a second wave and another lockdown 0:13:30.856,0:13:35.492 will have worse consequences on health 0:13:35.516,0:13:38.867 than a cycle of four-10. 0:13:38.891,0:13:44.268 And so it's really a question of[br]what you're comparing to. 0:13:45.858,0:13:47.542 DB: Sure. 0:13:47.566,0:13:51.199 Well, thank you so much[br]for sharing this idea, Uri. 0:13:53.170,0:13:55.505 CA: Indeed. 0:13:55.529,0:13:56.680 David, stay on. 0:13:56.704,0:13:58.235 But just before you go, 0:13:58.259,0:14:03.128 have any governments[br]expressed interest in exploring this? 0:14:03.152,0:14:06.052 Do you see people considering[br]actually implementing this 0:14:06.076,0:14:07.910 as national policy? 0:14:08.963,0:14:12.994 UA: Yes, we're in touch[br]with several European countries 0:14:13.018,0:14:16.703 and countries in South America[br]and Israel, of course. 0:14:16.727,0:14:20.812 Austria has adopted a similar program[br]for their school system, 0:14:21.875,0:14:26.560 which is five school days every two weeks. 0:14:26.584,0:14:32.193 And several companies and hospitals, etc. 0:14:32.217,0:14:36.664 And so we're very interested[br]to see how this will play out. 0:14:38.229,0:14:40.792 CA: Well, I love the basic start point 0:14:40.816,0:14:44.209 of starting by looking[br]at the enemy's weakness. 0:14:44.233,0:14:47.300 And you've got this four-day period 0:14:47.324,0:14:53.069 where it's not necessarily[br]that dangerous after an infection, 0:14:53.093,0:14:55.811 if you could figure out[br]a way to work then. 0:14:55.835,0:15:00.270 I assume that testing would actually[br]enhance this idea as well a lot, right? 0:15:00.294,0:15:02.168 To test people before they come back -- 0:15:02.192,0:15:03.899 UA: It's not predicated on testing. 0:15:03.923,0:15:06.091 You don't have to have[br]testing for this idea, 0:15:06.115,0:15:08.647 so that can apply to regions[br]without a lot of testing. 0:15:08.671,0:15:12.317 If you do have testing, it could help you[br]use testing in a more effective way 0:15:12.341,0:15:15.753 by concentrating testing on people[br]at the end of their 10 lockdown days, 0:15:15.777,0:15:17.503 just as they're about to go to work; 0:15:17.527,0:15:19.396 that could make[br]each test more impactful 0:15:19.420,0:15:21.646 in terms of reducing[br]their reproduction number. 0:15:21.670,0:15:24.542 CA: Indeed, instead of having[br]to test the whole population 0:15:24.566,0:15:25.883 every three or four days, 0:15:25.907,0:15:27.435 it's just once every two weeks. 0:15:27.459,0:15:30.283 That's a much more imaginable goal. 0:15:31.218,0:15:32.643 UA: Sure. 0:15:32.667,0:15:33.913 CA: Yeah. 0:15:33.937,0:15:37.015 Well, Uri Alon, thank you so much[br]for spending this time.