WEBVTT - #2093 Health Advice: Who To Trust? - David Gillespie 0:00:01.160 --> 0:00:01.800 Good Day team. 0:00:01.800 --> 0:00:04.640 It's the two champions, it's the two standout superstars of 0:00:04.680 --> 0:00:09.000 the U Project. And that doesn't include Tiff, even though 0:00:09.400 --> 0:00:11.440 IF's editing this, so I feel like I could get 0:00:11.480 --> 0:00:15.680 in trouble. Shout out to you anyway, Tiff, David, James Gillespie, Craig, 0:00:15.720 --> 0:00:18.919 Anthony Harper. Here at your service, folks. So what I firstly, 0:00:19.120 --> 0:00:21.000 hi mate? How are you good? 0:00:21.079 --> 0:00:21.400 How are you? 0:00:21.560 --> 0:00:21.880 How are you? 0:00:22.480 --> 0:00:25.840 It's good. I'm going to sound like a little winger 0:00:25.880 --> 0:00:27.880 for a moment, but it's not the worst wind. It's 0:00:27.920 --> 0:00:33.080 forty four degrees in Melbourne, so we are pretty hot 0:00:33.120 --> 0:00:36.519 as fires all over Victoria at the moment. What's the 0:00:36.640 --> 0:00:38.800 temperature up there in the Sunshine State. 0:00:39.200 --> 0:00:41.879 I think it's about thirty, but it was pretty hot yesterday. 0:00:42.080 --> 0:00:47.040 It's like thirty eight or something yesterday. But yeah, when 0:00:47.040 --> 0:00:49.559 I used to live in Victoria, I do remember not 0:00:49.680 --> 0:00:52.280 liking that particular aspect of it is that most of 0:00:52.320 --> 0:00:56.320 the time it's pretty nice, but when it gets hot, 0:00:56.400 --> 0:00:57.760 it really gets off. 0:00:57.880 --> 0:00:59.240 Yeah, it's yeah. 0:00:59.280 --> 0:01:02.960 Well the one days the predicted top well, it's still 0:01:02.960 --> 0:01:04.920 going to get hotter. They reckon, it's going to get 0:01:04.959 --> 0:01:09.959 forty six at about six o'clock, but the prediction was, well, 0:01:09.959 --> 0:01:11.400 the forecast was forty seven. 0:01:11.440 --> 0:01:15.360 I'm like, who does forty seven? Anyway, we're here. 0:01:15.840 --> 0:01:18.080 As long as everyone's safe, and everyone's and all the 0:01:18.120 --> 0:01:22.360 animals and forests are not getting burnt and devastated, that's 0:01:22.400 --> 0:01:27.320 all we care about. So I sent your message before 0:01:27.319 --> 0:01:29.320 and you went, I said, well, what are we talking about? 0:01:29.360 --> 0:01:29.959 What have you written? 0:01:29.959 --> 0:01:31.959 And you went, nothing, put on your thinking cap And 0:01:32.000 --> 0:01:34.319 I went, you know what I'm doing. I'm getting AI 0:01:34.400 --> 0:01:38.440 to put on its thinking cap. Oh no, so I 0:01:38.480 --> 0:01:41.199 said to Chatters. So if this is terrible, everyone don't 0:01:41.200 --> 0:01:44.240 blame me. It couldn't possibly be me. It's either David 0:01:44.720 --> 0:01:48.320 or chat GPT. So you're welcome, I said to Chatters, 0:01:48.360 --> 0:01:51.360 if you not me, because it always if I ask 0:01:51.440 --> 0:01:53.640 it stuff, it'll put its itself in my place. 0:01:53.880 --> 0:01:54.480 If you not me. 0:01:54.560 --> 0:01:57.480 Were interviewing author, researcher and lawyer David Gillespie, what would 0:01:57.480 --> 0:01:58.400 you ask him? 0:01:58.680 --> 0:01:59.560 And then I don't know why. 0:01:59.600 --> 0:02:02.960 It brought up sweet poison, like you've only ever done 0:02:03.000 --> 0:02:07.280 one thing, like there's a few other there's some deviations, 0:02:07.280 --> 0:02:09.359 but if I need to modified, I will. 0:02:10.720 --> 0:02:11.040 Craig. 0:02:11.160 --> 0:02:12.080 This is what he said to me. 0:02:12.240 --> 0:02:15.960 He or she if I were interviewing David Gilespiotim for 0:02:16.040 --> 0:02:18.160 questions that go beyond what he thinks and dig into 0:02:18.200 --> 0:02:21.040 how he thinks, why he changed his mind, and where 0:02:21.080 --> 0:02:26.079 people most misunderstand him. Here's how I'd structure it, origin, 0:02:26.200 --> 0:02:32.760 story and epistemology. I'm like, thanks, chatters, subtitle how a 0:02:32.880 --> 0:02:37.919 lawyer ends up waging war on sugar and big narratives. 0:02:37.560 --> 0:02:39.000 If these are ship will be in. 0:02:39.120 --> 0:02:39.960 But we'll see how we go. 0:02:40.560 --> 0:02:41.919 You were trained as a lawyer, not. 0:02:41.880 --> 0:02:44.560 A nutritionist or a neuroscientist. How did that shape the 0:02:44.600 --> 0:02:50.320 way you evaluate evidence compared to experts trained in those fields. 0:02:52.040 --> 0:02:54.320 It's not a bad. 0:02:54.480 --> 0:02:56.280 Well done chatters, Yeah, it's not bad. It's not bad. 0:02:56.320 --> 0:02:57.320 Good chatters. 0:02:57.960 --> 0:03:01.560 Well, I guess the important thing to know is that 0:03:01.639 --> 0:03:05.440 I don't have a dog in the fight, and lawyers 0:03:05.520 --> 0:03:06.079 rarely do. 0:03:06.360 --> 0:03:06.600 You know. 0:03:06.639 --> 0:03:11.720 There's that old saying, you know, if you represents yourself, 0:03:12.080 --> 0:03:19.560 you've got an idiot as a client. And that's kind 0:03:19.639 --> 0:03:22.359 of one of I think the most valuable things I 0:03:22.480 --> 0:03:26.079 bring to anything I look at. And it's also why 0:03:26.080 --> 0:03:29.960 I never really write about law. So I can't be 0:03:30.040 --> 0:03:34.160 fired for anything I write because I'm not employed as 0:03:34.200 --> 0:03:38.840 a dietitian or a psychologist or anything like that. I 0:03:39.720 --> 0:03:42.040 don't have a dog in the fight. I'm coming to 0:03:42.080 --> 0:03:46.200 the question and I'm saying, what does the evidence show us? 0:03:46.600 --> 0:03:48.760 And I look at the evidence on both sides of 0:03:48.760 --> 0:03:53.400 the argument and see what it says. And I think 0:03:53.520 --> 0:03:56.280 that's the skill that I'm bringing to the table here, 0:03:56.880 --> 0:04:00.400 which is that a so called expert in the area, 0:04:00.800 --> 0:04:04.400 say in biothet x or nutrition or something, has a 0:04:04.440 --> 0:04:08.520 dog in the fight. They have a job that and 0:04:09.160 --> 0:04:14.200 them keeping their job usually means that they have to 0:04:14.280 --> 0:04:17.599 stick with the consensus, the company line on this stuff. 0:04:17.600 --> 0:04:21.200 So if the consensus is fat makes you fat, then 0:04:21.640 --> 0:04:23.760 those are the glasses they will have on when they 0:04:23.760 --> 0:04:27.680 look at any of the evidence, Whereas I don't care. 0:04:28.680 --> 0:04:32.800 I just want to know what the evidence says. And 0:04:33.040 --> 0:04:35.680 if the evidence said fat makes you fat, then that's 0:04:35.680 --> 0:04:38.280 what I would have said. And nobody would have probably 0:04:38.279 --> 0:04:40.040 published a book written by me, because I would have 0:04:40.080 --> 0:04:41.640 been saying the same thing as everybody else. 0:04:42.680 --> 0:04:45.000 But I don't care about that either. 0:04:44.960 --> 0:04:47.359 In the sense that, as you well know as a 0:04:47.360 --> 0:04:51.799 fellow author, nobody gets rich writing books, and certainly nobody 0:04:51.800 --> 0:04:58.000 gets rich writing books in Australia. Yes, So all I'm 0:04:58.040 --> 0:05:03.080 doing is writing down in terms that people who are 0:05:03.080 --> 0:05:07.080 not experts in the area like me, can understand what 0:05:07.120 --> 0:05:11.800 the evidence actually says. And this is something probably the 0:05:11.839 --> 0:05:15.000 only useful thing that lawyers are trained to do, which 0:05:15.080 --> 0:05:20.800 is they are asked to understand the evidence in all 0:05:20.880 --> 0:05:24.359 manner of disciplines. They have to understand the evidence in 0:05:24.400 --> 0:05:28.600 a car accident, or the evidence in a bridge failure, 0:05:28.880 --> 0:05:33.520 or the evidence in a construction litigation, or the evidence 0:05:34.360 --> 0:05:36.599 you know, in whatever you name it. It's all different, 0:05:36.720 --> 0:05:39.520 and they're trained in one of us. They do have 0:05:39.600 --> 0:05:42.400 to understand the evidence and they have to be able 0:05:42.440 --> 0:05:47.160 to clearly state what the evidence says. And even more 0:05:47.320 --> 0:05:49.880 than that, and this is a side that most people 0:05:49.920 --> 0:05:52.919 don't see in lawyers, is they have an obligation to 0:05:52.960 --> 0:05:57.200 the court to present evidence to the court which contradicts 0:05:58.000 --> 0:06:02.920 their case, obligation not to mislead the court. So if 0:06:02.960 --> 0:06:06.760 there is evidence that they are aware of that contradicts 0:06:06.800 --> 0:06:09.000 something they are saying to the court, they must point 0:06:09.000 --> 0:06:15.000 it out. So I guess in that sense, I'm the 0:06:15.040 --> 0:06:18.159 best possible person to look at the questions that I'm 0:06:18.160 --> 0:06:23.400 writing about exactly because I'm not an expert in the area, 0:06:23.520 --> 0:06:26.120 and I am trained in assessing evidence. 0:06:26.880 --> 0:06:29.760 That is such a good answer, Tode. I hate paying 0:06:29.800 --> 0:06:33.120 you a compliment, but it's very good. Like I think 0:06:33.160 --> 0:06:40.520 also this because people firstly, you're possibly our most if 0:06:40.560 --> 0:06:44.279 not popular, definitely in the top three, right. I mean 0:06:44.279 --> 0:06:46.080 they are probably a couple of people who don't like you. 0:06:46.120 --> 0:06:48.560 There are thousand who don't like me. Who cares, right, 0:06:48.640 --> 0:06:51.840 But the thing about you is that you're not emotional 0:06:52.520 --> 0:06:54.680 and you don't have you said, you don't have a 0:06:54.680 --> 0:06:56.480 dog in the fight for me. You don't have any 0:06:57.480 --> 0:07:00.880 financial incentive to lean one way or the like. Whatever 0:07:00.920 --> 0:07:03.600 the data says, whatever, wherever the evidence goes. What a 0:07:03.680 --> 0:07:06.160 but buck up fill in the blank. It just is 0:07:06.160 --> 0:07:09.320 what it is. The next thing I want you to clarify, 0:07:09.360 --> 0:07:11.480 which I know and you know the answer to clearly. 0:07:12.200 --> 0:07:16.600 But when you say the evidence for this stuff, just clarify. 0:07:16.760 --> 0:07:20.200 For our listeners who are not scientists or researchers or 0:07:20.560 --> 0:07:24.520 gatherers of evidence in this sense, where do you get 0:07:24.560 --> 0:07:30.800 your evidence? I'm guessing not YouTube or Instagram, No. 0:07:32.320 --> 0:07:35.960 But mostly because those are media that people are using 0:07:36.000 --> 0:07:41.560 to summarize evidence. Hopefully they're not really presenting the original 0:07:41.600 --> 0:07:45.240 evidence because it would bore people rigid and no one 0:07:45.240 --> 0:07:50.200 would watch their YouTube channel. Yes, So what I mean 0:07:50.640 --> 0:07:53.360 by the evidence is I go to the source to 0:07:53.960 --> 0:07:58.480 the best quality of evidence that is available. So if 0:07:58.480 --> 0:08:01.679 I often, i'll see an artist call online and maybe 0:08:01.680 --> 0:08:03.880 it is even on Instagram post. I'm not sure, you know, 0:08:03.920 --> 0:08:06.320 but I'll see something that said that makes a claim. 0:08:06.960 --> 0:08:08.640 Like I'll give you an example. I saw one today. 0:08:08.880 --> 0:08:11.440 So there was an article I think was on Instagram. 0:08:11.600 --> 0:08:14.480 It's been doing the rounds. It says AI makes you dumber, 0:08:14.880 --> 0:08:19.320 all right, and has The one I saw on Instagram 0:08:19.360 --> 0:08:21.600 had a sort of a picture of a brain scan 0:08:21.720 --> 0:08:24.320 of someone who had apparently been using AI next to 0:08:24.400 --> 0:08:27.600 a picture of a brain scan of someone who hadn't 0:08:27.680 --> 0:08:31.160 been using AI, and it was much more colorful, so 0:08:31.240 --> 0:08:33.840 their brain was working better, apparently, and the other persons 0:08:33.840 --> 0:08:38.600 were just a dull gray now, you know. And so 0:08:39.240 --> 0:08:43.240 I thought, well, that sounds interesting. Is where is that 0:08:43.320 --> 0:08:47.160 coming from and what is the evidence underneath it? And 0:08:47.200 --> 0:08:50.120 so I went look. I went looking for what is 0:08:50.160 --> 0:08:52.439 this based on? And it is based on a study 0:08:53.400 --> 0:08:56.760 done by an American university I can't remember which one, 0:08:56.760 --> 0:08:59.520 but quite a reputable one. The study was only about 0:08:59.520 --> 0:09:07.280 forty p and it was that they found that in 0:09:07.320 --> 0:09:11.360 some circumstances they were able to test in a variety 0:09:11.360 --> 0:09:15.120 of ways and find that, for example, people had less 0:09:15.160 --> 0:09:17.880 recall when they used an AI to write something than 0:09:17.880 --> 0:09:19.040 if they wrote it themselves. 0:09:19.720 --> 0:09:21.079 And I'm kind of. 0:09:21.040 --> 0:09:26.319 Thinking, well, the yes, that seems to be logical. If 0:09:26.360 --> 0:09:28.520 you didn't write it, you're probably not going to remember 0:09:28.559 --> 0:09:30.559 what you didn't write, of. 0:09:30.679 --> 0:09:35.880 Course, whereas it's such a fucking dumb like claim out 0:09:35.920 --> 0:09:39.000 of that or that particular research design. 0:09:39.559 --> 0:09:44.920 Yeah, and so sure I did look at something. It 0:09:45.000 --> 0:09:48.640 got my attention initially in a popular media. But what 0:09:48.720 --> 0:09:51.160 I just did there is what I tend to do 0:09:51.200 --> 0:09:53.720 when I'm looking for research, which just keep drilling down 0:09:54.200 --> 0:09:57.360 until you find the source and then make judgments about 0:09:57.400 --> 0:09:58.600 the quality of that source. 0:09:59.120 --> 0:10:00.360 So first of all, all. 0:10:00.800 --> 0:10:05.640 You know handily journals are rated, so you can you 0:10:05.679 --> 0:10:08.480 don't have to guess is this a reputable source of information? 0:10:09.679 --> 0:10:14.280 They have ratings and you can decide, okay, well this 0:10:14.360 --> 0:10:17.280 is not a particularly highly rated journal. So I'm already 0:10:18.080 --> 0:10:20.400 a little bit cautious about whether this has been peer 0:10:20.440 --> 0:10:22.240 reviewed or not. So the particular study I was talking 0:10:22.240 --> 0:10:27.960 about there with AI not peer reviewed. So so straight 0:10:28.000 --> 0:10:30.440 away you're thinking, but that doesn't mean throw it in 0:10:30.440 --> 0:10:33.559 the bin. It just means start being really, really cautious 0:10:34.000 --> 0:10:38.360 and questioning about anything that's being said and put your 0:10:38.400 --> 0:10:39.640 critical thinking head on. 0:10:39.880 --> 0:10:41.880 Okay, and. 0:10:43.760 --> 0:10:46.160 So you apply that sort of logic to it, and 0:10:46.200 --> 0:10:49.280 then you start looking at, well, who's paying for this study? 0:10:49.559 --> 0:10:52.560 Do they have a vested interest in the outcome? All 0:10:52.600 --> 0:10:55.720 those kinds of questions as you're looking at it, and 0:10:55.760 --> 0:10:58.480 then if it withstands all of that, you start going 0:10:58.480 --> 0:11:01.600 to look for other research, and well, what other This 0:11:01.640 --> 0:11:04.360 can't possibly be the only person who's looked at this question. 0:11:04.440 --> 0:11:05.040 It never is. 0:11:05.080 --> 0:11:09.199 Thousands of studies are published every day. What do other 0:11:09.240 --> 0:11:13.160 people say about this particular area, What is their contradictory evidence? 0:11:13.160 --> 0:11:16.000 What does it say? And then I look at all 0:11:16.040 --> 0:11:19.040 of that and I start to form a view as 0:11:19.080 --> 0:11:24.480 to how reliable or otherwise what's being said is. And 0:11:24.520 --> 0:11:29.520 then I might form a hypothesis about whether it's you know, 0:11:29.800 --> 0:11:33.160 whether A plus B equals C, and then I'll start 0:11:33.200 --> 0:11:36.360 trying to prove it wrong. So I'll start I'll start 0:11:36.400 --> 0:11:38.840 looking for research that might prove that what I just 0:11:38.920 --> 0:11:43.040 said A plus B equal C is wrong. And if 0:11:43.040 --> 0:11:46.840 I can't find any despite my best efforts, then I 0:11:46.920 --> 0:11:49.960 start to think, okay, well this is starting to stack 0:11:50.040 --> 0:11:52.559 up logically. I can't find anything that says it's wrong. 0:11:52.840 --> 0:11:55.080 Can I find things that suggest that elements of this 0:11:55.200 --> 0:11:59.000 is right? Because you rarely ever find one that says 0:11:59.040 --> 0:12:01.200 all of it together. You say, well, if A is 0:12:01.200 --> 0:12:03.480 true and B is true, then C must be true. 0:12:04.320 --> 0:12:07.320 And so what I'll just describe to you there is 0:12:07.600 --> 0:12:10.800 kind of the process I go through over and over 0:12:10.880 --> 0:12:17.559 and over again with everything I look at, and it's 0:12:17.559 --> 0:12:20.520 an almost automatic thing that I do. And then every 0:12:20.559 --> 0:12:24.000 now and then, one in one hundred times I might 0:12:24.320 --> 0:12:26.080 find something that I think, you know, well, actually this 0:12:26.200 --> 0:12:30.120 is just surprising. This is the opposite of what I 0:12:30.200 --> 0:12:34.080 had believed and the opposite of what I think what 0:12:34.200 --> 0:12:36.719 most people believe. And then I'm all right about it. 0:12:38.000 --> 0:12:40.360 Yeah, it such a good process. 0:12:40.760 --> 0:12:45.560 I didn't even know until I started my PhD that 0:12:45.600 --> 0:12:49.120 there were generally there's like Q one through to Q 0:12:49.320 --> 0:12:51.640 four or Tier one to tier four papers which you 0:12:51.679 --> 0:12:54.160 know about. And just to give our listeners an idea, 0:12:54.960 --> 0:12:59.320 like to get accepted and published in a Tier one 0:12:59.440 --> 0:13:00.480 academic journal. 0:13:01.280 --> 0:13:02.640 It depends on the journal. 0:13:02.720 --> 0:13:06.200 They vary, but some of them take less than two 0:13:06.280 --> 0:13:09.600 percent of the papers that get submitted, and these are 0:13:09.640 --> 0:13:13.160 all obviously peer reviewed, and then down to Q four 0:13:13.400 --> 0:13:16.440 so the lowest level, which is where they're often not 0:13:16.559 --> 0:13:19.320 peer reviewed, and the way that they get published is 0:13:19.360 --> 0:13:22.640 by the authors paying for it to be published. 0:13:23.000 --> 0:13:27.199 Yes, you know play Yeah, so it's like oh. 0:13:26.960 --> 0:13:30.080 And then when you so it's and people don't know that, 0:13:30.240 --> 0:13:31.200 not that that's an insult. 0:13:31.200 --> 0:13:32.520 People aren't expected to know that. 0:13:32.600 --> 0:13:34.760 So when you go, oh, this was published in a 0:13:34.800 --> 0:13:38.120 scientific journal that could be analogous to the fucking Woman's 0:13:38.200 --> 0:13:41.240 Day or new Idea like it to be. Yeah, that's 0:13:41.559 --> 0:13:46.400 that's cool, but let's wish what was what was the journal? 0:13:46.400 --> 0:13:48.920 And some on. But even with your forty people. 0:13:48.679 --> 0:13:51.800 Research, you go, so, there's eight billion people in the 0:13:51.800 --> 0:13:55.840 world and we've done some research with forty it's that's 0:13:55.920 --> 0:13:59.439 not necessarily going to extrapolate or you know, and the other. 0:13:59.280 --> 0:14:02.360 Thing there's a problem with that than just this. I mean, 0:14:02.400 --> 0:14:07.280 small research can give you interesting insights that you might 0:14:07.320 --> 0:14:09.520 not other have seen, particularly when it's done in humans. 0:14:11.040 --> 0:14:13.520 So but then you start to look at, well, what 0:14:13.600 --> 0:14:19.040 exactly did they do? And if it becomes you look 0:14:19.080 --> 0:14:22.000 at it and you think, well, yeah, well that's kind 0:14:22.000 --> 0:14:26.120 of obvious, you know, And what are they proving by 0:14:26.280 --> 0:14:29.080 by showing that people who didn't write things don't know 0:14:29.120 --> 0:14:33.600 what they didn't write, you know, I guess they are. 0:14:33.520 --> 0:14:35.520 Proving that those people didn't think much. 0:14:36.760 --> 0:14:43.120 But so what yes, yeah, well I think also there 0:14:43.120 --> 0:14:45.640 are so many things that I never thought about until 0:14:45.640 --> 0:14:49.880 I started this journey, but like the quality of the participants, 0:14:50.040 --> 0:14:51.520 or the demographic. 0:14:50.960 --> 0:14:52.920 Or the age, or the buppa bappa bar fill in 0:14:52.960 --> 0:14:53.640 the blank. Right. 0:14:53.720 --> 0:14:57.760 So with I ran a few projects and in one 0:14:57.760 --> 0:14:59.920 of my projects, I had two hundred and ninety nine 0:15:00.080 --> 0:15:03.880 people and pretty much it was a bit skewed, but 0:15:03.960 --> 0:15:08.400 not a lot. It was like one hundred and forty students, 0:15:09.920 --> 0:15:16.280 so undergrad psychologist students, and the other group was general public. 0:15:16.840 --> 0:15:20.280 So you don't think about the limitations, and you don't 0:15:20.280 --> 0:15:23.640 think about, like when I explain to people, like good 0:15:23.760 --> 0:15:27.840 data and bad data. So this is true because most 0:15:28.480 --> 0:15:33.000 or many many research papers use students as their participants 0:15:33.040 --> 0:15:36.440 because they're so accessible and easy to motivate to do 0:15:36.480 --> 0:15:40.600 the research. But so the problem in inverted commas with 0:15:40.640 --> 0:15:43.960 some of my research is that I had one hundred 0:15:44.000 --> 0:15:47.360 and sixty odd general public who wanted to be there. 0:15:48.040 --> 0:15:50.520 They chose to be there. They're interested in the research, 0:15:50.640 --> 0:15:54.560 interested in the topic. They potentially follow me on social media. 0:15:55.720 --> 0:15:57.760 They were motivated by the fact that I was doing 0:15:57.760 --> 0:15:59.920 a public workshop at the end, which they paid zi 0:16:00.040 --> 0:16:03.440 zo four. How many came because of that, I'm not sure, 0:16:03.880 --> 0:16:07.560 but anyway, the average time, give or take, and this 0:16:07.640 --> 0:16:11.360 is this is written in in the paper, like you 0:16:11.440 --> 0:16:14.560 have to write what the limitations and what problems and 0:16:14.640 --> 0:16:17.400 you know, like in terms of the design, what did 0:16:17.440 --> 0:16:17.920 you fuck up? 0:16:17.960 --> 0:16:18.920 What would you do different? 0:16:20.440 --> 0:16:24.760 So the paper to answer all these questionnaires like paper 0:16:24.840 --> 0:16:27.440 questionnaires with a pen and paper took if you did 0:16:27.440 --> 0:16:31.160 it properly, about fifty minutes. And so the general public, 0:16:31.760 --> 0:16:33.640 it took all of them somewhere in the forty to 0:16:33.720 --> 0:16:34.800 sixty minute range. 0:16:35.560 --> 0:16:36.920 And then I won't. 0:16:36.720 --> 0:16:39.479 Even say how many, but a lot of the students 0:16:40.040 --> 0:16:43.880 were getting through in under fifteen minutes. And the reason, 0:16:44.120 --> 0:16:48.040 the main reason is because they were motivated to be 0:16:48.080 --> 0:16:51.920 there because they were getting academic credits, you know, and 0:16:52.000 --> 0:16:56.320 so you go, oh, so you have a ranging quality 0:16:56.480 --> 0:16:59.360 of data. And then then you've got to try and 0:16:59.400 --> 0:17:03.520 figure out which you literally remove some of the data 0:17:03.560 --> 0:17:06.760 because some of them didn't even finish the thing properly, 0:17:06.800 --> 0:17:08.679 so you just take that out. But yeah, there's a 0:17:08.760 --> 0:17:12.160 huge range of variables. Another thing that is a red 0:17:12.200 --> 0:17:15.159 flag for me is when you see some kind of 0:17:16.119 --> 0:17:21.680 pseudoscience that says, oh, fifty percent of people respond this 0:17:21.720 --> 0:17:24.359 way if they do that, or they see an improvement 0:17:24.440 --> 0:17:28.119 in that if they take this, and you're like, fifty percent, 0:17:28.240 --> 0:17:32.760 what are the chances that's thirty percent of the population, 0:17:33.080 --> 0:17:37.240 Like it's always an exact number, but in real research, 0:17:37.280 --> 0:17:39.760 it's almost never an exact number. 0:17:40.080 --> 0:17:42.280 Like never I didn't have any. 0:17:42.320 --> 0:17:46.320 Completely you know, like fifty or sixty or it's always 0:17:46.400 --> 0:17:50.119 like thirty seven point four or whatever it was. So yeah, 0:17:50.440 --> 0:17:53.360 and that's that's what you don't realize until you kind 0:17:53.400 --> 0:17:56.399 of open that door and try to understand how it 0:17:56.480 --> 0:17:57.920 works well. 0:17:58.119 --> 0:18:01.119 Data transparency is one of the are big problems in 0:18:01.200 --> 0:18:06.119 research is that good papers make their data available so 0:18:06.160 --> 0:18:09.520 that any other research can go in and use their 0:18:09.600 --> 0:18:12.560 data to see if they get the same results. 0:18:13.080 --> 0:18:14.840 And that's one of the big problems. 0:18:14.880 --> 0:18:16.920 And we talked about this briefly before when I think 0:18:16.920 --> 0:18:20.360 we talked about STATN research is STATIN research is almost 0:18:20.400 --> 0:18:24.840 exclusively controlled by consortium funded by the drug companies that 0:18:24.920 --> 0:18:30.479 make statins, and all of that research is published without 0:18:30.520 --> 0:18:34.760 access to the original data. In fact, it's explicitly and 0:18:34.840 --> 0:18:40.280 on purpose kept secret from any other researcher. So you 0:18:40.320 --> 0:18:45.000 will never ever see papers that contradict the results of 0:18:45.040 --> 0:18:48.719 STATIN papers because no one else can see the data, 0:18:48.760 --> 0:18:52.240 which means you can't tell who did they eliminate and why, 0:18:52.960 --> 0:18:55.919 what exactly did they do to the data set to 0:18:56.000 --> 0:18:58.520 get this result, what did they manipulate about it? And 0:18:58.560 --> 0:19:00.679 we know some of the things that they do, so 0:19:00.720 --> 0:19:03.919 there's a peculiar way that they count whether or not 0:19:03.960 --> 0:19:07.639 someone persisted with the drug or not, which is quite 0:19:07.680 --> 0:19:11.760 unusual in that kind of research. And that's just one 0:19:11.840 --> 0:19:14.320 example where data matters. And here's another one which is 0:19:14.359 --> 0:19:17.600 really important. There was a massive study in the United 0:19:17.640 --> 0:19:23.600 States called the Minnesota County Experiment in I think it 0:19:23.640 --> 0:19:25.960 was done in the nineteen sixties and early nineteen seventies, 0:19:26.359 --> 0:19:30.679 and in that what they were trying to prove was 0:19:30.760 --> 0:19:34.840 that if you took people off the traditional American diet 0:19:34.920 --> 0:19:37.520 at that time, which was heavy and saturated fat and 0:19:37.640 --> 0:19:42.159 replace the saturated fat with margarines and seed oils, you 0:19:42.200 --> 0:19:45.600 would reduce the incidents of heart attacks. Now, the study 0:19:45.880 --> 0:19:49.240 cost an absolute fortune because unlike the study we're talking 0:19:49.280 --> 0:19:51.400 about before with about forty people in it, this one 0:19:51.520 --> 0:19:55.320 had I think almost one hundred thousand people involved and 0:19:55.400 --> 0:19:59.440 it ran for the better part of a decade. So 0:19:59.560 --> 0:20:03.600 this thing was massive and hugely expensive and funded by 0:20:03.640 --> 0:20:10.359 the American government. And it proved nothing in the sense 0:20:10.440 --> 0:20:14.600 that it didn't prove the hypothesis, which is that there'd 0:20:14.600 --> 0:20:17.800 be less heart attacks and the people eating margarine, and 0:20:17.840 --> 0:20:24.359 so they published nothing. So they ran that study and 0:20:24.920 --> 0:20:27.960 produced a paper at the end of it, which was meaningless. 0:20:28.480 --> 0:20:30.639 They didn't produce a paper because I presume they had to, 0:20:31.240 --> 0:20:36.000 but it was meaningless. And recently, I say recently in 0:20:36.040 --> 0:20:39.240 the last decade, the National Institutes of Health in the 0:20:39.359 --> 0:20:43.880 US had access to the original data. So they went 0:20:44.040 --> 0:20:46.440 back looked at the data and said, well, it didn't 0:20:46.480 --> 0:20:49.800 prove the hypothesis, but did it prove anything else? Because 0:20:49.800 --> 0:20:52.200 it's a really important and useful data sets. You should 0:20:52.200 --> 0:20:57.399 make such a huge number of humans and have this effect, 0:20:58.080 --> 0:20:59.080 Did it prove anything? 0:20:59.160 --> 0:21:00.040 And what they found is. 0:21:00.840 --> 0:21:05.920 That it didn't. Actually, the people who were on the 0:21:05.960 --> 0:21:09.160 traditional diet had better heart disease outcomes than the people 0:21:09.240 --> 0:21:14.399 whose diet was changed, and the worse than that. The 0:21:14.440 --> 0:21:18.920 all cause mortality was much higher in the changed group. 0:21:19.600 --> 0:21:25.920 So reducing people's cholesterol, which was the aim, actually got 0:21:26.000 --> 0:21:29.440 worse outcomes, not just from heart disease but from everything. 0:21:30.080 --> 0:21:35.199 Now everything is usually code for cancer, so there were 0:21:35.280 --> 0:21:38.600 more people dying of cancer, more people are dying of 0:21:38.600 --> 0:21:42.600 heart disease. When they re looked at the data, and 0:21:43.119 --> 0:21:47.040 that was the research that got published by the National 0:21:47.119 --> 0:21:49.120 Institutes of Health about a decade ago. 0:21:49.800 --> 0:21:52.920 Now that's pretty big news, I would have thought. 0:21:53.520 --> 0:21:55.480 I mean, you think about. 0:21:56.640 --> 0:21:59.280 In reality, what they've done to those people by making 0:21:59.320 --> 0:22:03.159 the meat this alted diet, which was allegedly better and 0:22:03.240 --> 0:22:07.760 healthy and produce better health outcomes, was they overall increased 0:22:07.760 --> 0:22:09.240 morbidity and mortality. 0:22:09.640 --> 0:22:10.760 So it was bad. 0:22:10.880 --> 0:22:13.879 It was literally bad for people, but they buried that. 0:22:14.640 --> 0:22:20.080 I mean, look about talk about fucking unethical, if not evil. 0:22:20.200 --> 0:22:25.800 Imagine knowing that this is actually bad for people, but then, 0:22:25.880 --> 0:22:28.720 because of your ego or your commitment to some kind 0:22:28.760 --> 0:22:33.359 of authority or financial backing or whatever, you hide that 0:22:33.640 --> 0:22:35.639 actual evidence. 0:22:35.200 --> 0:22:38.720 Of what you know to be true for whatever range 0:22:38.760 --> 0:22:40.639 of reasons, like so immoral. 0:22:42.200 --> 0:22:46.440 Yeah, it's sorry. I don't want to mislead people. 0:22:46.480 --> 0:22:48.880 So just while you were rabbiting on there probably saying 0:22:49.359 --> 0:22:53.359 really important, I actually looked up the study and I 0:22:53.400 --> 0:22:55.840 want to be really accurate with some of the things 0:22:55.840 --> 0:22:57.680 that I just said to you there. So it wasn't 0:22:57.800 --> 0:23:00.639 one hundred thousand. The cohort was nine douns four hundred 0:23:00.640 --> 0:23:04.840