Sean Derrig.mp3 Sean Derrig: [00:00:00] But nobody's actually proved that the parachute is an effective intervention and indeed there are numerous injuries from parachutes every year. So until we have a randomized controlled trial, we can't actually say the parachutes are an effective intervention when dealing with the protective against gravitational trauma. Harpreet Sahota: [00:00:40] What's up, everybody? Welcome to the artists of Data Science podcast, the only self development podcast for Data scientists. You're going to learn from and be inspired by the people, ideas and conversations that'll encourage creativity and innovation in yourself so that you can do the same for others. I also host open office hours. You can register to attend by going to bitly.com/adsoh. I look forward to seeing you all there. Let's ride this beat out into another awesome episode. And don't forget to subscribe to the show and leave a five star review. Harpreet Sahota: [00:01:33] Our guest today is a microbiologist specializing in infection control systems, deployment's and novel bioremediation strategies. [00:01:43] He started his career studying biotechnology in the 1980s and has had a number of commercial roles, from startups to multinationals with over 25 years franchising experience. He's a former mainboard director of the British Franchise Association and remains a very practical scientist and has advised many high profile clients on microbiology and biotechnology. He's the CEO and founder of Radical Biotech, a company that is committed to providing sustainable, green, natural alternatives to the harmful and toxic chemicals used in waste treatment that are poisoning the planet. He's not an eco warrior. He's a scientist and he's on a mission. He's also the author of a wildly entertaining and informative blog called Recto Facile Ambiguity, where he takes on the alter ego of Recto FASA, a grumpy microbiologist who thinks writing his blog might be an antidote to all the stupid on the Internet. And apparently Recto Fatha is Latin for asshole. So please help me. In welcoming our guest today, a popular speaker, writer and science communicator, the radical Sean Derek Sean. Harpreet Sahota: [00:03:06] Hey, thank you so much for taking time out of your schedule earlier today. Sean, I really appreciate you coming on to the show. That's right. Sean Derrig: [00:03:12] And you just outed me and the which I was expecting to hear. But the blog has been supposedly anonymous for a number of years because I've lined up all the anti-vaxxers and assorted loonies out there. I don't care, bring it on. Harpreet Sahota: [00:03:28] I put you on blast, man. Alright. All right. Hey, so talk to us about your journey. How did you first get interested in microbiology and how did you get to where you are today? Sean Derrig: [00:03:36] Well, I started off I read biotech to start with when I was actually four and I started at uni. So it's still a reasonably young science at that point. I heard about an experiment in 1982 or Cajamarca in 1992 where when I was a kid, diabetics, I used to get insulin from Coull Pig Pancreas because obviously the human stuff wasn't available. Know some people couldn't tolerate it. Some people religious objections are causing some very smart people started breathing pure human insulin, although one was cool and I found out about how they did it. And what you do is you splice the gene for human insulin out of a white blood cell and your cells are pretty much and then you splice it into a bacteria and you basically reprogram the bacteria to what seems was a shit, pure human insulin, which is really cool. The really cool bit, though, is let's say you've got I mean, homebrewing, you've got like a generation of homebrewed world wide open or whatever, and you've got all these bacteria similar to yours, but you've got all these bacteria. There might be half a dozen or even one out of billions that you've actually managed to successfully introduce the insulin gene into. How do you find it? So what they did was they didn't just put the insulin gene in there. Sean Derrig: [00:05:06] They splice a gene for antibiotic resistance in there as well. So any of the bacteria that had taken up this is how you make insulin gene would also have taken up the here's how you deal with penicillin, Gene. So what you do is you've got your job full of billions of bacteria. Let's say one of them has taken up the plasma that you want. You put a spoonful of penicillin, piss off down the pub, have a couple of beers when you come back. The paracetamol have killed all of the other bugs, the ones that you will have been shagging below and multiply and multiply and literally get on with a few beers, come back and you've got a damage on full of insulin, crapping bacteria and just that thing of putting the antibiotic resistance marker on there so you could actually fight this damn thing when you've done the experiment. That is so cool. I want to do that. So I did. And then I kind of move more and more over the dark side of microbiology. So, yeah, that's probably how I came. It was like one experiment and that's what kind of started me on this rather demented path. Harpreet Sahota: [00:06:16] So you mentioned brewing and stuff like a lot of microbiologists that I know, especially around here in my city, they actually also tend to be wine makers and home brewers. The Bruny beer on your own? Sean Derrig: [00:06:28] No, no, no. A few of my mates do. I did toy with the idea of studying brewing science very early on, but no. I mean for a biologists, to my eternal shame, I can't grow anything. Sean Derrig: [00:06:42] I can't even grow weed for Christ's sake, it is ridiculous. So, you know, for biologists, it's pretty embarrassing being completely incapable of cultivating anything at all. Sean Derrig: [00:06:53] So, no, I've I've never really tried that. Harpreet Sahota: [00:06:56] Yeah I tried to - I brewed a couple of batches. And I figured I'm much better drinking it than I am making it. So I just stuck to drinking it. Yeah. So what's the dark side of microbiology that you mentioned? Sean Derrig: [00:07:09] I had a career away from it, from the scientific thinking. It was always that I of commercial stuff. And then about 15 years ago now, I was in the buyout team for a chemical company and then move into high end infection control, high end food, that kind of stuff, rather than just pouring chemicals. And that started me back on the science again. Then I got more and more into using biologicals to do stuff rather than chemicals. And that's why where Radical came from, where I sold out of the business four years ago. And retired. And I was shit at being retired. I was driving the wife nuts. So I started up something where rather than using chemicals to "clean" - technical term - to get clean shit off things, why don't use bacteria? You know, bacteria have been eating our shit for their interviews. They're really good at it. When you think about it, all of the natural processes of decomposition and you know, you see leaf litter in the garden and then it disappears and becomes soil. What do you think's doing that? Magic? No, it's all manner of microorganisms. So you can pretty much any organic matter, you can clear it up using bacteria. Sean Derrig: [00:08:19] You just need to know the right kind of bacteria. You need to know how to grow them, how to put them to sleep and put them in a product that will be stable on the shelf and then make them wake up again when you need to do something. So things like we have terrible problems with fat bugs. Steckel, which is restaurants flushing those increases down drains, blocks up the sewer. So we've got bugs that I love. It looks that will eat kind of used engine oil, for example, out of the concrete floor, that it's just kind of knowing the right bug, which is easy, and then trying to wrangle it into a product. And that's it. Really know basically nature has been designing these products for about three point eight billion years. So the more they work, that kind of cycle is fine. It's completely remarkable because I can miss them completely lacking in imagination. So they keep belching out these noxious chemicals and stuff, whereas the bugs are far better at it and you make them at room temperature. Sean Derrig: [00:09:18] So it's more energy efficient water-saving as well as a big thing in that I can't remember the number. Every couple of minutes a child somewhere dies through lack of access to safe water. We use flush toilets. I know that we can't solve the problem of water security globally overnight, but we can do something. So in a business we're not over here in like good old British pounds, you're probably spending for every urinal you've got. You're probably spending five hundred quid a year in water. So if you've got an office block with three or six urinals on every floor, that mounts up really quickly. So something else we've done is we've taken, you know, the sort of fragrance urinal box. You know, the technical term is "piss blocks". What we've done is we've taken this block sort of substrate and imprisoned about 70 billion friendly friendly bacteria that completely different to us, about 70 billion friendly bacteria in each of these. And these bacteria live there, watersports. They lived with love pest. So what what happens is there's all sorts of other microbiology going on inside the pipe coming out of your eye. There's also significant that because it's a nitrogen source, the rest of it, the ultimate breakdown of the nitrogen is is ammonia, which smells. So we've got some bugs that do everything except that last step of breaking it down, Somavia. So by releasing these into the arrival, they fuck off all of the other stuff. That's in a nutshell. No more smell like will smell. You can turn off the water and that saves money. Sean Derrig: [00:10:59] It saves water because CO2 saving as well. So the companies can put this into their annual account, saying that we've offset this much carbon dioxide, because if you think about it, water comes out of the tap or whatever, about 80 percent of it we send back to the water company sewage. And there's a cost and there's a quantified carbon cost of turning sewage back into potable water. So if we can prove it, there's an office in some ways saving them. I think it's about a thousand pounds a month or something. So they're saving that money. And it's eight tons of CO2 a year, but that's crazy, crazy numbers. So when you talk to people about it, so I can save water, I can save hard cash money and ready to carry bundles and reduce the carbon footprint, where's the catch? Well, we try thinking one of those, but we couldn't. So we just went down the pub instead. So I think that people are falling out of love with nasty chemicals in plastic bottles going around the place on 40 ton articulated lorries, semis. So this kind of stuff really is obvious. They we develop something now with logic where you get a bottle for life made out of recycled stuff. You get through the post sachets of concentrated book, that type of sachets with a liquid, which is quite clever. You put your Sachar into the trigger bottle, top it up with water from your tap, and then it eats everything you spice up. It's not out of showers. Sean Derrig: [00:12:34] It stops smelling all the great stuff the book basically talks to. And when it's empty, you snap off another type of substrate. But in that paper goes in in your paper recycling fill up with water. Because everything we have in common with coal companies is our water is more expensive than yours. Sean Derrig: [00:12:52] So think about it. Every time you do that, you've saved a plastic bottle, big made, you've saved it. Being transported across knows how many miles you've saved a plastic bottle going probably to a landfill. So we're launching this in a month or so because we have no business to business. But we could publish these. So also because using the postal system to send it, even though, yes, there are truck models associated with that, if it's just one envelope with five sachets in that goes through somebody from your that's rather more carbon efficient than five kilos of plastic bottles full of water going over all the truck. Yeah, it just strikes me as being a complete no brainer. So, yeah, that's that's that's kind of what we went do with Radical because I got bored. I thought of all the mistakes I've made in business, which are many. I mean, let me tell you about this. What if I could start with a completely blank slate and just have this big list of stuff that I have screwed up in the past and I don't do any of those. So that's that's kind of where I came from and sort of came out of the house as well. Harpreet Sahota: [00:14:01] So say your mission then saving the planet one day at a time, and you're doing that using pretty much just cool, quote, friendly bacteria. And that's awesome. Sean Derrig: [00:14:12] You mentioned something, but a little bit of SURFAXIN in there. But it's like from there it's it's it's like from coconuts, not not palm oil. Just to give it a little bit of it, all sustainable. It's all night from free range Armatix, orphan's ass cheeks kind of thing. So, you know, like departmentally friendly. Harpreet Sahota: [00:14:31] So what exactly is a Fatburg and how are they like poisoning and damaging... Sean Derrig: [00:14:36] So restaurants even though they're supposed to have grease management systems, most of don't maintain them. So you've got all these facts, all hot fatsos and Greeces from washing pans and stuff going down the drain. Of course, when it goes down the drain, the temperature drops. And what happens to a certain animal fats? What happens to them when they go cold, they go solid. And then you've got all these flushable flushable moir's whips and other stuff and, you know, sanitary pull ups and the rest of it. And the fat binds them all together. But if you just Google Fatburger and look at some of the pictures, you get these like bus sized blockages in there. There's one there's like 18 feet long. You've got these massive blockages in sewers. And certainly over here, most of the sewage infrastructure, certainly in cities like London, is Victorian. And it's it's struggling massively. So you get these enormous blockages made up of discarded sanitary products and wipes all bound together with hard animal fat. And it's a hard water area as well, which which doesn't help. It makes it does things that make the the thing even different. And they literally send people down there with pickaxes to chip away at this stuff. It's horrible. It's actually somebody's job to go into a sewer and chip away at. There's a massive blockage made out of little fat and shit and wasteful. You look horrible. So what they're now doing over here is on the principle that the producer pays. Sean Derrig: [00:16:17] If they can prove that your restaurant is discharging irresponsibly fossils and grace is not a proper system in place, they'll start fine again. So as of now, we can go in and say, well, we can sell that for about penny mail. That's. That's kind of what it works out and it will get the authorities off your back and stop being fined and the rest of your drains, locking and sticking as well as in foul water, flooded a commercial kitchen. It aint pretty Harpreet Sahota: [00:16:47] What's up Artists. We all know that cloud computing has changed the way we live, do business and stay connected with everyone using the same cloud platforms. Winning and losing comes down to having the talent to build products better and faster. So whether you're an aspiring Data scientist looking to build your skills or a seasoned veteran looking to level up developing tech skills and being comfortable working in cloud environments has never been more important than it is right now. Cloud Academy has thousands of video courses, learning pads and practical hands on labs in real world cloud environments designed to help you build critical cloud skills. They cover everything from major certifications to dev ops to security to programing languages. Cloud Academy is the cloud training platform of choice for Fortune 500 companies and thousands of tech professionals around the world. Don't just take my word for it. Check out the reviews on GTA and get started now at Cloud Academy Dotcom, members of the artist's loft can lock in 50 percent of the monthly price for life. Just putting the coupon code artist when checking out. It's a great way to pursue certifications or just build your cloud expertize again, go to Cloud Academy dotcom and use the coupon code artist to lock in 50 percent of the monthly price. Harpreet Sahota: [00:18:27] So we're wasting waters in some ways that we may not even realize. So what's something that we could start doing today to reduce our unnecessary consumption or misuse of water? Sean Derrig: [00:18:40] Most toilet systems now have a dual flush thing where you can either have like a short flush or long flush, depending on what you've been up to. Those are great. If you're got one of the old sort of ballcock systems, got a couple of house breaks and chuck them in there now being flush water down, you'd be amazed how much water dripping tap. And actually the I'm with the figures I have especially you've got a water meter and you're paying by usage rather than just a starting charge. So things like dripping taps mounts up really quickly. You know, that can be a lot of water. So domestically, those are the kind of things you can do commercially and stuff like waterless urinal. So that's sensible water management strategy. I was doing a project years ago for a big architectural practice and they were converting old municipal buildings into like really nice flats. Sean Derrig: [00:19:31] And they have this sort of water harvesting system. And the problem was rainwater is full of biology. Sean Derrig: [00:19:38] So what was happening was they were like using this as like flush water are very environmentally friendly. There's all these great I shit growing all over the place. You can see the spots where the water goes into the into the toilet lifecycle is algae growing everywhere because they were flush with rainwater. Sean Derrig: [00:19:54] So I invented something that would take care of that. There's a really interesting one, but I like where it is. But that was a project to build a very large water feature in a very large municipal building. Sean Derrig: [00:20:07] Can't say what it was. And they have this idea of harvesting rainwater. And so somebody points out I'm about an inch of rain a year. So I just trying I tell you what we'll use TSA is treated sewage effluent. So the idea is you get like treated sewage water, like you are recovered from that and you pump it into a big glass tank, will open salt in somewhere open to the sunlight. Sean Derrig: [00:20:35] And this is like treated sewage. I flood in one of the hottest countries in the world and you're going to put 500000 liters of treated sewage effluent open. What could possibly go wrong. So they're going to cover the system for them to stop it smelling as I have been to Japan and I don't go apparently every golf course that stinks of shit for the simple reason that using treated sewage effluent to water the grass. Sean Derrig: [00:20:59] So biologists will obsessed with shit, basically, especially microbiologists. Harpreet Sahota: [00:21:05] I like the interesting play on words for your company name. Can you talk to us a bit about that? Sean Derrig: [00:21:12] Yeah, it's radical ICLE rather than ICAL and it's a bit of a pun. And to be honest, thinking up names for the product company, whatever, it's really difficult and you know, oh, came out with it and then we sort of brainstorm that from the roots of getting rid of. It not so it was just like. We thought Wiki Barile as you play on words, but we couldn't think of anything else because science is not art. So we have no imagination of the complete Philistines. Harpreet Sahota: [00:21:42] So you wrote a story that pretty much debunking claims that Oseltamivir known as Tamiflu was useless and that the studies were rigged and you managed to upset a lot of people. Sean Derrig: [00:21:55] This was great fun, very, very long running story where the British Medical Journal, it used to be a journal of record and it used to be a proper scientific journal. Within the last decade or so. It's basically become a political campaigning organization. And they have this idea the drug company who was selling Tamiflu haven't put all of their data on all of the clinical trials Data out there. Sean Derrig: [00:22:21] And there was this ever escalating war of words about you haven't released all the data. We want to look at all this stuff and it got more and more heated so that eventually more so fuck it, they go, that's the law. Sean Derrig: [00:22:32] You know, they did it. So the BMJ analysis on it. And sure enough, it found that I mean, this is the press release, not the paper. They were saying that Tamiflu is a complete bust. It's no more effective than paracetamol. And the government, the British government has wasted half a billion quid stockpiling it for a pandemic, a pandemic. Sean Derrig: [00:22:58] So I look at this. I am not a data scientist and I don't like Stats. They make my head hurt, but I to look at this. So what they've done is you think, OK, so now they've got all the data, you analyze all of the data, no they didn't. When you do an analysis, which is where if you put lots of small studies together, if they're suitably similar in that in the way they've, you guys will get this measure. And, you know, you've got a bunch of similar studies, but all the numbers are going to take a big number and then you should get more powerful stats out of it. Sean Derrig: [00:23:36] The problem is, if, you know, if all of the studies that doctor so you put them together, it's not going to magically turn into a pot of gold. What they did was it's all about the inclusion criteria. Which studies do we lumped together to do this analysis? So in terms of the data, that is right. We're not going to have any published studies on it. So all of the good data that have been published that didn't use any of that. It's only from this that this stuff, what else is actually excluded? Sean Derrig: [00:24:07] The people who are actually sick when they're in in the trial, they just admit, yeah, they it was people who are otherwise well or a lot of chronic as opposed to an acute illness. So if they were sick, they weren't included. If their numbers have been in a previous published study, they weren't included. Basically what they did was a more cynical person than I might surmise that they picked the studies where basically it would give them the result they wanted. And it's known that Tamiflu is effective. If somebody is very sick in hospital and gets flu, there's a 30 percent chance it's going to kill them with flu serious. It's a it's a it's a proper killer. Sean Derrig: [00:24:54] So they excluded all of the people that actually have been proven to benefit from Tamiflu. They excluded all of the other good studies. They excluded a bunch of other stuff. And it was almost as though they were looking to the result they wanted. And my argument in the article was, these people have got a dog in the fight. They spent five years slagging off the drug company saying, you know, you're a bunch of crooks. You're speaking with about Data back and only only putting out the positive stuff. We want to see all of it. So at the end of that process, I started to say, oh, actually, you know, you are a bunch of bastards. And those were you know, you can see that politically, that might have been a slightly difficult to swallow to pardon the pun. Sean Derrig: [00:25:35] So I called them on that and basically pulled it to pay six of them five years to come up with this. I pulled it apart about 500 words. They got really pissed off. Sean Derrig: [00:25:45] The authors responded to my blog post. So I. Right. If you're so clever, here's a link. You to do the opposite. I've got neither the time nor the inclination to analysis. Not it's not your analysis. I'm arguing with it shall cracky inclusion criteria which are rigged to this result? Clearly. So don't don't try bluster with science, boys. It ain't going to work. I then got a response from the editor of the British Medical Journal asking if I would like to put a rapid response in the BMJ. And I had a comment. I can't find it immediately. There was a comment from an academic over your outlook on, say, why on earth would the author wish to contribute to your review process so clearly flawed? This was. They have had quite a bit of fun with that one. Sean Derrig: [00:26:35] It's not that I enjoy upsetting people, but... Harpreet Sahota: [00:26:38] I think a clear example of just numerous biases committed at first. You've got the selection bias, right? Who looks like doing r Sean Derrig: [00:26:48] esearch? Degrees of freedom, I think is the plight of the lots. Yeah. So they. Harpreet Sahota: [00:26:56] A little bit of publication bias and a little bit of selection bias and then. Yeah, definitely some p hacking going on there. So are you familiar with just, randomized controlled trials. Sean Derrig: [00:27:06] OK. OK, Harpreet Sahota: [00:27:08] So talk to us about the importance of those when it comes to like doing biological sciences. Sean Derrig: [00:27:14] Ok, so let's look at the ideal world. The ideal world is, let's say your launch, a new drug or so there are these things called the placebo effect now multiple very, very, very, very powerful. And this is how bullshit like homeopathy and acupuncture works. You tensorflow to see the doctor when the symptoms are. There was the natural history of pretty much anything is it turns up, hangs around a bit, fucks off, and all diseases are by definition self limiting and they'll piss off all those two ends of the scale. Sean Derrig: [00:27:52] So what happens is you go to see the doctor. I take this, you'll feel better, especially when it's something like pain or infection stuff. You take the pill, you get better. What you don't know is what you have got better anyway, or is it the pill? And everyone credits it to the intervention. That's just the way psychology works. So with a clinical trial, I mean, you need to ethically, but you have to witness you've got the people that get the sugar pill, the placebo, and you get the people to get the proper intervention. Now, interestingly, they're trialing the Kovik vaccine at the moment. The placebo they're using, that is the meningococcus vaccine. So people do get a bit of a saw and they do feel a bit fluid or whatever. So the placebo they're using will give a very similar, if you like, a clinical effect in the subject. But it's a different vaccine, which I thought was a really clever twist. Sean Derrig: [00:28:51] But the point is it's blinded, but it's not the people who have it. They don't know whether they've had the sugar pill or whether they've have the real drug, but it's double blinded. So the person administering it doesn't know either, because obviously, if they know that they're giving you the sugar pill or the fake injection or whatever rather than the real one, they might unconsciously give it away. So if the person who's administering it doesn't know whether it's the real one and the person getting it doesn't know it's the one, that's how you can usually quite accurately work out whether it's better than placebo, because then when you crunch the numbers, you've got the placebo group. What was the effect there? And you've got the real wing of the study, which is okay, let's say, you know, is that significant difference. So the placebo effect. So let's have a double blind trial looks now there are issues. That's the ideal world. But what I was doing, infection control stuff you got all these studies are where let's say you've got this magic machine and I've seen loads of them, either fogs of room all year or whatever, to get rid of hospital infection outbreaks of things. I see that or whatever. So what happens is the natural history of any outbreak is it turns up, it hangs around for a bit. It goes away. Sean Derrig: [00:30:13] That's what happens with outbreaks. And the problem is you've got lots of things going on. It's not like you can say, right. We're going to treat everyone on this one, but not on that what you can't do over. So what happens is people first of all, you get people to behave differently in an outbreak. They start doing things that they should have been doing but weren't like cleaning things. They stop doing things they were doing more assiduously, like hand washing. So they change their behaviors. And then you bring in another intervention, whether it's side by side with hydrogen peroxide or whatever. You bring in that if the outbreak then eases, it's the intervention. Sean Derrig: [00:30:57] So people don't report on, oh, yeah, we did this and it didn't work. Mean they should do it. They don't. Or well, we did this here work. We did it down the road. Useless. So the only the positive stuff gets published publication bias. But that's a known boss. We know about that. So we can kind of control for a bit. But the problem is it's something complex with a lot of moving parts like an outbreak in a hospital. Sean Derrig: [00:31:21] People tend to credit the intervention a bit like when the doctor gives you the pill and you were going to get better anyway. They credit the intervention with actually having an effect. When it raided, there's something called evidence based practice as well, which confounds this a bit where gone back a few years now when they were trying to get rid of quackery in medicine because they worked out the revolution, doctors strikes across the planet where the mortality rates came down because there was so much dogma and stuff in there. Is that right? What does the other side? The problem is you would often have things where the Data said something. That was one where it showed the homoeopathy cured mumps, which is just completely implausible, total ballocks. But because the numbers just happened to come out at random. The randomness in that would interpret it that homeopathy is an effective intervention. It can't be its water way cures. I'll give you that because I'm looking at try plausibility and just what the evidence what the data is telling them. You got some of these bullshit results. That was a really good paper that Smith in 1980. It was. And I'll look up the name of the standard so you can chuck it in the notes, but that they were getting pissed off with all this is evidence based stuff and they're working in jobs and going where it is kind of a bit more real estate. And there's brilliant paper out basically asking all parachute's an effective intervention when dealing with gravitational trauma. So I said, OK, so what we do know is we know that parachute's can cause injury. But what we do know as well is, you know, people say that parachutes protect people when they know who might be subject to gravitational trauma by jumping off point. But nobody's actually proved that the parachute is an effective intervention. And there are numerous there are numerous injuries from parachutes every year. Sean Derrig: [00:33:23] So until we have a randomized controlled trial, we can't actually say the parachutes are an effective intervention when dealing with that are protective against gravitational trauma. And they wrote to all the other proper scientific language basically sticking to things mergence. Sean Derrig: [00:33:43] And what they suggested was what we need to do is get all these people who are employed to shut out all this evidence based bullshit. We put them into two wings. We give half of them a parachute, half of the rucksack for the fuck out of a plane. And medicine will be a much better place for one simple trial, a great paper. Harpreet Sahota: [00:34:08] What's up artists I would love to hear from you. Feel free to send me an email to theartistsofDataScience@Gmail.com. Let me know what you love about the show. Let me know what you don't love about this show and let me know what you would like to see in the future. I absolutely would love to hear from you. I've also got open office hours that I will be hosting and you can register like going to Bitly.com/adsoh. I look forward to hearing from you all and look forward to seeing you in the office hours. Let's get back to the episode. Harpreet Sahota: [00:34:56] Ok, we're out there and we're doing research for whatever is decided to do. A lot of research as well, whether it's to gain some industry experience or just some domain expertize. What should we be asking ourselves when we're reading some of these research papers to ensure that it's not bullshit that we're reading? Right. Sean Derrig: [00:35:18] Ok, this is this is an interesting point when I'm thinking anti-vaxxers and sundry other loonies on the Internet. The problem is you can always say, you know, it's like if you interrogate the Data hard enough and for long enough, you can make it admit to anything and you can always find some bullshit study that will back up your bias if you fix enough. The problem is a lie to many people on this. They are really quite intelligent enough to read a scientific paper on a lot of subjects, especially medical stuff and that kind of thing, whether you are on string theory or something, really. OK, they can read a paper which is saying something, which I know is horseshit. But they can't critically evaluate. There's a world of difference between reading a scientific paper and understanding it and reading it and being able to critically evaluate it, which is like a box of Tamiflu. I mean, you guys to probably heard a minute look at the criteria that you used. I know. And the red flag comes up. So critical thinking is really, really, really important. Sean Derrig: [00:36:24] Another thing, the critical thinking skills and the cognitive biases that we fool ourselves with all the time, every day don't include a confirmation bias, all of that kind of stuff. It's really important that we're not just scientists, but anybody. There's this echo chamber bullshit out there called the Internet and, you know, in Microsoft. All right. Now, what I try to do with them is they ask a question. I don't give an answer. Sean Derrig: [00:36:52] How can we know that I'm trying to do the whole time is equip them with the critical thinking skills to be able to separate what's real from the ballocks if they were learning about the oldest, but when it was a primary school, learning about Noah's Ark from the religious studies teacher. So it was hilarious. I had more with the religious studies teacher on a screaming atheist. It was okay. So why didn't know all the non kosher animals on the ark that I asked Mrs. Guy? Why did the dinosaurs, the sheep that I asked Mrs. Colby. The one thing I didn't say, it's OK. So the whole world was populated by one incestuous family twice over. Didn't do that one, obviously, because I was thinking, OK, so how old is the flock? So I don't know. How long did it last? A year until the wolves were down. And when was this? A thousand years ago. So what happens to a tree, as you put it, under water for a year? Die? Absolutely. How come there are ten thousand year old trees in California to ask Mrs. Coming? So, you know, when you got a teacher, you can't handle questions like that. I've always been the wrong job. Love. That was the simple. So what I'm trying to do is teach them those skills. So hang on a minute. Let's just look at that. Sean Derrig: [00:38:13] How can we actually know that? And I thought there were scientists. It was that said, you know, don't fool yourself, especially if it's a research project. You're basically wedded to your you know, you're trying to prove rather than disprove your favorite theory. Don't fool yourself and you are the easiest person to fool. Harpreet Sahota: [00:38:34] Yeah, I think that was Richard Feynman that said that everything. So one thing we could do to make sure were flexing that critical thinking skill when we were going over and reading through research papers is to maybe come into the reading with a set of questions and anything that we come across actually just sit and think about it and try to find some contrary evidence to kind of either corroborated or it's it's very easy to go down a rabbit hole of confirmation bias where you just pull out stuff that confirms your existing bias and Facebook sort of handheld confirmation bias machine. Sean Derrig: [00:39:14] So, yeah, it is important to seek out the disconfirming evidence, but also it's about consistency. Often what'll happen is especially with a fairly new subject or a new way of looking at something or whatever is the other studies will be shit. Sean Derrig: [00:39:29] And then as it gets replicated and gets drilled down more and more, the latest studies have shown how far more and you get more of the consensus. You can always find an outlier to confirm your bias. But you know, the trick is not to go down, not biased rabbit hole, but it can be quite difficult, especially when you kind of want to believe something and you're looking to back it up rather than blow it out of the water. Harpreet Sahota: [00:39:53] This kind of leads me to the next thing I want to talk about is how you been freaking people out with bayes theorem, especially with respect to covid testing? Sean Derrig: [00:40:02] Yeah, I've been, because an awful lot of the kind of businesses we do is have been doing very much for months. I've been dragged back into infection control. And what I'm doing not a lot is advising boards on what they can and can't do because there's so much misinformation out there about it. Sean Derrig: [00:40:26] I mean, I've been advising the film and TV community about getting women on the couch, you on Netflix or whatever. So everyone's been paid and you've got all these content, none of it being bloody made. So they've just reopened Hollywood for the young, the while they're not starting. I think that's a break. Right. Production of the likes of Jurassic World Movie. There's all sorts of TV production that's now finally being started. I'll be doing a lot of that. And it's really it's there to let me help you navigate through this. There's all this guidance, some good. So I find that what people have been doing is they've been concentrating on poxy, inconsequential little risks and ignoring the great big bucks. That's because they don't they can't evaluate what the big risks are and what the small risks are. So I'm I've been helping a bunch of businesses, which is kind of interesting guy. Sean Derrig: [00:41:19] I get to go into lots of different businesses and talk to the you know, the senior team is, like I said, what makes you think about that? That way? I've I've done a fair bit on this now. I blogged a bit about it and there was talk of immunity, passports, and wouldn't that be a fantastic thing? I got a brief immunology lesson. When you got infected with something, your body's got a bunch of nonspecific responses, fever, all sorts of stuff that it will do. And it's purely based on like not like this is you know, this is a self. So that is a non self serving deal with it. And over time, it learns that the immune system learns to recognize a particular problem organism. This is how a vaccination works. Or if you get something in your childhood, your body loves to recognize it. By and large, you don't get it again. Sean Derrig: [00:42:15] It's how vaccination works. Vaccination does not cause autism. It causes adults. So the idea is you get exposed to something. The body learns how to deal with it so that there are antibodies that turn up during the infection and then sort of wieters the infection once. And there are other antibodies that are a little bit later, but they're the ones that stay in your bloodstream long term. And if you're really close to whatever infectious agent it was, they're the ones that the but the big red flags up and restart the immune system. So when you get your flu vaccine every winter and they say. Right. It's not going to be protective for a couple of weeks. So in the next couple of weeks, you're still at risk of catching flu. Flu probably isn't the best example because because of the way flu is, the flu vaccine isn't 100 percent protective. It's one percent crap, just not 100 percent protective. But if you have a vaccine against them, it's 100 percent protective. I will say to you it's going to be a couple of weeks before you're properly immune. But the thing is, when you've got that immunity and then you're exposed to whatever the infectious agent is a day or two, and the body can raise that specific response against that specific organism rather than just the general not, you know, not self must die kind of response. So this antibody could if you can measure that in somebody's blood, that will tell you that they've been exposed today. Wouldn't that be great if we had this idea that you do a simple blood, you know, like a pin prick blood test? It would look a bit like a pregnancy testing plastic thing, except you put a drop of blood on there and then wash it through with the reaction that comes with it. And that will tell you in 10 or 15 minutes whether you've been exposed to it and therefore one would hope have immunity to it. Wouldn't that be great? Really, really, really attractive idea. Sean Derrig: [00:44:09] The problem is the tests aren't perfect. They aren't 100 percent predictive. So let's say you've got a test. If you were offered a test that was 95 percent accurate. Yeah, that sounds. Yeah, I play, wouldn't you? OK, so now you know the answer is because you're not a data scientist, but if you've got a test that's 95 percent accurate for the accuracy of 95 percent accurate and you're testing for something, let's say one percent of the population of of how to got antibodies to it, all of your positive tests, not 95 percent accurate. What percentage of those positive test? Results are going to be false positives go. Harpreet Sahota: [00:44:56] I got a pen and paper for this one. How about you just tell me. Sean Derrig: [00:45:00] Eighty four percent of your positive test results will be false. People just cannot get there. How can a test that's 95 percent accurate give you a wrong result 84 percent of the time? And it's that something called Bayes Theorem, which is bloody interesting, but rather inconvenient so that people fall into the lottery fallacy here, which is people ask the question, what are the chances of me winning the lottery? Actually, the question is what are the chances of somebody winning the lottery? Yeah. So let's make this easy. You've got to test is 95 percent accurate and one percent of the population have got whatever you're testing for. The same thing applies to, for example, workplace drug testing. This is just math. This isn't biology. So if you test a million people and one percent of them have got what you're looking for, 10000 should test positive. Nine hundred ninety thousand should test negative. Yeah, OK. But the test is 95 percent accurate. So of those 10000 people that should test positive, nine and a half thousand will and five hundred will be false negative. The test won't pick it up in those 500 people out of the 10000, that should test positive because it tested. So you got five hundred false negatives, nine and a half thousand true positives out of that. Ten thousand. Let's look at the nine hundred and ninety thousand. So of those nine hundred ninety thousand, a ninety five percent of those will test negative. So that was not a forty thousand or whatever it is will test negative. Those are true negatives. About five percent of those nine hundred ninety thousand will be false positives. The test will come back positive because it's 95 percent accurate. So five percent of nine hundred ninety thousand is what, forty nine thousand or whatever it is. So of all those positive test results, nine and a half thousand are true positives. 49 and a half thousand are false positives. Sean Derrig: [00:47:26] So of all of your positives, only 16 percent of them are true positives. So if you tell 100 people that they've got the antibody, therefore they believe they are immune and only 16 percent of them have actually got immunity, they could go and catch it, spread it, whatever. And that's why false positives in the case of antibody tests to say that you've had something. It doesn't matter what you do. If you've been told you've got the I've got friends you've done even read about ahead with this is not you know, I've I've had a lot a me. So that's just the test. What we don't know is even if you do have an accurate test and you do have antibodies, we don't know how protective those antibodies are. We don't know what the antibody titer is that you need to guarantee that it's protective. We don't know how long immunity will last and the organism mutates. It could be that the antibodies become less protected. So there's all of that stuff that we don't know. So even if you have got antibodies and it's a pocket test, we don't know how protective those antibodies are and what that actually means. I will try to figure out over the next year or so. We just don't know that yet. The problem is, if you do a test and there are tests of more than 95 percent, not much more if you're going to test is ninety five percent accurate. Sean Derrig: [00:48:51] Eighty one percent of the population. That's the key thing is it's it's what's the actual incidence at large, even if it's five percent, doesn't get much better. So that's the issue with those tests. Sean Derrig: [00:49:02] You really can't you can kind of test to rely on the results at a population level, if you will do a bunch of testing, for example, the Data, to work out what percentage of the population might have an exposure. You can take a bit more comfort. But the challenge is if somebody thinks they are immune to it, their behaviors will change and those values of change will be the ones that can spread of that kind of. Yeah. So like in a case like this, the math is actually quite simple, but it's really counterintuitive. Harpreet Sahota: [00:49:34] Yeah. It's just it's a matter of just taking that kind of a decision tree approach and drawing the branches and then doing the multiplication back like that. But in a case like covid testing, like what's the error that would be most detrimental? Would it be the false positive or the false negative? And why. Sean Derrig: [00:49:51] Ok, so the. Two things that is testing whether you've got it like the nasal swab they use and they do the PCR on, it's not the type PDF antibodies in a minute with the PCR test, it's quite a finicky test to do. The person's technique used to be quite good. The there's two things you need to know about biosecurity. And you've got a viral load, which is how much virus is somebody producing, coughing out, whatever. And you've got infectious dose, which is how many virus particles I take against it in the low thousands like high hundreds of thousands. In terms of the amount of virus particles you need to ingest in order for the infection to take off with norovirus is really very low. A fewer than 10 seen numbers of one norovirus possible can can screw up. Sean Derrig: [00:50:47] With different bacteria, that's different. You might need tens of thousands of campylobacter, you might need a couple of dozen E. coli if one of them really nasty. So there's infectious disease. So with the swabbed test that tells you whether you've actually got it or not. I'm talking averages here takes just over five days from exposure to symptom onset for two to three days before symptoms that you're infectious or shedding virus. The problem is if you take a swab, just as people are starting to shed virus, you know, very infection and can infect others, you've got like a 20 percent chance of actually returning a positive result. So if you're testing for the illness, if you get a positive, you can be pretty bloody sure the person's got it with a negative. If you really can't take too much comfort from that, you've got to be quite careful. So false positives means people are going to get there. A few false positives with that, with the genetic test, with a nasal swab test they send off to a lab. False negatives means that people may well still be spreading the infection, which isn't a good look. If we're going to antibody testing, if you tell somebody their immune system and then not they might that it will spread to other people if people are immune to it and you tell them that they aren't. Well, chances are they're going to keep doing the behaviors that are going to be protected, the community at large. Sean Derrig: [00:52:16] So they're both bad. But the question is more what what is it you looking for? Or you're looking for an active infection? Are you looking for long term antibodies that show you that some degree of protection against infection or infection? Sean Derrig: [00:52:31] So if biology is complicated, sorry, what about sensitivity and specificity with regards to KOVA testing? What does that mean in this context? Sean Derrig: [00:52:42] So when I was very lazily talking about accuracy of the test, sensitivity on on a test is how likely is it that it will identify a positive? So with this, the sensitivity is, if you got it, how likely is the test to pick it up accurately? And then there's specificity, which is if you haven't got it, how likely is it the test will actually confirm that you haven't got it? So sensitive sensitivity. Will they test? How likely is something to test positive if they are positive specificity, how likely are they to test negative if they are negative? So you just need to split this accuracy's into these two separate wedges, that's how the stocks and work on it. Harpreet Sahota: [00:53:35] And what does bayes theorem have to do with all of this and covid testing? Sean Derrig: [00:53:41] Well, bayes theorem is the thing where you work out that if you've got a million people and you test them and it's one percent that's based, they're not. The case of math has been around since the 70s something or other. Sean Derrig: [00:53:55] So that's a very old piece of French math. So, yeah, that's that's that's based there. But it's used for pretty much all diagnostic testing in the. So what's the true incidence in the community? What's the sensitivity? What's the specificity? And then just plug it in. It'll tell you, for example, for gentlemanlike instead of the PSA test, prostate specific antigen, which was an indication of what you might have, prostate cancer or not. But because of this very problem, it was causing more, you know, more issues than it was solving because it was incorrectly identifying people at risk of all this kind of stuff. Sean Derrig: [00:54:31] So, unfortunately, this is a doctor's big, thick fingers of your chocolate starfish if you need your prostate exam doing it. So that's not a nice little pinprick test because that doesn't really tell you anything very much. So why is it that as a species we are so damn rubbish with probabilities and statistics and trying to conceptualize even though all evolution, we're really good at recognizing patterns, we greitzer patterns, we're going to tell stories with different numbers. It's just the way our brains work. But it's genetically hard wired. And in that I mean, you say power Data, which is where people can say, you know, Jesus on a tortilla or a face in a cloud or whatever part of you. That's because we're really good at recognizing patterns. Sean Derrig: [00:55:16] And the reason for this is if you could recognize the pattern of the tiger's face in a bush a split second before you might, you were the one that didn't get eaten. And as such, this this pattern recognition software, if you like, well, a mixture of hardware and software in the brain, it's genetically hardwired because back in the day, if you could recognize a harmful pattern like that more quickly than somebody else might get eaten and not you. Sean Derrig: [00:55:46] It's just the way we're wired up now. We think we're so clever with our opposable thumbs, our iPods or crap that actually which is very clever monkeys that, like humans, were just really just moist programable robots in a sense. Harpreet Sahota: [00:56:00] Right. So you've been involved in some. Well, yeah. You've been involved in some pretty cool, you know, entrepreneurial initiatives with the one that you're currently a part of. Like, do you have any advice or tips for anyone who's been done? Sean Derrig: [00:56:16] I think it's very important to surround yourself with the right people as a great book written by a friend of mine called The Beermat Entrepreneur. And he talks about having Cornerstone's energy, finance, cornerstone and just sales cornerstone, energy delivery cornerstone. And then you're the nutter in the coal is coming up with it. Eventually stuff something on every business I have on. I'm going to learn around the place kind of mad ideas. I need somebody walking by. I'm just going to move down and actually making out with the ownership with the goal. Not a good answer. So you do need to be honest with yourself about what your strengths and what your weaknesses are. And I mean, I'm not a salesman that put me in front of a reasonably senior person that a decision maker and I can buy things. You're going around and prospecting that and say, no, no, no. There's lots of people who I've been told to piss off like old salesmen. And it is always men. You know, the women are really good at it. So you need to be honest about what your weaknesses are and surround yourself with people who compensate for that. So my book, it's finance columnist Thomas Styles, Cornerstone, the delivery cornerstone and usually the latter in the quarter coming up with stuff. Also, you need to have people around you who will challenge you, not just say what they think you want to hear. Sean Derrig: [00:57:42] You know, I always work on them whenever I get parachuted in to take over a team is it's really, really simple, but I never like to do it once. And you go straight away. I don't care how ugly the truth is, I really don't. We need to establish what the facts are. Deal with it and then move on and preferably learn from it, and I don't care how many mistakes you make this month where you can get a shit, don't make the same ones next month. You know this you know, before I've let people do things which I knew were going to blow up in their face, but they were invested. And I really want to do okay. Sure. Okay, fine. Well, let me give you a little about this year. And I've let people make mistakes deliberately, and I knew that was going to happen. Yeah, but the way I love you, I've just invested in your training. Sean Derrig: [00:58:32] So you need to have that kind of mindset and you need to have the mindset where you need to understand that this took me a while. You need to understand that your success is entirely wrapped up in how successful you can make your team. It's not about you. It's about the people you put around you and what you can do to make them truly successful. And that's what you got your success from, because the company isn't going to grow based on finding your own ego. It just it doesn't - it would be great if it did. It would be marvelous. But life life doesn't work like that. Sean Derrig: [00:59:10] And you need to have the kind of culture life at which I was born and then challenge you on that stuff. And they know they can do it. I mean, they probably know not to do it in a shelter. Sean Derrig: [00:59:22] It's something they might need to work out that people do have egos, you do have feelings, and they might just do it gently. But you need to have that kind of environment where know they can wander in and just unload on you or whatever. So, yeah, that's kind of it. You need that culture where people don't sweep things under the carpet. Sean Derrig: [00:59:42] You know, if you make a mistake, you can tell me and then we can deal with it. So I so yeah, that was part of the culture that I try to engender. But just be honest with yourself about what you're talking about somebody else. Harpreet Sahota: [00:59:55] So I guess what would you say are some key traits that if somebody wants to be an entrepreneur, a full fledged entrepreneur, that they should be cultivating within themselves. So one of them you mentioned is just finding the right people, having the mindset that success is not about you. It's about making people around you, successful entrepreneurs, because we're all egomaniacs. Sean Derrig: [01:00:17] So you need to be quite honest with yourself and, you know, read up on it. And just just I mean, it's a difficult one. And without the ego and the drive of the lunatic at the helm, things don't happen. However, you do need somebody with their eye on the brakes and an eye on the cliff that you're hurtling towards, as well as. Always. I introduce people that he's the grown up in the room. So, yeah, you do need some some some grown ups around the place, unfortunately. Harpreet Sahota: [01:00:50] What's up, artists? I would love to hear from you. Feel free to send me an email to the artists of Data Science at Gmail dot com. Let me know what you love about the show. Let me know what you don't love about this show and let me know what you would like to see in the future. I absolutely would love to hear from you. I've also got open office hours that I will be hosting and you can register like going to bitly.com/adsoh. I look forward to hearing from you all and look forward to seeing you in the office hours. Harpreet Sahota: [01:01:31] Let's get back to the episode. Harpreet Sahota: [01:01:39] So in this COVID era, in terms of our entrepreneurship, what do you see being some interesting problems that are worth tackling that might be Data related or not Data related? Harpreet Sahota: [01:01:50] That would be kind of interesting inside. Sean Derrig: [01:01:53] The blog, which is demanding mystery. So I don't feel often enough. It's about countering the bullshit with Clover. That's the key thing. There is so much misinformation out there. Bloody Facebook and it's this death. The people have no respect for expertize anymore in that they think that their opinion they've pulled out of their ass is is equally valid to somebody who's been working in the field for four years. Sean Derrig: [01:02:21] As an unrelated example, I shout at the BBC on the radio, and there's a big issue with false balance in science reporting lots about on things like climate change now. Well, OK, so we've got all of these NLP professionals who've been studying climate science for all of their careers and their consensus is this. Yeah, but we need to find some random nutter who's going to say the opposite. It is even qualified in the field. I'm going to walk out without talking about up in the north east of England area, all the kids Data falling out. And there's this very eminent academic you've been in clinical practice. Sean Derrig: [01:03:00] 40 years, he said, look, we need to put fluoride in the water. This is one of the simplest, safest interventions that will stop this not overnight, but quite quickly. And I've taken too many kids too far in my career. I know that people are concerned about chemicals in the water. This has been proven throughout the world to be the safest one of the most effective interventions we can possibly do. And it's really badly. No. Thank you, Professor, with your 40 year clinical experience. So what we've done is we've been speaking to Sharon in Redcar who's been who's come up with completely the opposite view by sitting on her sofa watching daytime soaps and shooting crotch fruit. I thought, you know, worst chemicals in it. It's not just me screaming at the radio every time I to. Sean Derrig: [01:03:56] Back to the data science thing, I'd say just keep canceling the bullshit out there. Let's get the good studies out there and make sure that people understand what they mean and what they don't. And in terms of science communication, what you need to suss out really quickly is the level you need to pitch something at. I mean, I'm equally comfortable speaking to the kitchen porter in a hotel if there's a hygiene outbreak or whatever or chief executive and what you need to do, it could be, you never know in the foreign kitchen for what you're talking to, might have a degree in astrophysics, in the wrong language, but the sea could be thick as shit, but just be quite good at bullshitting as well. So you need to read people quite well and work out where to put your Data and what you know, just just put it in language they both understand. But you just just keep from the bullshit. They it's a bunch of Data and that's a crunch. Harpreet Sahota: [01:04:52] So being an executive yourself, do you have any tips for Data scientist for the appropriate way to communicate their ideas when they're presenting to members of the C Suite or the board of their scientists tend to just get way into the weeds of the numbers in the modeling. Do you have any tips on how we can be effective with our communication? Sean Derrig: [01:05:11] Yeah, you need to. It's very helpful when you're presented with options and that can be as simple as, OK, either we can do this, which we think is quite a good idea, or we can hold our hands in a bucket of shit and give ourselves a huge round of applause. Which would you prefer to do? Well, I think I'll be the first one. The other problem I find is and you get this with computer geeks all the bloody time, which is fantastic. New platform. Great. How is it going to help us sell more shit and make more money? Well, that's not why it's a fantastic no. How's it going to help us sell more shit, spend less money and make more, which isn't going to help us sell more shit? Is it going to mean we spend less on other shit? Is it going to make us more money? Is it going to make me happy? Yeah, we know. But, you know, just putting in big print for the hard of hearing, we reckon this could do this. Sean Derrig: [01:06:09] Really how is it going to do that? And then you start drilling down to the level of detail level of understanding that the audience wants to start, you know, start with the money shot and go work back from that. And it needs to be relevant to the business in that if it's completely irrelevant, chances are not only you're going to get ignored, you might be looking for what the fuck is this you got, you know, that kind of question. Sean Derrig: [01:06:36] So keep it keep it relevant, keep it succinct, but then be ready when you get their attention, which if you get your own life in one of the greatest advertising copywriters ever once said when you're writing it out, that's only 10 percent of people read beyond the first line. So you've blown 90 percent of your budget on the first sentence. Make it a fucking good one. And in that, I think sums it up, it's you need to make sure you grab their attention, right. Here's here's what we've been doing. Here's what we think. This this this really. And then you need to know your subject, which, of course, you should do, but you then need to have a lot of layers of information. Here's your top tier. Here's your second tier. And any of the bullet points you put that just be ready to go down that rabbit hole because they are the final stretch. Sean Derrig: [01:07:34] We'll have a different set of questions to the ops director, to the CIA or whatever. So you need to be aware that the different people around the table are there for different reasons and none of them are idiots. But they'll have a different set of questions because I'm looking at a different way, which is a good one perspective. You need to understand the angle that they're coming from. Otherwise you not answer their questions wrongly, but just partly because often there's a particular nugget that that they're after. Unless you don't understand kind of what their role is and what they bring, you have a lot more difficulty identifying quickly why they're asking that question. So it's just a bit of everything else. Harpreet Sahota: [01:08:15] Thank you very much. That's some really, really valuable advice I think is really going to benefit from that. So you get a lot of patents credited to you. So what would you say is your most favorite one or the one that you're most proud of? Sean Derrig: [01:08:31] The one I do like. I've got one for frac fluid. So depending on how fracking was or what it is they're going to get across to. Sean Derrig: [01:08:39] Well, and they pump a loaded down there under pressure. The idea as it recovers and the last knockings of oil and gas down that so do is you get like a sort of 80 ton tank and you fill it full of water and then you put some usually Qualcomm in there to fix it up a bit and then you put a bit of it. But it's the fact that you have to spin that to make it work. And of course, this is a bug food. So you get bacteria growing on. All this stuff is a huge form, these biofilms and frac fluid. So then you can get past that. You pump it out of the pressure down into a crack on oil fields and then recover it under pressure and you end up with 80000 liters of oily water and some gas. Sean Derrig: [01:09:27] So you recover the best you can sell. And then you're left with like an eighty thousand liters of hazardous waste, which needs disposing of hazardous waste because it's hazardous waste crews in the night. Sean Derrig: [01:09:40] So the bug system where you could put a bag of bugs, it looks like looks like a bit of trash. Sean Derrig: [01:09:47] The watching, obviously how you smores you, Chuck, is a big thing. The bacteria shit the surfactants, they hang around chemicals to it and they get rid of all the bugs that have been causing the biofilm. So you're using less chemical. You haven't got biofilm issues. It goes down the pipe like a dolphin. Then you pump it back out again. There's more bugs and the bugs then eat all of the shit in the water that makes it hazardous waste. You can't put it into a river. It'll be a bit funny. Sean Derrig: [01:10:23] What you can do is you can use it to irrigate fields so you can take 80000 liters of frack fluid and take a lot of the chemicals out of the box at the other end. When it comes out, you can then remediate. It was a wonderful friend of bacteria and friend and then you can turn it into wastewater that's safe enough to make crops corkscrewing cool. Sean Derrig: [01:10:47] So that's other similar things for the treated sewage effluent, for these sort of competition and in the Middle East, because it was such a huge lake of oil with a tiny bit of sand over it. Let's see if we can build sustainable to build cities and stuff. But it won't for them where you could get treated sewage effluent and make it nice enough to use in fountains and on fruit and veg and that kind of thing. But yeah, the pattern for that was on the on the frac fluid, which is kind of fun because I got to play in oil fields and stuff and see all that. Texas, Texas, which is tremendously entertaining. Harpreet Sahota: [01:11:27] The last kind of formal question before we jump into quick lightning round. And that is what's the one thing you want people to learn from your story on. Sean Derrig: [01:11:35] I'm basically unemployable, and I haven't done too badly. Sean Derrig: [01:11:38] So if you're completely unemployable like me, franchising are a good home. "Everybody's unemployable in franchising" which which is kind of my sort of semi other career. But yeah, I mean, just yeah, it's good. I'm completely unemployable and I haven't done too badly, so don't fret about it too much. I would say don't don't overthink things too much. Harpreet Sahota: [01:12:02] So jumping into lightning round here if. Harpreet Sahota: [01:12:04] You could meet any historical figure, who would it be? Sean Derrig: [01:12:07] I've always found that meeting of scientists whose discoveries that don't realize until after their death is not I mean, mentally either prospering, not Ignacio, by the guy that is a doctor that ran a couple of maternity hospitals. And all these kids dying of personal favorite was known as. I wonder if we wash our hands between us. I wonder and sure enough, if I scientifically call the midwives in one hospital to wash their hands and just left as if they were, why it went away. And it wasn't until 12 years later just a germ theory. And I was like, oh, you're calling us dirty bastards, telling us to wash our hands because he has got because he was absolutely right and I never knew. I think I would kind of like to make it so I can have any idea how I call natural selection is a look at the stuff that Daniel Bennett's doing with, like cognition and that kind of stuff and how it's percolated across all of science and not just biology. That would be me talking to him. Harpreet Sahota: [01:13:17] So if he could put up a billboard anywhere, what would it say? Sean Derrig: [01:13:22] I would like to get a billboard on AIs put on it either thought what would Satan do? Or have a big LGBT Jesus and put it anywhere in Alabama. That's what I'd like to do with a billboard. Yeah. LGBT Jesus. Right. But fuck Alabama or something fantastic. I very much appeal to my sense of justice in general. Do you know the Alabama of the North? Obviously. Harpreet Sahota: [01:13:50] Anyway, let's what something you believe that other people think is crazy. Pretty much anything going through my head at any given time, really, you wouldn't want to spend time. So what's the most bizarre aspect or quality of human nature? Sean Derrig: [01:14:12] Let's say it's our ability for cognitive dissonance, the idea that we can completely believe an idea and completely believe a completely conflicting idea. But at the same time, it's sort of Orwell doublethink saying, don't get me started because, you know, just pick any religion you want. I can give you a bunch of them for any of those. Harpreet Sahota: [01:14:35] So what's the academic topic or area of research that you think Data scientist should spend some time studying or researching up on. Sean Derrig: [01:14:45] Critical thinking Sean Derrig: [01:14:46] That's a very that's an all encompassing term. If you don't listen to the skeptics guide to the universe, start listening to it. If you don't know the book, get it. It just packages everything that you need to know about critical thinking, analyzing data, making sure that the subconscious biases that we all have creeping in. If you do nothing else, subscribe to that podcast as well as this one, because it's absolutely brilliant and the book just encapsulates all of it. Coffee would be a copy, which is very nice. Harpreet Sahota: [01:15:21] The next question is going to be the number one book you'd recommend our audience read. So would you say it would be that one? The. Sean Derrig: [01:15:27] Yeah, that was kind of necessary. There is a book I was a cold on, I can say multitudes, which is a broad sweep through the fascinating, fascinating world of microbiology. It's I mean, you know, phonon sciences as well. It's absolutely brilliant. Sean Derrig: [01:15:48] It will tell you things that will freak you out. I mean, I can pull microbiological facts out, but this is there are bacteria that live in rocks that shit your goals. Let's just it's that there are there are bacteria that can change your sex, is all I can say. Multitudes and young. My the favorite is about the 1918 flu pandemic. Gruesomely fascinating. So on a Sunday. Harpreet Sahota: [01:16:18] So that was I contain multiples by Ed Gold. Sean Derrig: [01:16:22] I contain multitudes by Ed Yong Harpreet Sahota: [01:16:24] OK, awesome. I'll definitely include that into the show notes. So if we could somehow get a magic telephone that allowed us to contact 20 year old Shawn, what would you tell him. Sean Derrig: [01:16:35] Stop being a dick. Sean Derrig: [01:16:39] No, I don't particularly like my 20 year old self, I think it was, and people don't change now. You need to continually reinvent and improve yourself because there's always a superior model. Sean Derrig: [01:16:50] It's a bit like women, women for men, a bit like houses. It's it's not how it is now. It's how it's going to look after extensive remodeling. I've been extensively remodeled by my wife over the years, obviously very much for the better. Harpreet Sahota: [01:17:08] What's the best advice you've ever received? Sean Derrig: [01:17:10] I think in terms of management, when you start managing people, when I first started on that tangent, my then boss and great still great mentor said to me, here's what I want you to go away and make a list of the 10 things or at least 10 things of things that your previous bosses have done that really pissed you off, whether it's giving somebody public in your father to change their mind or you just anything that former bosses of yours have done that really, really got to write them down. So I get what you want. Yes. Write them down. Will come up when you're done. You right write the also have good management is avoiding bad management. Take that list. Never do any of those things. And that's a really good start. And I've imparted that same advice to everybody that I've ever sat on a sort of management career. As you know, it's not a secret science. Sean Derrig: [01:18:05] It's just trying to do things and it kind of turns out all right. Harpreet Sahota: [01:18:11] So what motivates you? Sean Derrig: [01:18:13] Greed! Sheer, naked greed! There's just so much stuff out there that I don't know. I'm totally disappearing in a bubble of my own confirmation bias. And I don't listen to the radio.I listen to the news a bit, but sometimes podcasts, especially about disease and pestilence and stuff and audio books, you know, no fiction. I'm a complete philistine. So there's just so much stuff, you know, that you can tell somebody there's an invisible man in the sky who is all powerful and all knowing. And he's got this list of ten things. And if you do that, you're got completely fucked you over. But he loves you. But even though he's all powerful, he's shit with money. And we know this because you need more reviews. So what's the money that people will believe they will give you money on the person? Well, I see a sign that says, well, I've got to touch it. So that's just kind of how my brain works. Harpreet Sahota: [01:19:10] What song do you have on repeat right now? Sean Derrig: [01:19:13] Well, it's of Frank Zappa greatest song ever written. Harpreet Sahota: [01:19:17] So how could people connect with you? Where could they find you online. Sean Derrig: [01:19:22] With caution? Sean Derrig: [01:19:25] You can find me on LinkedIn. I'm sure you're going to link to the blog on the show notes or whatever. Sean Derrig: [01:19:31] Always happy to connect with people if you need introducing to any of my network. No problem. As long as you're not a dick. I'm happy to introduce people around. I'm at that stage in my previous career now where for me it's kind of more about legacy. It's more about, OK, so how can I leave the place a bit better than I found it? Because I've done all of the corporate stuff. I've I've done the boring, greasy pole stuff. So I don't need, you know, so anything I can do to give anybody a bit of a nudge or, you know, to help them out or whether it's advice, whatever the moment, I'm doing a lot of covid related stuff because the problem the organizations have in Sabitha. Right. We need to such a big organization. Right. Curvilinear. Right. Let's get Professor So-and-so and Professor So-and-so has never had a fucking job in his life. And it wasn't no commercial reality if it slapped an about face on it. So what you need is somebody who understands the sides really well and the bit where. Yeah, that sounds really plausible. Here's why, folks. Here's what you probably should be doing. A bit singularities this way. So it's quite rare to have a not not just all just nice, but having those scientific chops, but understanding the realities of running a business and the fact you can't just put on a massive, great control over your office building and say nothing's ever going to get in. It doesn't it doesn't work that way. So you need somebody who can do the science, but also is commercially where there's underclothes at the moment. So if I was to reach out on that, if they're struggling, I'm sure that I'll be able to phone call with you. Harpreet Sahota: [01:21:23] Yeah, definitely. I'll include a link to your blog, which is I encourage when to check it out. Like I don't like biology whatsoever, but I was going through some of the stuff in your blog and I was just laughing the entire way because it's really entertaining and well-written. And I'm sure if you guys have made it this far into this episode, you get some sense of... Harpreet Sahota: [01:21:43] He's got a talk about testicles having taste buds. Sean Derrig: [01:21:46] And what the other title of the article that was done there was this yogurt of a few years ago when Obama legalized gay marriage or to give it its proper term marriage, all of the religious wingnuts went fucking mental about anyway. That was your concept of, you know, clearly a young lady clearly in a state of undress under a yogurt and tubs, the football partner and there's a woman of the religious nurses went absolutely mental about it. And this one literally went, yeah, yeah. It's biology that is lactobacillus like yogurt. So, you know, so that was one that just came out as a stream of consciousness and didn't even need adjusting. I just revenge on these people, unlike the point to look at. It's not natural. So I debunk all of that and it really was a stream of consciousness. I make the point that, look, if if you're uncomfortable with the idea of two people of. If if you're uncomfortable with gay marriage or opposite marriage, it's fine, you can be as uncomfortable with that as you want, just don't marry a gay person and it will never, ever, ever, ever affect you, you know. Yeah, that they have. And they have a right to be outraged. We have a need for what? Not to give a flying fuck what they think it does. Go ahead. And so, yeah, that was that was that was a lot of fun to write. Yeah. That's a clue to links to that one as well. It's it's it it just it just does my head and it is all these, all these women like, you know, they, these people are so ignorant like Ben-Israel. This sense that Lightman's around them absolutely does more hating and claiming that they've got justification for them from their imaginary friend in the sky to be of bigots while pretending they're holding the rest of us on. Harpreet Sahota: [01:23:57] So, Sean, thank you so much, man, for taking time out, is scheduled to be on the show. I really appreciate it. It had a very entertaining morning. Sean Derrig: [01:24:10] Thank you very much for the invitation.