WEBVTT 00:00:00.000 --> 00:00:03.080 OK, let's get into this. We're taking a deep 00:00:03.080 --> 00:00:06.280 dive today into some, well, some really cutting 00:00:06.280 --> 00:00:08.900 edge areas that are fundamentally changing how 00:00:08.900 --> 00:00:11.500 we think about health and medicine. Absolutely. 00:00:11.800 --> 00:00:13.720 We're pulling insights from some fascinating 00:00:13.720 --> 00:00:16.320 conversations Eric Tupel hosted recently. That's 00:00:16.320 --> 00:00:18.539 right. First up, we'll be digging into the latest 00:00:18.539 --> 00:00:21.140 science on sleep with Professor Matthew Walker. 00:00:21.480 --> 00:00:24.980 And we're talking updates that go... quite a 00:00:24.980 --> 00:00:27.820 bit beyond just counting hours. Then we'll switch 00:00:27.820 --> 00:00:29.899 gears a bit, look at the bigger picture, the 00:00:29.899 --> 00:00:32.439 future landscape of life science. Things like 00:00:32.439 --> 00:00:37.000 genomics, AI, and our understanding of the immune 00:00:37.000 --> 00:00:39.500 system, which is evolving so rapidly. And for 00:00:39.500 --> 00:00:41.719 that, we're drawing on insights from Sir John 00:00:41.719 --> 00:00:43.899 Bell, who's a real visionary in those fields. 00:00:44.179 --> 00:00:47.140 Exactly. So the mission for this deep dive, really, 00:00:47.240 --> 00:00:50.420 is to pull out the most impactful bits of knowledge 00:00:50.420 --> 00:00:52.439 from what these experts are saying. Yeah, translating 00:00:52.439 --> 00:00:54.609 some quite complex research into to hopefully 00:00:54.609 --> 00:00:56.689 surprising facts and insights that are actually 00:00:56.689 --> 00:00:59.030 relevant, especially for professionals, working 00:00:59.030 --> 00:01:01.829 in industries touched by all these changes. Because 00:01:01.829 --> 00:01:03.890 understanding this stuff isn't just academic 00:01:03.890 --> 00:01:07.069 curiosity anymore, is it? It feels key to navigating, 00:01:07.069 --> 00:01:09.849 well, both our personal health futures and the 00:01:09.849 --> 00:01:11.950 professional world in the next few years. Couldn't 00:01:11.950 --> 00:01:14.549 agree more. It really is. Right then. Let's jump 00:01:14.549 --> 00:01:16.269 straight in. We'll start with the science of 00:01:16.269 --> 00:01:18.590 sleep, which is often, I think, underappreciated, 00:01:18.790 --> 00:01:21.230 and maybe begin with something quite unexpected 00:01:21.230 --> 00:01:25.540 at social side. Yes. It's become increasingly 00:01:25.540 --> 00:01:28.760 clear that sleep's influence goes way beyond 00:01:28.760 --> 00:01:32.239 just us as individuals. Recent work has really 00:01:32.239 --> 00:01:35.519 shown a light on its crucial role in how we interact 00:01:35.519 --> 00:01:38.079 socially. Which is surprising, isn't it? Because 00:01:38.079 --> 00:01:39.980 you tend to think of sleep deprivation effects 00:01:39.980 --> 00:01:43.760 as your mood, your concentration, maybe physical 00:01:43.760 --> 00:01:46.239 health. That's the traditional view, yes. But 00:01:46.239 --> 00:01:48.900 it's about more than just how you feel when you 00:01:48.900 --> 00:01:52.019 haven't slept. It's actually about how that lack 00:01:52.019 --> 00:01:54.700 of sleep affects the people around you. Okay, 00:01:54.780 --> 00:01:56.620 tell me more about that social dimension. What's 00:01:56.620 --> 00:01:58.799 the new thinking? Well, there's some really compelling 00:01:58.799 --> 00:02:01.560 research, particularly from Dr. Eddie Ben Simon 00:02:01.560 --> 00:02:04.840 over at Berkeley. Her work demonstrates pretty 00:02:04.840 --> 00:02:08.740 clearly that sleep deprivation makes people measurably 00:02:08.740 --> 00:02:13.199 more asocial. Asocial. Yes. Essentially, not 00:02:13.199 --> 00:02:15.099 getting enough sleep seems to make you, well, 00:02:15.280 --> 00:02:17.780 she used the term, socially repellent. Socially 00:02:17.780 --> 00:02:19.919 repellent? That sounds quite harsh. It is a strong 00:02:19.919 --> 00:02:22.699 term, yes. But the data suggests that people 00:02:22.699 --> 00:02:25.280 who are under -stumped tend to withdraw socially. 00:02:25.319 --> 00:02:28.319 They report feeling lonelier. And crucially, 00:02:28.699 --> 00:02:30.620 they seem to project something that makes other 00:02:30.620 --> 00:02:32.919 people less inclined to want to engage with them. 00:02:33.319 --> 00:02:35.539 So people can actually sense it, even if they 00:02:35.539 --> 00:02:37.599 don't know you pulled an all -nighter. It seems 00:02:37.599 --> 00:02:40.259 that way, yes. The research showed that even 00:02:40.259 --> 00:02:42.099 when observers didn't know the person was sleep 00:02:42.099 --> 00:02:45.240 -deprived, they rated them as less socially appealing, 00:02:45.560 --> 00:02:48.419 less attractive to interact with. And here's 00:02:48.419 --> 00:02:50.460 where it gets, well, frankly, a bit alarming. 00:02:51.000 --> 00:02:53.099 The studies also found that after interacting 00:02:53.099 --> 00:02:55.259 with someone who was sleep deprived, a person 00:02:55.259 --> 00:02:57.340 who was actually well -rested walked away feeling 00:02:57.340 --> 00:03:00.800 more lonely themselves. No way. So a kind of 00:03:00.800 --> 00:03:03.419 social loneliness contagion. Your bad night's 00:03:03.419 --> 00:03:05.520 sleep can actually make someone else feel lonely 00:03:05.520 --> 00:03:08.020 just by talking to them. Exactly. It's like a 00:03:08.020 --> 00:03:11.639 ripple effect. A sort of viral transmission of 00:03:11.639 --> 00:03:14.949 that feeling. And the work also uncovered another 00:03:14.949 --> 00:03:18.110 really significant social impact, a reduction 00:03:18.110 --> 00:03:20.349 in what they call helping behavior. Meaning? 00:03:20.729 --> 00:03:23.569 Meaning, when you're insufficiently slept, you're 00:03:23.569 --> 00:03:26.509 simply less likely to offer help to others, even 00:03:26.509 --> 00:03:29.110 strangers. Your sort of default setting shifts 00:03:29.110 --> 00:03:31.930 towards being less altruistic, less pro -social. 00:03:32.219 --> 00:03:34.300 And I remember reading there was population level 00:03:34.300 --> 00:03:36.199 data on this too, something about daylight savings. 00:03:36.319 --> 00:03:38.639 That's right, yes. The daylight sevens time study 00:03:38.639 --> 00:03:41.080 is a really striking example. They looked at 00:03:41.080 --> 00:03:43.729 charitable donations across the U .S. And in 00:03:43.729 --> 00:03:45.909 the days immediately after the clocks go forward 00:03:45.909 --> 00:03:48.430 in spring, when everyone loses that hour of sleep 00:03:48.430 --> 00:03:51.189 opportunity, there was a statistically significant 00:03:51.189 --> 00:03:54.250 drop. A drop in donations. A big dent, as Professor 00:03:54.250 --> 00:03:56.750 Walker described it, in national charitable giving. 00:03:56.990 --> 00:03:59.210 Just from losing one hour of potential sleep 00:03:59.210 --> 00:04:01.870 across the population, it shows this reduced 00:04:01.870 --> 00:04:03.990 helping behavior playing out on a massive scale. 00:04:04.229 --> 00:04:06.270 That's incredible. It really speaks volumes, 00:04:06.270 --> 00:04:09.830 doesn't it, about how fundamental sleep is for 00:04:09.830 --> 00:04:13.000 just Well, basic human cooperation and empathy. 00:04:13.199 --> 00:04:15.699 It does. And you can immediately see the professional 00:04:15.699 --> 00:04:18.939 implications. Teamwork, collaboration, effective 00:04:18.939 --> 00:04:22.800 leadership. They all rely so heavily on positive 00:04:22.800 --> 00:04:25.399 social interaction, on that willingness to help 00:04:25.399 --> 00:04:28.839 each other out. Absolutely. If your team or even 00:04:28.839 --> 00:04:32.060 just key people in it are chronically under -slept, 00:04:32.300 --> 00:04:34.899 it's bound to have a tangible, negative effect 00:04:34.899 --> 00:04:37.199 on how everyone functions together. It's not 00:04:37.199 --> 00:04:38.879 just about individual performance, it's about 00:04:38.879 --> 00:04:41.769 the collective. It really reframes sleep health, 00:04:41.870 --> 00:04:43.750 doesn't it? Not just a personal wellness thing, 00:04:43.889 --> 00:04:47.930 but almost a social or organizational imperative. 00:04:48.129 --> 00:04:49.910 I think that's exactly right. A social imperative. 00:04:50.009 --> 00:04:51.990 OK, so that's the surprising social angle. What 00:04:51.990 --> 00:04:54.129 about the underlying biology? There's been a 00:04:54.129 --> 00:04:55.990 lot of talk about genetic short sleepers. Is 00:04:55.990 --> 00:04:58.050 that a real thing? Are some people genuinely 00:04:58.050 --> 00:05:01.870 OK on much less sleep? Yes, it is real. But, 00:05:02.089 --> 00:05:05.089 and this is a huge, but it's incredibly exceedingly 00:05:05.089 --> 00:05:08.430 rare, for the vast, vast majority of us, that 00:05:08.430 --> 00:05:10.970 need for seven to nine hours is biologically 00:05:10.970 --> 00:05:13.689 baked in. The evidence for impairment below that 00:05:13.689 --> 00:05:15.910 is just overwhelming. But there are some people. 00:05:16.129 --> 00:05:18.569 There are. Science has identified this tiny, 00:05:18.629 --> 00:05:21.569 tiny subset of individuals who seem to genuinely 00:05:21.569 --> 00:05:25.350 thrive on significantly less sleep. Maybe five 00:05:25.350 --> 00:05:28.560 hours. Maybe even a bit less. And Critically, 00:05:28.920 --> 00:05:31.079 they show no apparent signs of impairment. No 00:05:31.079 --> 00:05:33.120 impairment at all. That sounds like a superpower. 00:05:33.339 --> 00:05:35.379 It almost defies everything else we know about 00:05:35.379 --> 00:05:37.399 sleep debt. It certainly seems that way, and 00:05:37.399 --> 00:05:39.379 researchers have actually pinpointed specific 00:05:39.379 --> 00:05:41.939 genes linked to this. The first one identified 00:05:41.939 --> 00:05:45.439 was called the DEC2 gene. A specific mutation 00:05:45.439 --> 00:05:47.560 in that gene seems strongly connected to this 00:05:47.560 --> 00:05:49.899 natural short sleep ability. But you stressed 00:05:49.899 --> 00:05:52.019 how rare it is. How rare are we talking? Professor 00:05:52.019 --> 00:05:54.079 Walker put it very starkly. He said, you are 00:05:54.079 --> 00:05:56.220 statistically more likely to be struck by lightning 00:05:56.220 --> 00:05:59.019 in your lifetime than to have this specific genetic 00:05:59.019 --> 00:06:00.899 makeup that allows you to genuinely function 00:06:00.899 --> 00:06:03.600 perfectly well on very little sleep. Struck by 00:06:03.600 --> 00:06:06.019 lightning. OK. It puts it in perspective. So 00:06:06.019 --> 00:06:07.839 most people who think they're fine on five hours 00:06:07.839 --> 00:06:09.980 probably aren't. That's the overwhelming likelihood, 00:06:10.180 --> 00:06:13.300 yes. They might feel okay, or they've adapted 00:06:13.300 --> 00:06:15.819 subjectively, but objectively they're likely 00:06:15.819 --> 00:06:18.500 accumulating significant sleep debt with all 00:06:18.500 --> 00:06:20.980 the downstream negative consequences for their 00:06:20.980 --> 00:06:23.519 health and cognitive function, even if they don't 00:06:23.519 --> 00:06:27.199 realize it. Right. So why is this tiny group 00:06:27.199 --> 00:06:30.920 so interesting to science, then, if it's that 00:06:30.920 --> 00:06:34.040 uncommon? Because it poses this fundamental biological 00:06:34.040 --> 00:06:37.339 puzzle. How has Mother Nature, through just random 00:06:37.339 --> 00:06:41.199 genetic mutation, found a way in these few individuals 00:06:41.199 --> 00:06:45.040 to essentially, well, ZIP file sleep? ZIP file 00:06:45.040 --> 00:06:47.240 sleep? I like that analogy. They somehow compress 00:06:47.240 --> 00:06:49.319 all the necessary functions that normally take 00:06:49.319 --> 00:06:51.639 seven to nine hours into maybe five or six and 00:06:51.639 --> 00:06:53.540 still seem to get all the benefits. The huge 00:06:53.540 --> 00:06:56.160 question is, how? What are the molecular tricks? 00:06:56.560 --> 00:06:58.379 What are the mechanisms that let them do that? 00:06:58.639 --> 00:07:00.459 And if we could figure that out? If we could 00:07:00.459 --> 00:07:03.040 figure that out. It could unlock profound secrets 00:07:03.040 --> 00:07:05.579 about what sleep is actually for at a fundamental 00:07:05.579 --> 00:07:09.519 level, and potentially lead to new ways to help 00:07:09.519 --> 00:07:12.579 people with sleep problems, or maybe even enhance 00:07:12.579 --> 00:07:15.860 sleep efficiency for everyone else. Which I suppose 00:07:15.860 --> 00:07:18.300 immediately raises thoughts about future applications, 00:07:18.439 --> 00:07:20.519 maybe even trying to replicate it artificially. 00:07:20.899 --> 00:07:22.939 Exactly. And that's where it gets complicated, 00:07:23.199 --> 00:07:26.120 ethically and societally. While the main scientific 00:07:26.120 --> 00:07:28.779 goal is understanding, the knowledge could lead 00:07:28.779 --> 00:07:32.420 down paths like developing drugs or even genome 00:07:32.420 --> 00:07:35.800 editing to try and induce this short sleep ability. 00:07:36.180 --> 00:07:38.079 And Professor Walker raised concerns about that. 00:07:38.319 --> 00:07:40.620 He did. He talked about the risk of an arms race 00:07:40.620 --> 00:07:43.879 of de -escalation and sleep. A scenario where 00:07:43.879 --> 00:07:47.180 society, chasing productivity gains, pushes people 00:07:47.180 --> 00:07:49.720 towards less and less sleep, potentially overriding 00:07:49.720 --> 00:07:51.819 other vital functions of sleep that we don't 00:07:51.819 --> 00:07:54.089 even fully understand yet. It highlights why 00:07:54.089 --> 00:07:58.250 we need to really grasp the why and how before 00:07:58.250 --> 00:08:00.370 we start thinking about artificial manipulation. 00:08:00.810 --> 00:08:02.949 Which is a perfect segue to another critical 00:08:02.949 --> 00:08:05.110 biological mechanism, the glymphatic system. 00:08:05.209 --> 00:08:07.050 This is the brain's cleaning crew, right? That's 00:08:07.050 --> 00:08:09.449 a great way to put it, yes. It's basically the 00:08:09.449 --> 00:08:11.629 brain's equivalent of the lymphatic system that 00:08:11.629 --> 00:08:14.750 cleans waste from the rest of the body. The glymphatic 00:08:14.750 --> 00:08:18.740 system's job is to flush out metabolic byproducts 00:08:18.740 --> 00:08:20.620 that accumulate in the brain while we're awake. 00:08:20.839 --> 00:08:22.939 Including things like the proteins linked to 00:08:22.939 --> 00:08:25.720 Alzheimer's. Precisely. Things like beta amyloid 00:08:25.720 --> 00:08:28.500 and tau proteins. They're essentially the molecular 00:08:28.500 --> 00:08:30.579 garbage that the brain produces through its normal 00:08:30.579 --> 00:08:32.899 activity. And it needs to be cleared out regularly, 00:08:33.100 --> 00:08:35.379 otherwise it builds up and causes problems. And 00:08:35.379 --> 00:08:37.539 the established view, the overwhelming evidence, 00:08:37.679 --> 00:08:40.120 has pointed to sleep being absolutely crucial 00:08:40.120 --> 00:08:42.299 for this process. Yes, that's been the prevailing 00:08:42.299 --> 00:08:44.919 understanding for some time now, based on a huge 00:08:44.919 --> 00:08:47.379 amount of evidence from many different research 00:08:47.379 --> 00:08:51.480 groups across different animal species and increasingly 00:08:51.480 --> 00:08:54.259 in humans too. The idea being that deep sleep 00:08:54.259 --> 00:08:57.080 is the prime time for this brainwashing. Exactly. 00:08:57.360 --> 00:09:00.100 The data strongly suggested that sleep, particularly 00:09:00.100 --> 00:09:03.919 that deep, non -REM sleep, stage 3 sleep, provides 00:09:03.919 --> 00:09:06.580 the optimal window for this glymphatic system 00:09:06.580 --> 00:09:09.100 to really kick into high gear. Why then? Why 00:09:09.100 --> 00:09:11.480 not when we're awake? Well, the thinking is that 00:09:11.480 --> 00:09:13.960 when we're awake, the brain is just too busy 00:09:13.960 --> 00:09:17.159 processing information. There's less space between 00:09:17.159 --> 00:09:20.059 the brain cells, and the fluid flow needed for 00:09:20.059 --> 00:09:22.620 cleansing seems to be significantly reduced. 00:09:23.360 --> 00:09:25.440 Some researchers even describe wakefulness as 00:09:25.440 --> 00:09:28.399 a state of sort of low -level biochemical damage, 00:09:28.899 --> 00:09:31.679 and sleep provides the essential sanitary salvation. 00:09:31.850 --> 00:09:34.190 But wasn't there a study recently that caused 00:09:34.190 --> 00:09:36.850 a bit of a stir, suggesting maybe clearance was 00:09:36.850 --> 00:09:39.009 higher when we were awake? That's correct. A 00:09:39.009 --> 00:09:43.149 study published I think around 2022 or 2023 used 00:09:43.149 --> 00:09:46.169 some specific imaging techniques in humans and 00:09:46.169 --> 00:09:49.009 reported findings suggesting higher fluid flow 00:09:49.009 --> 00:09:52.070 and potentially higher clearance during wakefulness 00:09:52.070 --> 00:09:55.129 compared to sleep, which naturally seemed to 00:09:55.129 --> 00:09:56.669 contradict quite a lot of the previous work. 00:09:56.809 --> 00:09:58.750 And what was the reaction from other scientists? 00:09:58.870 --> 00:10:01.529 Were there criticisms of that study? Yes, researchers 00:10:01.529 --> 00:10:03.289 who'd been proponents of the sleep -dependent 00:10:03.289 --> 00:10:05.309 model raised several points about the methodology. 00:10:05.909 --> 00:10:08.049 They noted the use of imaging methods that weren't 00:10:08.049 --> 00:10:10.070 standard, or at least different from those used 00:10:10.070 --> 00:10:12.330 in many of the animal studies. There were also 00:10:12.330 --> 00:10:14.710 questions about an artificial method used to 00:10:14.710 --> 00:10:17.070 perfuse fluid through the brain, rather than 00:10:17.070 --> 00:10:19.590 observing the natural flow. And importantly, 00:10:19.970 --> 00:10:21.950 concerns about the sampling rate, whether it 00:10:21.950 --> 00:10:25.129 was fast enough to capture the crucial slow brainwave 00:10:25.129 --> 00:10:28.080 oscillations that happen during deep sleep. These 00:10:28.080 --> 00:10:30.059 slow waves being important for driving the fluid 00:10:30.059 --> 00:10:32.580 flow. That's the hypothesis, yes, that these 00:10:32.580 --> 00:10:36.139 very slow, large amplitude brain waves physically 00:10:36.139 --> 00:10:39.120 help to pump the cerebral spinal fluid through 00:10:39.120 --> 00:10:41.659 the brain tissue, facilitating that waste removal. 00:10:42.279 --> 00:10:44.139 So if the measurement technique missed those, 00:10:44.460 --> 00:10:46.940 it could skew the results. So weighing everything 00:10:46.940 --> 00:10:49.419 up, what's the current sort of nuanced understanding? 00:10:49.539 --> 00:10:51.879 Is it all about sleep or is it more complicated? 00:10:52.139 --> 00:10:54.580 I think the picture is becoming a bit more nuanced 00:10:54.580 --> 00:10:56.799 than just saying it only happens during sleep. 00:10:57.279 --> 00:11:00.440 The glymphatic system is clearly dynamic. It's 00:11:00.440 --> 00:11:02.879 possible there's a kind of low battery mode clearance 00:11:02.879 --> 00:11:05.480 that can happen during quiet wakefulness. But 00:11:05.480 --> 00:11:08.440 sleep is still king. But the overwhelming weight 00:11:08.440 --> 00:11:10.620 of evidence, when you look across all the different 00:11:10.620 --> 00:11:12.299 studies, different species, different methods, 00:11:12.899 --> 00:11:15.559 still points very strongly towards sleep, especially 00:11:15.559 --> 00:11:18.929 deep and REM sleep. providing the optimal opportunity 00:11:18.929 --> 00:11:22.769 for a really thorough cleanse. A full biochemical 00:11:22.769 --> 00:11:25.129 reboot. Like the difference between putting your 00:11:25.129 --> 00:11:27.950 phone on low power mode versus actually switching 00:11:27.950 --> 00:11:30.409 it off for a proper update and refresh. That's 00:11:30.409 --> 00:11:33.009 a great analogy, yes. You might get some background 00:11:33.009 --> 00:11:35.990 tidying in low power mode, but the deep clean, 00:11:36.070 --> 00:11:38.750 the system overhaul, really happens when it's 00:11:38.750 --> 00:11:41.509 properly off. The bulk of evidence still firmly 00:11:41.509 --> 00:11:44.490 supports sleep as the preferential most effective 00:11:44.490 --> 00:11:47.080 state for this vital brain maintenance. It just 00:11:47.080 --> 00:11:48.879 makes intuitive sense, doesn't it, that such 00:11:48.879 --> 00:11:52.039 a complex, active organ needs dedicated downtime 00:11:52.039 --> 00:11:54.840 for housekeeping? And you mentioned a link between 00:11:54.840 --> 00:11:57.080 this glymphatic system and other parts of the 00:11:57.080 --> 00:11:59.539 body, like the heart. Yes, that was fascinating. 00:12:00.080 --> 00:12:02.419 A paper in a cardiology journal reported that 00:12:02.419 --> 00:12:05.220 patients with atrial fibrillation, or AFib, a 00:12:05.220 --> 00:12:07.299 common heart rhythm problem, actually showed 00:12:07.299 --> 00:12:09.820 less active glymphatic clearance during their 00:12:09.820 --> 00:12:12.460 sleep. Really? How does that connect? It highlights 00:12:12.460 --> 00:12:15.559 this idea of cardiorespiratory neurophysiological 00:12:15.559 --> 00:12:19.090 coupling. It suggests it's not just the brain's 00:12:19.090 --> 00:12:20.950 electrical activity driving this fluid flow. 00:12:21.649 --> 00:12:23.830 The heart and the lungs, their rhythms during 00:12:23.830 --> 00:12:27.009 sleep, seem to synchronize with the brain's slow 00:12:27.009 --> 00:12:29.629 waves to help drive the pulsations needed for 00:12:29.629 --> 00:12:31.570 efficient clearance. So it's the whole system 00:12:31.570 --> 00:12:34.070 working together heart, lungs, brain to wash 00:12:34.070 --> 00:12:36.950 the brain during sleep. Exactly. It's a beautiful 00:12:36.950 --> 00:12:39.309 illustration of how sleep benefits the entire 00:12:39.309 --> 00:12:41.990 embodied organism, as Professor Walker put it. 00:12:42.090 --> 00:12:44.330 Your body and your brain are deeply interconnected 00:12:44.330 --> 00:12:47.779 in this process. good cardiovascular and respiratory 00:12:47.779 --> 00:12:50.139 health might actually contribute to better brain 00:12:50.139 --> 00:12:52.779 cleansing overnight. That link is quite profound. 00:12:53.120 --> 00:12:55.200 A heart condition potentially impacting brain 00:12:55.200 --> 00:12:58.360 maintenance. Wow. And speaking of things that 00:12:58.360 --> 00:13:00.820 impact sleep and potentially this glymphatic 00:13:00.820 --> 00:13:03.879 function, what about sleep medications? Ah, yes. 00:13:03.980 --> 00:13:06.559 A critical area. And the research here has thrown 00:13:06.559 --> 00:13:09.509 up some quite surprising and frankly concerning 00:13:09.509 --> 00:13:12.009 findings, especially regarding the older, more 00:13:12.009 --> 00:13:14.090 established classes of sleeping pills. You mean 00:13:14.090 --> 00:13:17.190 things like Ambien, the Z drugs, benzodiazepines? 00:13:17.509 --> 00:13:20.789 Exactly. Those drugs, the ones often called sedative 00:13:20.789 --> 00:13:24.009 notics, zolpidem, ezolpaclone, zeleplon, and 00:13:24.009 --> 00:13:27.409 the older benzodiazepines like diazepam or temezepam, 00:13:27.850 --> 00:13:31.110 they primarily work by, well... broadly sedating 00:13:31.110 --> 00:13:33.529 the brain's cortex. They essentially just dampen 00:13:33.529 --> 00:13:35.769 down overall brain activity. And they do help 00:13:35.769 --> 00:13:37.950 people sleep or at least feel like they're sleeping 00:13:37.950 --> 00:13:40.149 more. They certainly make you fall asleep faster 00:13:40.149 --> 00:13:42.950 and they can increase the total time spent apparently 00:13:42.950 --> 00:13:45.720 asleep. On a standard sleep study, they might 00:13:45.720 --> 00:13:47.980 even show an increase in deep sleep minutes. 00:13:48.720 --> 00:13:51.679 However, and this is crucial, they seem to significantly 00:13:51.679 --> 00:13:54.340 alter the electrical quality or the signature 00:13:54.340 --> 00:13:56.480 of that deep sleep. Alter the signature. What 00:13:56.480 --> 00:13:58.500 does that mean in practice? It means they seem 00:13:58.500 --> 00:14:00.980 to create a substantial dent, as Professor Walker 00:14:00.980 --> 00:14:04.039 termed it, in those very large, very slow brain 00:14:04.039 --> 00:14:06.919 waves that are the hallmark of truly restorative 00:14:06.919 --> 00:14:10.480 deep NREM sleep. They suppress those critical 00:14:10.480 --> 00:14:12.779 slow oscillations. The same oscillations thought 00:14:12.779 --> 00:14:16.120 to drive the glymphatic system. Precisely. So 00:14:16.120 --> 00:14:19.139 by suppressing those key brain waves, even if 00:14:19.139 --> 00:14:21.120 the sleep looks like deep sleep on a superficial 00:14:21.120 --> 00:14:23.820 level, the underlying functional machinery needed 00:14:23.820 --> 00:14:26.559 for optimal brain cleansing seems to be impaired. 00:14:27.240 --> 00:14:29.659 So even if your sleep tracker says you got more 00:14:29.659 --> 00:14:32.080 deep sleep after taking an older sleeping pill, 00:14:32.379 --> 00:14:34.759 you might not actually be getting the brain cleaning 00:14:34.759 --> 00:14:37.460 benefits. That's exactly what the evidence suggests. 00:14:38.299 --> 00:14:40.720 Pioneering work, particularly from making Natagard's 00:14:40.720 --> 00:14:43.779 lab using animal models, found that despite increasing 00:14:43.779 --> 00:14:46.480 sleep time and apparent deep sleep time, the 00:14:46.480 --> 00:14:48.460 actual brain glymphatic cleansing was reduced 00:14:48.460 --> 00:14:51.740 perhaps by 30 % to 40 % when these older sedative 00:14:51.740 --> 00:14:54.039 hypnotics were used. Reduced? That's incredibly 00:14:54.039 --> 00:14:55.740 counterintuitive. You take something to help 00:14:55.740 --> 00:14:58.720 you sleep better. And it hinders one of sleep's 00:14:58.720 --> 00:15:01.279 key functions. It is profoundly counterintuitive, 00:15:01.460 --> 00:15:03.679 yes. But it strongly suggests these drugs might 00:15:03.679 --> 00:15:05.480 be generating what's been called junk sleep, 00:15:06.000 --> 00:15:08.159 sleep that ticks the boxes on basic duration 00:15:08.159 --> 00:15:10.600 measures but lacks some of the crucial functional 00:15:10.600 --> 00:15:13.799 properties, like efficient waste clearance. That's 00:15:13.799 --> 00:15:16.299 a really significant finding, given how widely 00:15:16.299 --> 00:15:18.220 prescribed these medications are and have been 00:15:18.220 --> 00:15:21.970 for decades. Is there a newer approach, a different 00:15:21.970 --> 00:15:24.330 class of drug that might avoid this problem? 00:15:24.669 --> 00:15:26.789 Yes, thankfully there is. There's a newer class 00:15:26.789 --> 00:15:29.490 of medications called DORAZ. That stands for 00:15:29.490 --> 00:15:32.330 dualorexin receptor antagonists. DORAZ? Okay. 00:15:32.750 --> 00:15:35.129 How do they work differently? They represent 00:15:35.129 --> 00:15:38.139 a much more... well, elegant and targeted way 00:15:38.139 --> 00:15:40.639 to promote sleep. Instead of just blanketing 00:15:40.639 --> 00:15:43.139 the cortex with sedation, they target a very 00:15:43.139 --> 00:15:46.000 specific chemical system in the brain, the orexin 00:15:46.000 --> 00:15:49.440 system. Orexin. Orexin, sometimes called hypocretin, 00:15:49.779 --> 00:15:52.200 is a key neurotransmitter that acts like a wakefulness 00:15:52.200 --> 00:15:55.860 switch or volume control. Keeps us alert. Doors 00:15:55.860 --> 00:15:58.100 work by blocking the receptors that erect and 00:15:58.100 --> 00:16:00.279 binds to, so they essentially just turn down 00:16:00.279 --> 00:16:03.100 the volume on wakefulness. So less forcing sedation, 00:16:03.220 --> 00:16:05.820 more allowing sleep to happen naturally. Exactly. 00:16:06.220 --> 00:16:07.820 It doesn't push the brain into an artificial 00:16:07.820 --> 00:16:10.139 state, but rather removes the wakeup signal, 00:16:10.700 --> 00:16:12.820 allowing the natural process of sleep onset and 00:16:12.820 --> 00:16:15.100 maintenance to emerge more easily when conditions 00:16:15.100 --> 00:16:17.220 are right. It's a much more subtle mechanism. 00:16:17.399 --> 00:16:20.519 And have these newer Dora drugs been tested for 00:16:20.519 --> 00:16:23.159 their effects on glymphatic activity or brain 00:16:23.159 --> 00:16:26.120 cleansing? Yes. And there's a very exciting and 00:16:26.120 --> 00:16:28.059 promising study that Professor Walker highlighted. 00:16:28.679 --> 00:16:32.159 It involved older adults, age 65 and up, using 00:16:32.159 --> 00:16:35.519 a Dora drug called SuverXant brand name Balsamra. 00:16:35.679 --> 00:16:38.080 What did that study involve? It was a proper, 00:16:38.600 --> 00:16:40.940 rigorous, randomized placebo -controlled trial. 00:16:41.679 --> 00:16:44.139 They gave the participants either SuverXant or 00:16:44.139 --> 00:16:46.740 a placebo. They measured not only how much sleep 00:16:46.740 --> 00:16:50.179 they got and how much deep NREM sleep, but crucially, 00:16:50.539 --> 00:16:52.659 they also measured the levels of those Alzheimer's 00:16:52.659 --> 00:16:55.799 biomarkers, beta amyloid and tau in the bloodstream 00:16:55.799 --> 00:16:58.299 before and after the night of sleep. Okay, and 00:16:58.299 --> 00:17:00.080 what did they find regarding those biomarkers? 00:17:00.179 --> 00:17:02.139 Did the Dora help clear them? They found that 00:17:02.139 --> 00:17:04.299 on the night the participants took the Dora drug, 00:17:04.619 --> 00:17:07.339 Suvarexan, there was significantly greater clearance 00:17:07.339 --> 00:17:10.279 of both beta amyloid and tau protein from the 00:17:10.279 --> 00:17:12.200 brain into the bloodstream compared to when they 00:17:12.200 --> 00:17:14.319 took the placebo. The opposite effect to the 00:17:14.319 --> 00:17:17.339 older drugs. Exactly the opposite. It wasn't 00:17:17.339 --> 00:17:19.539 just increasing sleep duration or deep sleep 00:17:19.539 --> 00:17:21.900 minutes, it seemed to be generating genuinely 00:17:21.900 --> 00:17:25.509 functional sleep. sleep that was actually facilitating 00:17:25.509 --> 00:17:29.009 the removal of these potentially harmful proteins 00:17:29.009 --> 00:17:31.589 linked to Alzheimer's disease. That's a massive 00:17:31.589 --> 00:17:34.549 development. It really is. Professor Walker framed 00:17:34.549 --> 00:17:36.910 it as potentially the first evidence we have 00:17:36.910 --> 00:17:39.450 of a sleep medication that not only helps you 00:17:39.450 --> 00:17:42.349 sleep in a more naturalistic way, but also seems 00:17:42.349 --> 00:17:45.109 to enhance the brain's own restorative cleansing 00:17:45.109 --> 00:17:47.710 processes. That could have huge implications 00:17:47.710 --> 00:17:50.410 for brain health, especially as we age. It really 00:17:50.410 --> 00:17:52.690 drives home that not all sleep is created equal, 00:17:52.829 --> 00:17:55.869 especially when medication is involved. So quantity 00:17:55.869 --> 00:17:57.950 is important, obviously, but it sounds like we 00:17:57.950 --> 00:18:00.109 absolutely have to focus just as much, if not 00:18:00.109 --> 00:18:03.609 more, on sleep quality and also sleep regularity. 00:18:03.910 --> 00:18:06.529 Absolutely. For years, the public health message, 00:18:06.609 --> 00:18:09.430 the whole conversation, was heavily skewed towards 00:18:09.430 --> 00:18:12.430 just getting enough hours The quantity. And don't 00:18:12.430 --> 00:18:14.670 get me wrong, sufficient quantity is non -negotiable. 00:18:14.769 --> 00:18:16.390 You need that foundation. But it's not the whole 00:18:16.390 --> 00:18:19.509 story. Not by a long shot. Research, especially 00:18:19.509 --> 00:18:22.230 over the last decade or so, has made it increasingly 00:18:22.230 --> 00:18:25.250 clear that sleep quality and sleep regularity 00:18:25.250 --> 00:18:28.690 are equally, and in some cases more, predictive 00:18:28.690 --> 00:18:31.349 of critical health outcomes. Like what kind of 00:18:31.349 --> 00:18:34.509 outcomes? Things like all -cause mortality, your 00:18:34.509 --> 00:18:38.390 overall risk of dying from any cause, also cardiovascular 00:18:38.390 --> 00:18:41.509 disease risk, metabolic health, think diabetes 00:18:41.509 --> 00:18:44.210 risk, and potentially even cancer mortality as 00:18:44.210 --> 00:18:47.859 well. Quality and regularity are powerful predictors 00:18:47.859 --> 00:18:50.720 across the board. OK, let's unpack quality first. 00:18:50.900 --> 00:18:53.400 How is that usually defined or measured? A common 00:18:53.400 --> 00:18:55.740 and useful metric for sleep quality is sleep 00:18:55.740 --> 00:18:58.480 efficiency. It's actually quite simple. It's 00:18:58.480 --> 00:19:00.339 the percentage of time you spend actually asleep 00:19:00.339 --> 00:19:02.259 while you're in bed trying to sleep. Right. So 00:19:02.259 --> 00:19:04.140 if you're in bed for eight hours, but tossing 00:19:04.140 --> 00:19:06.240 and turning for two, your efficiency isn't great. 00:19:06.579 --> 00:19:09.259 Exactly. Generally, a sleep efficiency of 85 00:19:09.259 --> 00:19:12.359 % or higher is considered good, indicating solid, 00:19:12.480 --> 00:19:15.200 consolidated sleep. When efficiency drops much 00:19:15.200 --> 00:19:17.740 below that, usually because your sleep is fragmented 00:19:17.740 --> 00:19:19.859 with lots of awakenings, that's when you start 00:19:19.859 --> 00:19:22.140 seeing problems. And what kind of problems are 00:19:22.140 --> 00:19:25.200 linked specifically to poor quality fragmented 00:19:25.200 --> 00:19:27.759 sleep? Pretty much system -wide impairments, 00:19:28.440 --> 00:19:30.920 increased risk of those metabolic and cardiovascular 00:19:30.920 --> 00:19:33.960 issues we mentioned, hormonal imbalances. But 00:19:33.960 --> 00:19:36.299 perhaps most strikingly, the effects on mood 00:19:36.299 --> 00:19:39.269 and mental health are profound. How so? It's 00:19:39.269 --> 00:19:42.369 a remarkable finding, but sleep is abnormal in 00:19:42.369 --> 00:19:44.650 every single psychiatric condition that's ever 00:19:44.650 --> 00:19:48.450 been studied. Depression, anxiety, PTSD, bipolar 00:19:48.450 --> 00:19:52.109 disorder, schizophrenia, you name it, sleep is 00:19:52.109 --> 00:19:55.019 disrupted. And very often, the key feature isn't 00:19:55.019 --> 00:19:57.640 necessarily just short sleep duration, but severely 00:19:57.640 --> 00:20:00.759 fragmented poor quality sleep. Poor sleep efficiency 00:20:00.759 --> 00:20:02.839 is a major red flag for mental health issues. 00:20:03.299 --> 00:20:05.240 That really highlights how fundamental good quality 00:20:05.240 --> 00:20:08.019 sleep is for just, well, keeping everything running 00:20:08.019 --> 00:20:10.220 smoothly mentally and physically. Okay, what 00:20:10.220 --> 00:20:12.099 about regularity then? You said that's also vital, 00:20:12.220 --> 00:20:14.740 maybe even underappreciated. Hugely underappreciated, 00:20:14.759 --> 00:20:18.140 I think. but absolutely vital. There was a really 00:20:18.140 --> 00:20:21.180 stunning study using data from the UK Biobank, 00:20:21.359 --> 00:20:23.460 this massive data set with hundreds of thousands 00:20:23.460 --> 00:20:25.119 of people. What did they look at in that study? 00:20:25.400 --> 00:20:27.480 They tracked not just how much sleep people were 00:20:27.480 --> 00:20:30.099 getting on average, but also how regular their 00:20:30.099 --> 00:20:33.480 sleep schedules were over many years. So how 00:20:33.480 --> 00:20:35.759 consistent were their bedtimes and wake times 00:20:35.759 --> 00:20:38.279 from day to day, including weekends? And what 00:20:38.279 --> 00:20:40.319 did they find about regularity versus quantity? 00:20:40.569 --> 00:20:43.650 The results were quite remarkable. People who 00:20:43.650 --> 00:20:47.210 maintained high sleep regularity, basically going 00:20:47.210 --> 00:20:49.309 to bed and waking up around the same time every 00:20:49.309 --> 00:20:52.109 day, had a significantly lower risk of dying 00:20:52.109 --> 00:20:55.069 from any cause, from cancer and from cardiovascular 00:20:55.069 --> 00:20:57.450 disease, compared to those with very irregular 00:20:57.450 --> 00:20:59.250 schedules. Even accounting for how much sleep 00:20:59.250 --> 00:21:01.849 they got overall. Yes. And here's the kicker. 00:21:01.980 --> 00:21:04.940 When they statistically compared the predictive 00:21:04.940 --> 00:21:07.799 power of regularity versus quantity for all -cause 00:21:07.799 --> 00:21:11.079 mortality, regularity almost beat quantity. Really? 00:21:11.279 --> 00:21:13.160 Almost more important than total hours slept. 00:21:13.400 --> 00:21:15.240 That feels counterintuitive if we've always been 00:21:15.240 --> 00:21:17.359 told to get your eight hours. It does, doesn't 00:21:17.359 --> 00:21:20.319 it? But it highlights something crucial. While 00:21:20.319 --> 00:21:23.039 you absolutely need sufficient quantity, being 00:21:23.039 --> 00:21:25.400 perfectly regular but only sleeping four hours 00:21:25.400 --> 00:21:28.680 a night isn't going to work. Regularity acts 00:21:28.680 --> 00:21:31.430 as a powerful anchor for your entire internal 00:21:31.430 --> 00:21:34.450 biological clock, your 24 -hour circadian rhythm. 00:21:34.809 --> 00:21:37.210 So keeping a consistent schedule helps keep your 00:21:37.210 --> 00:21:40.130 body clock stable. Exactly. It sends strong, 00:21:40.269 --> 00:21:42.410 reliable timing signals to your brain and body 00:21:42.410 --> 00:21:45.190 every day. And having a well -aligned, robust 00:21:45.190 --> 00:21:47.630 circadian rhythm doesn't just directly benefit 00:21:47.630 --> 00:21:50.109 health. It also feeds back to improve both your 00:21:50.109 --> 00:21:52.720 sleep quantity and your sleep quality. They all 00:21:52.720 --> 00:21:55.920 reinforce each other quantity quality and regularity 00:21:55.920 --> 00:21:57.819 They're the three pillars of healthy sleep and 00:21:57.819 --> 00:21:59.740 we need to pay attention to all of them Okay, 00:21:59.759 --> 00:22:02.559 so with all this detail stages quality efficiency 00:22:02.559 --> 00:22:04.700 regularity It's no surprise people are turning 00:22:04.700 --> 00:22:07.980 to technology sleep trackers rings watches even 00:22:07.980 --> 00:22:10.440 smart mattresses now Are these things genuinely 00:22:10.440 --> 00:22:12.839 helpful or can they sometimes make things worse? 00:22:13.059 --> 00:22:15.619 It's a great question and the answer is Well, 00:22:15.680 --> 00:22:18.400 it depends For the average person, someone who 00:22:18.400 --> 00:22:20.960 doesn't already have clinical insomnia or a lot 00:22:20.960 --> 00:22:23.619 of anxiety around sleep, tracking can generally 00:22:23.619 --> 00:22:26.859 be pretty useful. Why is that? Because sleep 00:22:26.859 --> 00:22:30.200 is notoriously difficult for us to judge subjectively. 00:22:30.720 --> 00:22:32.880 We're often quite bad at estimating how long 00:22:32.880 --> 00:22:34.779 it took us to fall asleep, how many times we 00:22:34.779 --> 00:22:37.460 woke up, how much deep sleep we got. Unlike, 00:22:37.480 --> 00:22:39.980 say, counting steps or tracking calories, which 00:22:39.980 --> 00:22:42.910 are more concrete. So the old adage, what gets 00:22:42.910 --> 00:22:45.109 measured gets managed, can definitely apply. 00:22:45.269 --> 00:22:47.849 Can you give an example of how seeing the data 00:22:47.849 --> 00:22:50.650 might help someone change their behavior? A classic 00:22:50.650 --> 00:22:53.430 one is alcohol. Many people are quite surprised, 00:22:53.470 --> 00:22:55.789 maybe even shocked, when they see the objective 00:22:55.789 --> 00:22:58.210 data from their tracker showing just how consistently 00:22:58.210 --> 00:23:01.130 even moderate alcohol consumption close to bedtime 00:23:01.130 --> 00:23:03.970 absolutely tanks their sleep quality. How does 00:23:03.970 --> 00:23:06.289 it show up in the data? Typically you'll see 00:23:06.289 --> 00:23:08.799 reduced REM sleep. Increased heart rate throughout 00:23:08.799 --> 00:23:11.220 the night, more awakenings, poor sleep efficiency. 00:23:11.859 --> 00:23:14.279 Seeing that pattern night after night can be 00:23:14.279 --> 00:23:17.140 a really powerful motivator to, say, cut back 00:23:17.140 --> 00:23:19.319 on evening drinking or stop drinking earlier. 00:23:19.619 --> 00:23:21.900 Right, makes sense. If someone is thinking about 00:23:21.900 --> 00:23:24.400 getting a tracker, any tips on what to look for? 00:23:24.859 --> 00:23:27.880 I think the key things are low friction and reasonable 00:23:27.880 --> 00:23:30.859 accuracy. Low friction just means it needs to 00:23:30.859 --> 00:23:33.200 be easy and comfortable enough that you'll actually 00:23:33.200 --> 00:23:35.619 wear it or use it every single night without 00:23:35.619 --> 00:23:38.349 it being a hassle. So something like a ring might 00:23:38.349 --> 00:23:40.809 be lower friction than a watch for some people. 00:23:41.069 --> 00:23:44.829 Potentially, yes. Or a smart mattress is perhaps 00:23:44.829 --> 00:23:47.789 the ultimate in low friction. It's just there. 00:23:47.829 --> 00:23:50.369 You don't wear anything. In terms of accuracy, 00:23:50.549 --> 00:23:53.049 especially for sleep staging, the technology 00:23:53.049 --> 00:23:55.819 is getting better all the time. Data generally 00:23:55.819 --> 00:23:58.400 suggests devices like the Aura Ring are currently 00:23:58.400 --> 00:24:00.519 among the more accurate wearables compared to 00:24:00.519 --> 00:24:03.000 gold standard sleep lab polysomnography. But 00:24:03.000 --> 00:24:05.599 others like Fitbit, Apple Watch, Whoop are also 00:24:05.599 --> 00:24:08.220 pretty good and constantly improving. Full disclosure, 00:24:08.400 --> 00:24:10.400 I should mention I have previously done some 00:24:10.400 --> 00:24:12.500 advisory work for Aura in the past. Okay, useful 00:24:12.500 --> 00:24:15.039 context. But you mentioned a potential downside, 00:24:15.279 --> 00:24:17.509 particularly for some people. Yes, and this is 00:24:17.509 --> 00:24:19.950 really important. For individuals who are already 00:24:19.950 --> 00:24:22.630 prone to anxiety, especially anxiety about sleep 00:24:22.630 --> 00:24:25.589 or maybe have perfectionistic tendencies, these 00:24:25.589 --> 00:24:28.450 trackers can sometimes backfire. There's a recognized 00:24:28.450 --> 00:24:31.779 phenomenon now called orthosomia. Orthosomnia? 00:24:32.039 --> 00:24:34.279 What's that? It's basically where someone becomes 00:24:34.279 --> 00:24:37.720 so fixated, so obsessed with achieving the perfect 00:24:37.720 --> 00:24:40.460 sleep score or hitting ideal numbers on their 00:24:40.460 --> 00:24:42.640 sleep tracker app. That the anxiety about it 00:24:42.640 --> 00:24:45.480 actually makes their sleep worse. Exactly. The 00:24:45.480 --> 00:24:47.460 pressure to perform to get that perfect data 00:24:47.460 --> 00:24:50.960 creates stress. And stress is the absolute enemy 00:24:50.960 --> 00:24:53.460 of sleep. It becomes this vicious cycle. You 00:24:53.460 --> 00:24:55.720 worry about the score. That makes you sleep poorly. 00:24:55.759 --> 00:24:58.390 You get a bad score. You worry more. So what 00:24:58.390 --> 00:25:00.769 should someone do if they suspect they might 00:25:00.769 --> 00:25:02.990 be falling into that trap? Well, the first thing 00:25:02.990 --> 00:25:04.829 might be to simply take a break from tracking 00:25:04.829 --> 00:25:07.450 altogether for a while. See if that reduces the 00:25:07.450 --> 00:25:11.190 anxiety. If the sleep problems persist, especially 00:25:11.190 --> 00:25:14.130 if it's chronic insomnia, the gold standard treatment 00:25:14.130 --> 00:25:16.809 is cognitive behavioral therapy for insomnia, 00:25:17.250 --> 00:25:20.569 or CBTI. That directly addresses the unhelpful 00:25:20.569 --> 00:25:23.410 thoughts and behaviors around sleep. Another 00:25:23.410 --> 00:25:25.990 strategy could be to just check the data much 00:25:25.990 --> 00:25:28.660 less frequently. maybe only once a week, looking 00:25:28.660 --> 00:25:30.700 for broad trends rather than getting caught up 00:25:30.700 --> 00:25:33.480 in the nightly numbers. It's a really valuable 00:25:33.480 --> 00:25:35.660 reminder, isn't it, that the tracker data is 00:25:35.660 --> 00:25:39.079 just a tool. It's information. It's not the ultimate 00:25:39.079 --> 00:25:42.039 judge of your sleep health. How you feel is still 00:25:42.039 --> 00:25:44.900 paramount. Precisely. You always need to integrate 00:25:44.900 --> 00:25:47.690 the objective data with your subjective experience. 00:25:48.130 --> 00:25:50.170 How rested do you actually feel during the day? 00:25:50.230 --> 00:25:52.890 How's your energy, your mood? Does it align with 00:25:52.890 --> 00:25:55.089 what the tracker says? Sometimes the data might 00:25:55.089 --> 00:25:57.930 look bad, but you feel fine or vice versa. Your 00:25:57.930 --> 00:26:00.369 subjective feeling of restfulness is incredibly 00:26:00.369 --> 00:26:02.490 important. And just finally, on tech, you mentioned 00:26:02.490 --> 00:26:04.470 smart mattresses. How are they different? What's 00:26:04.470 --> 00:26:06.950 the promise there? This is where the technology 00:26:06.950 --> 00:26:10.809 starts to move beyond just tracking sleep to 00:26:10.809 --> 00:26:14.059 actively augmenting or improving it. Devices 00:26:14.059 --> 00:26:15.940 like the eight -sleep mattress, for example, 00:26:16.279 --> 00:26:18.380 they don't just measure your sleep stages, heart 00:26:18.380 --> 00:26:20.539 rate, breathing rate. They actually do something 00:26:20.539 --> 00:26:23.240 with that data. Yes. They use it to intelligently 00:26:23.240 --> 00:26:25.819 adjust the temperature of the mattress surface, 00:26:26.119 --> 00:26:28.460 zone by zone, throughout the night. Temperature. 00:26:28.460 --> 00:26:31.339 Why is that important? Because our body temperature 00:26:31.339 --> 00:26:34.299 regulation is incredibly tightly linked to our 00:26:34.299 --> 00:26:37.740 sleep cycle. To fall asleep easily, your core 00:26:37.740 --> 00:26:40.480 body temperature needs to drop slightly. To maintain 00:26:40.480 --> 00:26:42.420 deep sleep, you generally need to stay slightly 00:26:42.420 --> 00:26:44.789 cool. and then your temperature needs to rise 00:26:44.789 --> 00:26:47.190 again towards morning to help you wake up naturally. 00:26:47.410 --> 00:26:49.170 And the mattress can manage that automatically 00:26:49.170 --> 00:26:51.710 based on your individual sleep. That's the idea. 00:26:52.130 --> 00:26:54.569 It learns your patterns, senses your current 00:26:54.569 --> 00:26:56.690 sleep stage and physiology minute by minute, 00:26:57.130 --> 00:26:59.309 and then subtly adjusts the heating or cooling 00:26:59.309 --> 00:27:02.150 of the mattress surface to keep you in your optimal 00:27:02.150 --> 00:27:04.880 thermal zone for each stage of sleep. It's a 00:27:04.880 --> 00:27:07.259 really elegant use of technology, creating this 00:27:07.259 --> 00:27:10.799 constant personalized thermal feedback loop to 00:27:10.799 --> 00:27:12.960 improve sleep quality and maybe even quantity. 00:27:13.160 --> 00:27:15.880 That sounds quite sophisticated. It is. And they're 00:27:15.880 --> 00:27:17.839 adding other features, too, like detecting snoring 00:27:17.839 --> 00:27:20.140 and then maybe subtly adjusting the head elevation 00:27:20.140 --> 00:27:22.299 to help open the airway. Yeah. These kinds of 00:27:22.299 --> 00:27:24.599 smart beds, I think, represent a really promising 00:27:24.599 --> 00:27:27.480 new direction for actually enhancing sleep, not 00:27:27.480 --> 00:27:29.940 just measuring it. OK. That brings us to the 00:27:29.940 --> 00:27:32.980 end of our deep dive into sleep science. Some 00:27:32.980 --> 00:27:35.500 truly surprising stuff there. The social impacts, 00:27:35.720 --> 00:27:38.539 the genetic quirks, the brain's cleaning system, 00:27:38.799 --> 00:27:41.299 how meds affect it, the importance of quality 00:27:41.299 --> 00:27:44.319 and regularity, and how technology fits in. It 00:27:44.319 --> 00:27:46.299 certainly shows how much more we're learning 00:27:46.299 --> 00:27:49.039 about this fundamental biological need and just 00:27:49.039 --> 00:27:51.660 how fast the science is moving on so many different 00:27:51.660 --> 00:27:54.890 fronts at once. Right. So... Let's shift gears 00:27:54.890 --> 00:27:57.009 now. We're staying with cutting edge science, 00:27:57.529 --> 00:28:00.250 but broadening the view to the whole future of 00:28:00.250 --> 00:28:02.950 life science, drawing on that conversation with 00:28:02.950 --> 00:28:05.529 Sir John Bell. Yes, a really influential figure, 00:28:06.170 --> 00:28:08.730 long distinguished career in genetics, immunology, 00:28:09.109 --> 00:28:11.390 formerly the Regis Professor of Medicine at Oxford. 00:28:11.869 --> 00:28:13.990 And now he's heading up this significant new 00:28:13.990 --> 00:28:16.549 venture called the Ellison Institute of Technology, 00:28:16.710 --> 00:28:19.890 the EIT, based over here in the UK. OK, tell 00:28:19.890 --> 00:28:21.710 us about the EIT. What's its mission? What makes 00:28:21.710 --> 00:28:23.890 it different? Well, it's incredibly ambitious. 00:28:24.309 --> 00:28:26.390 It's funded by Larry Ellison, the tech entrepreneur. 00:28:27.470 --> 00:28:30.029 And his mission is really broad, using state 00:28:30.029 --> 00:28:31.829 -of -the -art science and technology, especially 00:28:31.829 --> 00:28:34.269 AI and machine learning, to tackle some of the 00:28:34.269 --> 00:28:36.089 world's biggest problems. Not just health? Not 00:28:36.089 --> 00:28:38.490 just health, no. Health care and disease are 00:28:38.490 --> 00:28:41.250 major focuses, obviously, but also critical areas 00:28:41.250 --> 00:28:45.029 like food security, climate change, even things 00:28:45.029 --> 00:28:47.430 like global governance. It's a very wide scope. 00:28:47.569 --> 00:28:49.410 And you mentioned a unique approach or model. 00:28:49.650 --> 00:28:52.779 Yes, that's a key difference. The vision is that, 00:28:52.799 --> 00:28:55.299 where possible, the solutions they develop should 00:28:55.299 --> 00:28:57.359 actually be commercially viable and sustainable. 00:28:57.880 --> 00:28:59.980 It's structured more like a company than a traditional 00:28:59.980 --> 00:29:02.880 academic foundation. The aim is to create tangible 00:29:02.880 --> 00:29:05.779 products, platforms, services that have real 00:29:05.779 --> 00:29:08.839 -world impact and can sustain themselves, rather 00:29:08.839 --> 00:29:11.660 than relying solely on grants or donations. It's 00:29:11.660 --> 00:29:14.829 a more entrepreneurial take on Philanthropic 00:29:14.829 --> 00:29:17.269 science, perhaps. And the scale sounds pretty 00:29:17.269 --> 00:29:19.329 huge. Oh, absolutely massive. They're building 00:29:19.329 --> 00:29:21.670 this enormous campus, potentially millions of 00:29:21.670 --> 00:29:23.970 square feet eventually, and planning to recruit 00:29:23.970 --> 00:29:26.430 around 5 ,000 people. Really top talent from 00:29:26.430 --> 00:29:29.750 across the globe. 5 ,000 people, wow. Yeah. The 00:29:29.750 --> 00:29:31.990 focus is on getting the absolute best minds, 00:29:32.150 --> 00:29:34.630 the top 1 % in their fields, and giving them 00:29:34.630 --> 00:29:37.589 the resources and freedom to take risks on really 00:29:37.589 --> 00:29:40.289 novel, potentially disruptive ideas that might 00:29:40.289 --> 00:29:41.910 struggle to get funding through more traditional 00:29:41.910 --> 00:29:44.190 channels. And it works closely with Oxford University. 00:29:44.289 --> 00:29:46.609 How does that relationship function? Yes, there's 00:29:46.609 --> 00:29:49.069 a very strong synergistic link. Oxford's been 00:29:49.069 --> 00:29:53.009 very supportive. The idea is that the EIT complements 00:29:53.009 --> 00:29:56.410 Oxford's strengths and the wider UK life science 00:29:56.410 --> 00:29:59.670 ecosystem. One specific gap Sir John mentioned 00:29:59.670 --> 00:30:02.809 EIT aims to fill is helping promising UK science 00:30:02.809 --> 00:30:05.470 scale up. Historically the UK has been great 00:30:05.470 --> 00:30:08.289 at discovery but less good at growing those discoveries 00:30:08.289 --> 00:30:11.650 into large impactful companies. EI key structure 00:30:11.650 --> 00:30:13.490 and funding are designed to allow them to build 00:30:13.490 --> 00:30:15.809 full stack solutions taking things right through 00:30:15.809 --> 00:30:17.980 development. That makes sense. And speaking of 00:30:17.980 --> 00:30:20.619 the UK's life science landscape, Sir John Bell 00:30:20.619 --> 00:30:22.980 himself was hugely influential in establishing 00:30:22.980 --> 00:30:25.799 the UK as a world leader in human genomics, wasn't 00:30:25.799 --> 00:30:29.079 he? Absolutely pivotal. His influence was immense. 00:30:29.599 --> 00:30:32.440 There were really three massive, successive initiatives 00:30:32.440 --> 00:30:35.099 that built that leadership in large -scale human 00:30:35.099 --> 00:30:37.940 genomics data. Which were? The UK Biobank, then 00:30:37.940 --> 00:30:40.319 Genomes England, and the newest one, our Future 00:30:40.319 --> 00:30:42.660 Health. He was kind of built on the last tackling 00:30:42.660 --> 00:30:44.740 different aspects of using genomic information. 00:30:44.910 --> 00:30:47.269 Let's start with UK Biobank. That's become this 00:30:47.269 --> 00:30:49.849 incredible global resource for researchers. How 00:30:49.849 --> 00:30:51.650 did that come about? Well, it really grew out 00:30:51.650 --> 00:30:54.130 of Oxford's traditional strength in large -scale 00:30:54.130 --> 00:30:56.930 epidemiology, studying disease patterns in big 00:30:56.930 --> 00:31:00.269 populations. The key insight was realizing you 00:31:00.269 --> 00:31:03.009 needed enormous numbers of people, huge cohorts, 00:31:03.309 --> 00:31:05.910 to reliably detect the often small effects of 00:31:05.910 --> 00:31:08.289 individual genetic variations on common diseases. 00:31:08.589 --> 00:31:11.029 So the idea was born to recruit a massive group. 00:31:11.529 --> 00:31:14.069 Exactly. The incredibly ambitious plan was to 00:31:14.069 --> 00:31:16.230 recruit half a million people across the UK, 00:31:16.829 --> 00:31:18.549 collect loads of health data questionnaires, 00:31:18.890 --> 00:31:20.809 physical measures but critically, also collect 00:31:20.809 --> 00:31:23.849 biological samples, like DNA, and get consent 00:31:23.849 --> 00:31:26.289 to link all that to their ongoing NHS health 00:31:26.289 --> 00:31:28.390 records. Half a million people linked to health 00:31:28.390 --> 00:31:30.390 records. That sounds like a logistical nightmare 00:31:30.390 --> 00:31:33.259 to set up. Was it difficult? Extremely difficult. 00:31:33.519 --> 00:31:35.539 Sir John talked about facing massive resistance 00:31:35.539 --> 00:31:38.440 initially from peer reviewers. People doubted 00:31:38.440 --> 00:31:40.500 it was even feasible. Could you recruit that 00:31:40.500 --> 00:31:43.940 many? Get reliable data, manage the NHS linkage, 00:31:44.160 --> 00:31:46.299 even do the genetics on that scale back then. 00:31:46.619 --> 00:31:48.940 But they persevered, got crucial funding from 00:31:48.940 --> 00:31:52.119 the MRC and Wellcome Trust, and built this resource 00:31:52.119 --> 00:31:55.359 that is now genuinely used by tens of thousands 00:31:55.359 --> 00:31:58.039 of researchers around the world every single 00:31:58.039 --> 00:32:00.460 day. It completely changed the game for genetic 00:32:00.460 --> 00:32:04.210 epidemiology. Amazing achievement. And then came 00:32:04.210 --> 00:32:06.970 Genomes England. That shifted the focus more 00:32:06.970 --> 00:32:09.410 towards using genomics in actual clinical care. 00:32:09.609 --> 00:32:12.190 Yes, exactly. As the cost of sequencing a whole 00:32:12.190 --> 00:32:14.589 genome just plummeted thanks to new technology, 00:32:15.130 --> 00:32:18.210 the idea emerged. Could we use this powerful 00:32:18.210 --> 00:32:21.230 tool right now within the NHS to help diagnose 00:32:21.230 --> 00:32:23.710 rare diseases? Because many rare diseases have 00:32:23.710 --> 00:32:26.670 a genetic cause, but finding it is hard. Precisely. 00:32:27.029 --> 00:32:28.890 It can be like finding a needle in a genetic 00:32:28.890 --> 00:32:31.410 haystack. Whole genome sequencing gives you the 00:32:31.410 --> 00:32:33.710 whole code. This project got really important 00:32:33.710 --> 00:32:36.309 backing from the then Prime Minister, David Cameron, 00:32:36.809 --> 00:32:38.710 partly driven apparently by his own family's 00:32:38.710 --> 00:32:40.480 experience with a rare condition. And did that 00:32:40.480 --> 00:32:43.160 face resistance, too? Oh, yes. So John mentioned 00:32:43.160 --> 00:32:45.240 massive resistance, again, this time from parts 00:32:45.240 --> 00:32:47.299 of the medical establishment, perhaps understandably 00:32:47.299 --> 00:32:49.920 wary of the complexity, the cost, the interpretation 00:32:49.920 --> 00:32:53.500 challenges. But again, they pushed through, sequenced 00:32:53.500 --> 00:32:56.819 100 ,000 genomes from patients with rare diseases 00:32:56.819 --> 00:32:59.859 and their families, made huge numbers of diagnoses. 00:33:00.039 --> 00:33:02.839 And it continues today with things like a program 00:33:02.839 --> 00:33:05.400 for sequencing newborns to spot treatable conditions 00:33:05.400 --> 00:33:09.029 incredibly early. And the third and most recent 00:33:09.029 --> 00:33:12.190 big initiative is our future health. What's the 00:33:12.190 --> 00:33:14.950 ambition there? Our future health really represents 00:33:14.950 --> 00:33:17.369 another shift in focus, moving towards prevention 00:33:17.369 --> 00:33:20.069 and early diagnosis. It came out of the UK's 00:33:20.069 --> 00:33:23.230 life science strategy back in 2017. A key push 00:33:23.230 --> 00:33:25.589 came from the pharmaceutical industry, who basically 00:33:25.589 --> 00:33:27.970 said the UK needs to focus on the future of health 00:33:27.970 --> 00:33:29.869 care, stopping people getting sick in the first 00:33:29.869 --> 00:33:32.710 place or catching diseases much, much earlier. 00:33:32.829 --> 00:33:35.089 Moving away from just treating established disease. 00:33:35.289 --> 00:33:38.309 Exactly. Moving decisively towards identifying 00:33:38.309 --> 00:33:41.230 people at high risk before symptoms appear or 00:33:41.230 --> 00:33:43.549 diagnosing conditions at stage zero or stage 00:33:43.549 --> 00:33:46.029 one when treatments are often far more effective. 00:33:46.589 --> 00:33:50.289 The scale here is just enormous. The goal is 00:33:50.289 --> 00:33:53.720 to recruit five million adults in the UK. Five 00:33:53.720 --> 00:33:56.359 million. That's about 10 % of the adult population. 00:33:56.519 --> 00:33:59.000 It is. It's staggering. The idea is to track 00:33:59.000 --> 00:34:01.460 their health over many years, combine genetic 00:34:01.460 --> 00:34:04.339 data with lots of other health information, lifestyle 00:34:04.339 --> 00:34:06.720 data, maybe even data from wearables. To do what, 00:34:06.819 --> 00:34:09.400 exactly? Primarily to build incredibly detailed 00:34:09.400 --> 00:34:12.099 models of who is likely to develop which diseases, 00:34:12.460 --> 00:34:14.480 enabling much earlier identification of risk. 00:34:14.960 --> 00:34:17.179 And secondly, to create a massive platform for 00:34:17.179 --> 00:34:19.480 efficiently testing new preventative therapies 00:34:19.480 --> 00:34:22.130 or early diagnostic tools on a huge scale. And 00:34:22.130 --> 00:34:24.389 has recruitment going for that. Despite delays 00:34:24.389 --> 00:34:26.409 due to the pandemic, it's been incredibly fast. 00:34:27.030 --> 00:34:29.309 Millions have already registered and large numbers 00:34:29.309 --> 00:34:31.469 have provided samples and data. It's going to 00:34:31.469 --> 00:34:34.630 generate this unbelievably rich multiomics data 00:34:34.630 --> 00:34:37.750 set. Multiomics. So beyond just the genome sequence. 00:34:38.150 --> 00:34:41.280 Yes. much beyond. Looking at proteins, proteomics, 00:34:41.679 --> 00:34:44.440 metabolites, metabolomics, maybe epigenetic markers, 00:34:44.900 --> 00:34:47.380 even things like organ clocks, biological markers, 00:34:47.480 --> 00:34:49.900 trying to estimate the functional age of different 00:34:49.900 --> 00:34:52.699 organs. It's about building a much more holistic, 00:34:52.820 --> 00:34:55.460 dynamic picture of health and disease risk over 00:34:55.460 --> 00:34:58.239 time. That sounds like it could truly revolutionize 00:34:58.239 --> 00:35:00.889 preventative medicine. But it must require new 00:35:00.889 --> 00:35:03.610 ways of thinking, clinically and regulatory -wise. 00:35:04.010 --> 00:35:06.630 Absolutely. You need new types of clinical trials 00:35:06.630 --> 00:35:09.030 designed for prevention in healthy populations. 00:35:09.590 --> 00:35:11.750 You need regulators to think about approving 00:35:11.750 --> 00:35:14.250 diagnostics based on predicting future risk, 00:35:14.730 --> 00:35:17.030 not just confirming present disease. It's a huge 00:35:17.030 --> 00:35:19.789 shift. And driving a lot of this potential, you 00:35:19.789 --> 00:35:22.610 mentioned AI is absolutely central to the Ellison 00:35:22.610 --> 00:35:25.510 Institute's vision. Yes. AI is seen as a core 00:35:25.510 --> 00:35:27.889 enabling platform, underpinning pretty much everything 00:35:27.889 --> 00:35:29.710 they do, whether it's health, climate, food, 00:35:29.869 --> 00:35:31.889 whatever. And they have a massive advantage here 00:35:31.889 --> 00:35:34.730 because of the huge compute power funded by Larry 00:35:34.730 --> 00:35:37.050 Ellison. Access to that level of computing being 00:35:37.050 --> 00:35:39.869 essential for modern AI. Absolutely critical. 00:35:40.289 --> 00:35:42.969 Training these incredibly complex machine learning 00:35:42.969 --> 00:35:45.230 models on the vast data sets we're talking about 00:35:45.230 --> 00:35:49.050 requires enormous computational resources. Few 00:35:49.050 --> 00:35:51.369 universities or traditional institutes can compete 00:35:51.369 --> 00:35:53.329 on that scale. And they're bringing in the talent 00:35:53.329 --> 00:35:56.530 to use it? Yes, actively recruiting top AI and 00:35:56.530 --> 00:35:58.250 machine learning talent from around the world, 00:35:58.610 --> 00:36:00.590 including people from places like Google DeepMind. 00:36:00.889 --> 00:36:04.289 So you have this potent combination, world -class 00:36:04.289 --> 00:36:08.010 talent, massive compute, and access to the UK's 00:36:08.010 --> 00:36:11.320 unique rich healthcare, and genomic data assets. 00:36:11.699 --> 00:36:13.780 Those are the three essential pillars. Talent, 00:36:13.880 --> 00:36:17.139 compute, data. Makes sense. Eric Topol's review 00:36:17.139 --> 00:36:19.920 for the NHS actually highlighted the UK's potential 00:36:19.920 --> 00:36:22.559 to lead in AI for health, didn't it? Building 00:36:22.559 --> 00:36:25.340 on that genomic strength. It did, yes. The potential 00:36:25.340 --> 00:36:27.820 is clearly there. And you mentioned a second 00:36:27.820 --> 00:36:30.760 key technology platform at EIT working alongside 00:36:30.760 --> 00:36:33.900 AI. That's right. Generative biology or synthetic 00:36:33.900 --> 00:36:35.860 biology, as it's often called. That's the other 00:36:35.860 --> 00:36:38.320 core platform. They've recruited a real leader 00:36:38.320 --> 00:36:40.719 in the field, Professor Jason Chin from Cambridge, 00:36:40.900 --> 00:36:43.099 to head that up. Generative or synthetic biology, 00:36:43.159 --> 00:36:44.800 what does that involve? It's essentially about 00:36:44.800 --> 00:36:46.940 designing and building new biological parts, 00:36:47.219 --> 00:36:49.980 devices and systems, or redesigning existing 00:36:49.980 --> 00:36:53.000 ones for useful purposes. So engineering life, 00:36:53.099 --> 00:36:55.119 basically. What kind of applications are we talking 00:36:55.119 --> 00:36:58.429 about? Things like being able to synthesize huge 00:36:58.429 --> 00:37:01.449 stretches of DNA from scratch to encode new functions, 00:37:02.190 --> 00:37:05.349 engineering cells or microbes to act as diagnostics 00:37:05.349 --> 00:37:07.929 or therapeutics, or to produce valuable chemicals, 00:37:08.369 --> 00:37:10.130 even potentially transforming plants to make 00:37:10.130 --> 00:37:12.090 them more nutritious or resilient to climate 00:37:12.090 --> 00:37:14.909 change. It's about using engineering principles 00:37:14.909 --> 00:37:17.849 to manipulate biology in predictable ways. And 00:37:17.849 --> 00:37:21.110 how does that connect with AI? They work synergistically. 00:37:21.289 --> 00:37:24.309 AI can analyze vast amounts of biological data 00:37:24.309 --> 00:37:26.230 to help design these new biological systems, 00:37:26.409 --> 00:37:28.590 predict how they'll behave, and then generative 00:37:28.590 --> 00:37:30.710 biology provides the tools to actually build 00:37:30.710 --> 00:37:33.389 and test those designs in the real world. This 00:37:33.389 --> 00:37:35.829 capability can then feed into all of EIT's main 00:37:35.829 --> 00:37:38.130 focus areas, creating new health solutions, new 00:37:38.130 --> 00:37:40.889 food production methods, new environmental technologies. 00:37:41.230 --> 00:37:43.590 It's moving from just reading biology to actively 00:37:43.590 --> 00:37:46.119 writing and editing it. Okay, let's move to another 00:37:46.119 --> 00:37:48.960 area where Sir John has deep expertise and where 00:37:48.960 --> 00:37:52.420 EIT is making a big push. Understanding the immune 00:37:52.420 --> 00:37:55.320 system. You called it the next great frontier. 00:37:56.159 --> 00:37:58.599 I think it really is. If you think about it, 00:37:58.880 --> 00:38:01.139 our clinical tools for assessing someone's immune 00:38:01.139 --> 00:38:04.599 status are still surprisingly crude. Often it 00:38:04.599 --> 00:38:06.860 boils down to just counting different types of 00:38:06.860 --> 00:38:08.840 white blood cells. Which doesn't tell you much 00:38:08.840 --> 00:38:10.320 about how well they're actually functioning. 00:38:10.539 --> 00:38:13.059 Exactly. The immune system is this incredibly 00:38:13.059 --> 00:38:16.690 complex network of different cell types, T cells, 00:38:16.829 --> 00:38:19.309 B cells, innate immune cells interacting through 00:38:19.309 --> 00:38:22.469 countless molecular signals, antibodies, including 00:38:22.469 --> 00:38:25.530 autoantibodies that can cause disease. It's vastly 00:38:25.530 --> 00:38:28.030 complex. And without a really deep molecular 00:38:28.030 --> 00:38:30.090 understanding of an individual's immune state, 00:38:30.590 --> 00:38:32.809 making big breakthroughs in immune -related diseases 00:38:32.809 --> 00:38:36.789 is really hard. Which covers a huge range. Infections, 00:38:36.949 --> 00:38:39.550 cancer, autoimmune diseases. A massive range. 00:38:40.110 --> 00:38:42.030 Sir John described our current approach in this 00:38:42.030 --> 00:38:44.789 area as sometimes feeling like pinning the tail 00:38:44.789 --> 00:38:47.070 on the donkey. A bit of guesswork in the dark. 00:38:47.349 --> 00:38:50.150 We need a much more detailed map. So what's EIT 00:38:50.150 --> 00:38:52.590 doing specifically to try and build that map 00:38:52.590 --> 00:38:54.699 to get beyond the guesswork? They're establishing 00:38:54.699 --> 00:38:58.519 a very significant long -term project focused 00:38:58.519 --> 00:39:03.219 on deep immunophenotyping, basically characterizing 00:39:03.219 --> 00:39:06.380 immune cells and responses at a really detailed 00:39:06.380 --> 00:39:08.599 molecular level. And how are they going about 00:39:08.599 --> 00:39:11.500 that? What's the methodology? A key part involves 00:39:11.500 --> 00:39:14.219 using human challenge models in a very careful 00:39:14.219 --> 00:39:16.380 and controlled way. Human challenge models? What 00:39:16.380 --> 00:39:18.340 does that mean, exposing people to diseases? 00:39:18.679 --> 00:39:22.000 Yes, essentially. It involves carefully selecting 00:39:22.000 --> 00:39:25.199 healthy volunteers and exposing them to controlled, 00:39:25.360 --> 00:39:28.639 safe doses of specific pathogens like common 00:39:28.639 --> 00:39:31.219 cold viruses or bacteria, or giving them vaccines. 00:39:31.420 --> 00:39:33.239 And the purpose is to watch the immune system 00:39:33.239 --> 00:39:36.619 react in real time. Precisely. It allows researchers 00:39:36.619 --> 00:39:39.320 to observe the entire immune response right from 00:39:39.320 --> 00:39:41.719 the beginning in a controlled setting. They can 00:39:41.719 --> 00:39:43.639 track exactly how different parts of the immune 00:39:43.639 --> 00:39:45.800 system, different cell types, different molecules 00:39:45.800 --> 00:39:48.920 respond over time at a very fine -grained molecular 00:39:48.920 --> 00:39:51.300 level. It gives you much richer, more dynamic 00:39:51.300 --> 00:39:53.920 information than just taking a snapshot blood 00:39:53.920 --> 00:39:56.139 sample from someone who's already sick or even 00:39:56.139 --> 00:39:58.599 someone healthy. And I'm guessing AI is absolutely 00:39:58.599 --> 00:40:00.699 essential for making sense of all the data that 00:40:00.699 --> 00:40:04.159 generates. Utterly essential. The sheer volume 00:40:04.159 --> 00:40:06.300 and complexity of the data from these kinds of 00:40:06.300 --> 00:40:09.000 studies is immense. You're looking at changes 00:40:09.000 --> 00:40:11.800 in gene expression across thousands of genes 00:40:11.800 --> 00:40:14.760 in different immune cells, sequencing the receptors 00:40:14.760 --> 00:40:17.840 on T cells and B cells to understand their diversity, 00:40:18.559 --> 00:40:20.619 measuring hundreds of thousands of different 00:40:20.619 --> 00:40:23.079 proteins and metabolites in the blood. It sounds 00:40:23.079 --> 00:40:26.860 overwhelming. It would be, without powerful computational 00:40:26.860 --> 00:40:31.380 tools. AI and machine learning are critical for 00:40:31.380 --> 00:40:34.500 finding the patterns in that complexity. Understanding 00:40:34.500 --> 00:40:37.039 how all the different components, T cells, B 00:40:37.039 --> 00:40:39.559 cells, innate immunity, antibodies interact. 00:40:40.079 --> 00:40:41.880 Figuring out which interactions actually lead 00:40:41.880 --> 00:40:44.099 to a protective immune response versus those 00:40:44.099 --> 00:40:46.360 that might lead to disease or side effects. So 00:40:46.360 --> 00:40:48.900 the ultimate goal is to build a kind of predictive 00:40:48.900 --> 00:40:51.619 model of the immune system. Yes, to build a really 00:40:51.619 --> 00:40:53.980 solid scientific foundation for predicting and 00:40:53.980 --> 00:40:56.219 anticipating what kind of immune response you 00:40:56.219 --> 00:40:58.400 need to generate for effective protection against 00:40:58.400 --> 00:41:00.599 a specific infection or for effective therapy 00:41:00.599 --> 00:41:03.000 against cancer or to calm down an autoimmune 00:41:03.000 --> 00:41:05.460 attack. And if you can predict it, you can potentially 00:41:05.460 --> 00:41:07.739 design interventions to achieve it. That's the 00:41:07.739 --> 00:41:11.039 hope. If we understand exactly what a good immune 00:41:11.039 --> 00:41:13.980 response looks like at the molecular level for, 00:41:13.980 --> 00:41:17.550 say, flu, We can design much better vaccines 00:41:17.550 --> 00:41:20.940 to elicit exactly that response. If we know how 00:41:20.940 --> 00:41:23.340 cancer evades the immune system, we can design 00:41:23.340 --> 00:41:25.900 immunotherapies to overcome those evasion tactics. 00:41:26.579 --> 00:41:29.420 Or for autoimmune diseases, if we can pinpoint 00:41:29.420 --> 00:41:31.900 the specific rogue immune cells or pathways, 00:41:32.320 --> 00:41:34.440 we can develop much more targeted therapies. 00:41:34.780 --> 00:41:36.840 Like the recent breakthroughs in treating diseases 00:41:36.840 --> 00:41:39.320 like lupus you mentioned. Exactly. Treatments 00:41:39.320 --> 00:41:42.159 targeting specific types of B cells, like those 00:41:42.159 --> 00:41:44.960 expressing CD19, have shown remarkable results 00:41:44.960 --> 00:41:47.619 in some severe autoimmune diseases, illustrating 00:41:47.619 --> 00:41:49.300 the power of this kind of molecular process. 00:41:49.130 --> 00:41:52.210 precision. It's about moving beyond broad immunosuppression 00:41:52.210 --> 00:41:54.789 to highly targeted interventions. And Sir John 00:41:54.789 --> 00:41:57.510 mentioned this major immunology project at EIT 00:41:57.510 --> 00:42:00.070 is starting soon. Yes. He said they expect data 00:42:00.070 --> 00:42:02.190 collection using these challenge models and deep 00:42:02.190 --> 00:42:04.349 phenotyping to begin within the next six months 00:42:04.349 --> 00:42:07.489 or so. It's a major undertaking, but the potential 00:42:07.489 --> 00:42:09.869 payoff for understanding health and disease is 00:42:09.869 --> 00:42:13.210 enormous. It really is exciting work. All right, 00:42:13.309 --> 00:42:15.309 just finally, the conversation ended with a brief 00:42:15.309 --> 00:42:17.789 reflection on the broader state of global science, 00:42:17.949 --> 00:42:20.429 didn't it? Potential challenges ahead. Yes, there 00:42:20.429 --> 00:42:23.269 was a brief, quite poignant exchange near the 00:42:23.269 --> 00:42:26.409 end, alluding to potential headwinds facing science 00:42:26.409 --> 00:42:29.329 globally, maybe political disruption, challenges 00:42:29.329 --> 00:42:31.809 to international collaboration. Sir John shared 00:42:31.809 --> 00:42:33.929 his perspective on that. And what was his take? 00:42:34.309 --> 00:42:37.349 Was he pessimistic? Not entirely pessimistic, 00:42:37.550 --> 00:42:39.949 but realistic perhaps. He strongly emphasized 00:42:39.949 --> 00:42:42.570 the inherently global nature of science. That 00:42:42.570 --> 00:42:45.289 despite national borders or politics, scientists 00:42:45.289 --> 00:42:47.329 everywhere are fundamentally part of a shared 00:42:47.329 --> 00:42:50.449 community working on common problems. He stressed 00:42:50.449 --> 00:42:52.750 America's indispensable role as a scientific 00:42:52.750 --> 00:42:55.650 powerhouse given its scale and impact. And his 00:42:55.650 --> 00:42:57.989 hope for the future. His view was that the global 00:42:57.989 --> 00:42:59.929 scientific community needs to stick together 00:42:59.929 --> 00:43:02.969 to lean in collectively to support the scientific 00:43:02.969 --> 00:43:05.809 enterprise, hoping that any political disruptions 00:43:05.809 --> 00:43:09.070 are temporary. He made the point that the complex 00:43:09.070 --> 00:43:11.510 structure supporting international science collaborations 00:43:11.510 --> 00:43:15.010 take a long time to build but can be damaged 00:43:15.010 --> 00:43:17.610 quite quickly, underlining the need for continued 00:43:17.610 --> 00:43:20.070 vigilance and support from society and governments 00:43:20.070 --> 00:43:22.969 everywhere. A thoughtful note to end on. Well, 00:43:23.090 --> 00:43:25.250 this has been quite the journey from the unexpected 00:43:25.250 --> 00:43:28.250 ways sleep affects our social lives and the intricate 00:43:28.250 --> 00:43:30.929 mechanics of brain cleaning right through to 00:43:30.929 --> 00:43:33.909 the future being forged by genomics, AI, synthetic 00:43:33.909 --> 00:43:36.909 biology, and a deeper understanding of our own 00:43:36.909 --> 00:43:39.730 immune systems at places like the EIT. It really 00:43:39.730 --> 00:43:42.050 covers a vast landscape, doesn't it? And it makes 00:43:42.050 --> 00:43:44.090 it clear these aren't just abstract topics for 00:43:44.090 --> 00:43:46.210 scientists. They are fundamentally reshaping 00:43:46.210 --> 00:43:48.710 our understanding of health. of aging and they 00:43:48.710 --> 00:43:50.630 have profound implications for how we might live 00:43:50.630 --> 00:43:53.429 and work in the future. The pace of change is 00:43:53.429 --> 00:43:56.070 just incredible. It certainly gives you, the 00:43:56.070 --> 00:43:58.989 listener, a lot to think about. Given these rapid 00:43:58.989 --> 00:44:01.429 advances, this power to understand health at 00:44:01.429 --> 00:44:04.849 such a deep level, how might our personal responsibility 00:44:04.849 --> 00:44:07.090 for managing health change in the coming years? 00:44:07.510 --> 00:44:09.670 And what does all this mean for you professionally, 00:44:10.050 --> 00:44:12.409 especially if you're in an industry touched by 00:44:12.409 --> 00:44:15.630 biology, data, and technology? The ripple effects 00:44:15.630 --> 00:44:17.559 are going to be immense. over the next decade. 00:44:17.739 --> 00:44:20.139 They really are. The future of health is being 00:44:20.139 --> 00:44:23.239 written right now in labs and data centers across 00:44:23.239 --> 00:44:25.659 the world. Absolutely. And if you found this 00:44:25.659 --> 00:44:27.920 deep dive insightful and valuable, please do 00:44:27.920 --> 00:44:29.940 consider taking a moment to rate and share the 00:44:29.940 --> 00:44:31.860 show. It really helps others find it.