Latest / Second Nature / E179 Allister Furey: Manufacturing Trust

Transcript

E179 Allister Furey: Manufacturing Trust · Second Nature

• 0:06
  SPEAKER_01: It's a shame we haven't got you in Devon for this one.
• 0:08
  SPEAKER_00: Yeah, it would have been nice to get out into the country.
• 0:10
  SPEAKER_01: We'll invite you back again, it's fine.
• 0:12
  SPEAKER_01: You can apply the Rob Gardner rule.
• 0:14
  SPEAKER_01: He's invited himself several times to the point whereby he said, How many times do I have to appear before you let me host?
• 0:19
  SPEAKER_01: And we agreed at five.
• 0:22
  SPEAKER_01: So once he's done five, he can he can he can host it up.
• 0:25
  SPEAKER_01: Um I'll thank you for coming in and seeing us.
• 0:27
  SPEAKER_01: Um should probably start by telling people who you are and what you do.
• 0:31
  SPEAKER_00: Yeah, so um I'm Alistair Fieri.
• 0:33
  SPEAKER_00: I'm the CEO and co-founder of Silvera.
• 0:36
  SPEAKER_00: And uh yeah, Silvera, we're a we're really a data business for uh carbon and commodities.
• 0:41
  SPEAKER_00: Um we provide uh ratings, pricing on carbon credits, we provide um geospatial data and um data for uh the emissions uh status of commodity products like steel, cement, um this kind of stuff.
• 0:59
  SPEAKER_00: And yeah, that's us.
• 1:01
  SPEAKER_01: I I think you've probably got one of the most eclectic and fascinating backgrounds and routes to getting there.
• 1:08
  SPEAKER_01: Can you step people through the sort of chapters that have made up your career ahead of Silvera and how you came to start the business?
• 1:15
  SPEAKER_00: Yeah, I've just kind of followed my nose of what what is kind of interesting and where there's a bit of passion there.
• 1:21
  SPEAKER_00: Um always been interested in building uh the common threads, building businesses and doing impactful things and technology, probably the other three common threads.
• 1:33
  SPEAKER_00: Um I'm biologist originally, but um uh studied neuroscience, interested in AI since back in the day, focused on neural computation.
• 1:42
  SPEAKER_00: And what does that mean for people, neural computation?
• 1:44
  SPEAKER_00: It's basically uh the at the heart of most modern AI systems is kind of a kind of a very simplified version of um uh the way brain uh work the brain works.
• 1:56
  SPEAKER_00: So it's a neural network, there's little processing units which kind of are roughly analogous to neurons in your brain.
• 2:02
  SPEAKER_00: Uh they're connected up um uh with certain weights or strengths of connection, and it's the change of those weights of connections that like uh embody knowledge and allow the network to do something from some input to some output.
• 2:15
  SPEAKER_00: Um so yeah, when when when and people were trying to figure this out for for probably about a hundred years.
• 2:24
  SPEAKER_00: Um and yeah, so I picked that up maybe sort of in the early 2000s when I was at uni.
• 2:31
  SPEAKER_01: And and that was it was not cool then, right?
• 2:33
  SPEAKER_01: No one else was running towards this uh in the general sense.
• 2:36
  SPEAKER_00: No, it's it it's kind of come in and out of fashion as it's kind of like overpromised, slightly under delivered, under overpromised, undelivered.
• 2:44
  SPEAKER_00: But it at each stage there's been utility.
• 2:46
  SPEAKER_00: I think probably the same as true now, except that that the utility that's there is all actually uh obviously huge.
• 2:53
  SPEAKER_00: So, you know, AGI, uh, do you need it for it to be extremely, extremely valuable and useful?
• 2:59
  SPEAKER_00: No, actually, like there's so much you can do with what what is um what's up and running now that you don't need like um superhumans.
• 3:08
  SPEAKER_00: Uh you can already do like insane, and and there's so much value to create there, but but yeah, there has been these waves.
• 3:14
  SPEAKER_00: And at that time, it was more out of interest than anything that that I did it.
• 3:20
  SPEAKER_00: And also we didn't really want to cut you, you basically went the computational track, or you went the sort of slicing up mice brains, uh rat brains track, or putting um uh electrodes into the brains of rats while they went around mazes.
• 3:35
  SPEAKER_00: And I kind of like that's not really for me.
• 3:37
  SPEAKER_00: Uh like the the um the what you learn from a rat's brain by slicing it up and uh and then putting uh not that much, turns out.
• 3:45
  SPEAKER_00: So but you have to kill a lot of rats in the process.
• 3:48
  SPEAKER_00: So I wasn't really into that.
• 3:49
  SPEAKER_00: Whereas with um uh with uh software, you can model these systems and your learning rate is way, way faster.
• 3:56
  SPEAKER_00: And they did learn a lot of uh super interesting stuff that hasn't even made it in through to like widely deployed AI.
• 4:03
  SPEAKER_00: Um the computational neurosciences um has a huge toolkit.
• 4:07
  SPEAKER_00: And this is tends to be in what happened with this is a massive digression from the original question, but the um what tends to have happened is um the engineers hit a plateau and then they often look back to the biology, get an inspiration, like don't want to touch anything to do with the biology, horrible, complex, like messy stuff, and then milk everything they possibly can out of that sort of paradigm.
• 4:30
  SPEAKER_00: And then they hit plateau.
• 4:31
  SPEAKER_00: And and so this is kind of reinforcement learning, also comes from uh from biology.
• 4:36
  SPEAKER_00: So um, but there's literally like this massive, huge, long list of things that exist in biology that have yet to be harnessed and implemented in AI.
• 4:44
  SPEAKER_00: Um, and this is probably because of the power drawer of compute is so extreme.
• 4:50
  SPEAKER_00: That's probably where people are going to look back to the biology again.
• 4:54
  SPEAKER_00: It's like, how can I have something that doesn't like double the power requirements of the planet to get the use out of AI?
• 5:00
  SPEAKER_00: And it's more about efficiency probably than about uh performance.
• 5:04
  SPEAKER_01: But yeah, that's kind of my that pattern you just articulated is exactly the story that DeepMind tell about like the the Alpha Go programs and things that they literally used one branch as much as possible to try and achieve a step forward.
• 5:18
  SPEAKER_01: Okay, we were in a dead end, let's go back to the biology.
• 5:21
  SPEAKER_01: Because uh Demi Sasabis, he's a neuroscientist by background as well.
• 5:24
  SPEAKER_01: So whereas actually the more of the reinforcement learning lobbyists tend to be more mathematicians as opposed to the neuroscientists.
• 5:32
  SPEAKER_01: So I'm just I think it's so awesome that you were doing that in the early 2000s.
• 5:36
  SPEAKER_01: You know, DeepMind didn't start till like 14 or something, so you had a long run ahead of that, which is super cool.
• 5:43
  SPEAKER_00: Yeah, it's at 2014 to 2014 they started.
• 5:46
  SPEAKER_00: And yeah, so de Demis, he did his, he was doing his postdoc at the Gatsby Institute at UCL.
• 5:50
  SPEAKER_00: And UCL was is like a huge uh hub for uh neuroscience, and that so that's where I was at.
• 5:55
  SPEAKER_00: And um, so they were pretty good.
• 5:56
  SPEAKER_00: And they they let me and this other guy in my course, we were just we we kind of said they're doing stuff in computer science um that is relevant for us.
• 6:06
  SPEAKER_00: Can we kind of defect, partly defect from the um biology, it was anatomy school, and and do classes?
• 6:13
  SPEAKER_00: And there wasn't really an option to that, but they just let us do it.
• 6:16
  SPEAKER_00: So so that was kind of cool of them at the time.
• 6:19
  SPEAKER_00: Um and then yeah, I honestly I I um I didn't really like I didn't really know what to do, and I didn't really see like a kind of graduate track as the one that I wanted to go down, graduate intake sort of track.
• 6:33
  SPEAKER_00: Um, I didn't really appeal.
• 6:35
  SPEAKER_00: So I started a company, and that was just with my friends, and I knew absolutely nothing about entrepreneurship.
• 6:39
  SPEAKER_00: So 2003, we started a clothing company because it was You know where that itch came from to start a company.
• 6:45
  SPEAKER_00: It was more like it's something you can do that isn't a nine to five working for somebody else.
• 6:51
  SPEAKER_00: Uh and you is more active creation and building something, I don't know, is always uh appealed to me.
• 6:57
  SPEAKER_00: And uh you can fail uh well, you you have this hard failure function, and that kind of like um, you know, either it works or it dies, right?
• 7:05
  SPEAKER_00: And whereas in sort of a career, you can kind of muddle on for like most of your life in this sort of gray area of like getting paid, but does it really mean anything?
• 7:15
  SPEAKER_00: You only get to live once.
• 7:16
  SPEAKER_01: So back to your reinforcement learning there, right?
• 7:18
  SPEAKER_01: You either pass or fail.
• 7:19
  SPEAKER_00: Yeah, you have a good signal.
• 7:20
  SPEAKER_00: I mean, that you have a very strong signal when you're building a company, like, well, and if if the signal is like no, no, no, then you just stop doing it and do something else.
• 7:29
  SPEAKER_00: Whereas in yeah, in a kind of corporate career, it's just a bit of kind of grey, grey sort of middleness, isn't it?
• 7:36
  SPEAKER_00: Um so so I I that never really appear appealed to me.
• 7:39
  SPEAKER_00: Um academia also I don't I don't know, there's probably more I could do there, to be honest, but but not in them, it's just a bit slow paced, and you know, unless it does something in the real world, it's kind of I'm more interested in the applied end of things.
• 7:55
  SPEAKER_00: So we started this company and um uh we we we scaled it up, and then basically uh you need to care about fashion to run a fashion company.
• 8:05
  SPEAKER_00: And and um we we we kind of came to this mutual understanding that we didn't really care about fashion.
• 8:10
  SPEAKER_00: So I went back to the AI stuff and and I um but more on the computer science side, and and that Sussex is uh where I live now.
• 8:17
  SPEAKER_00: Um there's a really amazing uh group.
• 8:19
  SPEAKER_00: It was a Center for Computational Neuroscience and Robotics, and they had this program, and it was very much square in this sort of middle zone of philosophy, um, biology, and computer science.
• 8:30
  SPEAKER_00: And there's really amazing people um uh doing very interesting work in robotics, uh, AI, and philosophy all kind of in one group.
• 8:40
  SPEAKER_00: So I did a master's there uh actually kind of while I was building this company.
• 8:46
  SPEAKER_00: And um, at some point I was when we ran finished the company, I was like, okay, I either get a real job or I go and um try and level up in some way or build another company.
• 8:56
  SPEAKER_00: I didn't really know what to build at that point, so so I actually had a job offer doing um corporate finance, like MA.
• 9:02
  SPEAKER_00: Um uh and then I also had an offer to do a PhD.
• 9:06
  SPEAKER_00: Uh, and my kind of plan for the PhD was like build something that I could build a company around that would um uh have some defensible, something defensible to it.
• 9:16
  SPEAKER_00: Because with clothing, it's really your brand and your operations and your margins and stuff.
• 9:20
  SPEAKER_00: It kind of you have this life cycle of a clothing brand.
• 9:22
  SPEAKER_01: Were you were you aware enough of academia to understand the realities of PhD at that point?
• 9:27
  SPEAKER_00: Uh well, yeah, I mean, so so yeah, I think so.
• 9:30
  SPEAKER_00: So so my um my supervisor is a guy called Inman Harvey, amazing guy, extremely inspirational.
• 9:37
  SPEAKER_00: And um uh he was a philosopher, uh mathematician, actually, really, by training, Cambridge mathematician, kind of very interested in the philosophy of mind.
• 9:47
  SPEAKER_00: And um, but he would build uh models of um evolution and this kind of and and and neural next and this kind of stuff.
• 9:54
  SPEAKER_00: Um very unusual.
• 9:56
  SPEAKER_00: He used to work for MI6, and he'd he so he had a completely different life and then come into academia.
• 10:03
  SPEAKER_00: And I remember his talk on doing a PhD was like, don't do a PhD, career suicide, um, never earn any money, you've got to be mad.
• 10:12
  SPEAKER_00: You the only reason you should do it is if you feel like you have to do it.
• 10:15
  SPEAKER_00: Uh and I was like, yeah, I like that.
• 10:18
  SPEAKER_01: I do a very similar speech.
• 10:20
  SPEAKER_00: It's just a terrible idea.
• 10:21
  SPEAKER_00: It's just for most people, um, especially if you don't know what to do, then you should not do a PhD.
• 10:27
  SPEAKER_01: Um It doesn't go away.
• 10:28
  SPEAKER_01: It's just I remember writing up on Christmas Day and my in-laws being like, Oh, do you need to do that now?
• 10:33
  SPEAKER_01: And I'm like, it's not going anywhere else.
• 10:35
  SPEAKER_01: No, no, no.
• 10:36
  SPEAKER_01: Like it's just always here.
• 10:37
  SPEAKER_01: Yeah.
• 10:38
  SPEAKER_01: You just need to get through it.
• 10:39
  SPEAKER_00: You you kind of live with that.
• 10:41
  SPEAKER_00: You never don't feel like you shouldn't be working on it for the whole period.
• 10:44
  SPEAKER_00: Just go, yeah, yeah.
• 10:45
  SPEAKER_00: I don't know how you're how yours went awful.
• 10:48
  SPEAKER_01: Well, you um the thing about undergrad is you can write a terrible essay that you don't like, you can get a low score and it goes away.
• 10:55
  SPEAKER_01: Yeah.
• 10:55
  SPEAKER_01: It's yours forever.
• 10:57
  SPEAKER_01: Like it I guess it's the it's almost like having a child that just yells at you all the time.
• 11:02
  SPEAKER_01: It just keeps coming back with red writing on until it gets to that point.
• 11:06
  SPEAKER_01: Albeit I couldn't write when I started my PhD.
• 11:09
  SPEAKER_01: And I think that's the that and the abil ironically, the ability to synthesize a huge amount of information and pull out the nuggets are the things that I took the most away from it.
• 11:17
  SPEAKER_01: Yeah.
• 11:17
  SPEAKER_01: I mean AI can help with that now, but the ability to write and communicate was the thing that because of all those repetitions that that I really got from it.
• 11:25
  SPEAKER_01: Um and actually I was I wasn't ready for a a career after undergrad at all.
• 11:31
  SPEAKER_01: So I think it was I matured a lot during the PhD period.
• 11:34
  SPEAKER_01: And actually, that's when I properly started playing American football internationally.
• 11:38
  SPEAKER_01: It's where I went and visited the West Coast for the first time.
• 11:41
  SPEAKER_01: And that was 2000, literally 20 years ago.
• 11:44
  SPEAKER_01: And it was when the year after YouTube, sorry, the year after Google had IPO'd, YouTube launched that year, Facebook did.
• 11:51
  SPEAKER_01: If it hadn't been for the PhD and going and spending time in the US then, OC wouldn't be a thing now.
• 11:56
  SPEAKER_01: Because that was like the first realization that, oh, you can be a data geek, love tech, and start a business.
• 12:02
  SPEAKER_01: Because no one in the UK was in 2006.
• 12:05
  SPEAKER_01: Um so yeah, super glad I did it.
• 12:07
  SPEAKER_01: But my advice to anyone now is you have to really love what you're doing.
• 12:10
  SPEAKER_01: You really do.
• 12:11
  SPEAKER_01: Because otherwise it's a sentence.
• 12:13
  SPEAKER_00: Yeah, yeah.
• 12:14
  SPEAKER_00: Yeah, yeah.
• 12:14
  SPEAKER_00: You you you really you really do.
• 12:16
  SPEAKER_00: Um and uh yeah, I think I was trying to put a lot into it.
• 12:20
  SPEAKER_00: So it was like, I want a positive in environmental impact because it's something I had been um uh interested in for a long time.
• 12:27
  SPEAKER_00: And um I wanted to use the AI skills and I wanted to be it's something you could commercialise.
• 12:33
  SPEAKER_00: Uh and so really at that time AI skills.
• 12:36
  SPEAKER_01: Hell of a Venn diagram.
• 12:37
  SPEAKER_00: Yeah, exactly.
• 12:37
  SPEAKER_00: It's like this sliver is almost almost nothing.
• 12:41
  SPEAKER_00: So um so uh most of the applications at that point, sort of mid 2000s, were basically in uh finance or defense.
• 12:51
  SPEAKER_00: So it's like, but I I don't that doesn't really tick the the box of any positive social impact at all, right?
• 12:58
  SPEAKER_00: Um and uh that you're not you might as well not do a PhD, you might as well just go and work for a hedge fund or something.
• 13:04
  SPEAKER_00: Um so so uh what what was interesting at the time was there was a sort of a uh new class of wind uh energy system emerging that had a big control problem.
• 13:14
  SPEAKER_00: And I was like, well, uh AI can be used to solve control problems, uh especially um where the s like what you're trying to control isn't very well defined.
• 13:25
  SPEAKER_00: So so um so you uh you have an adaptive controller.
• 13:28
  SPEAKER_00: So what I was working on was kite power.
• 13:30
  SPEAKER_00: So you instead of using a wind turbine, um, you use a kite to intercept the wind.
• 13:36
  SPEAKER_00: That you basically need to fly acrobatically, something the size of a 747 without engines on it for 20 years without crashing the thing, without knowing the turbulence field that you're flying through, and everything's kind of squishy and like uh like the you have a lot, you have a tether and the the wing's very flexible.
• 13:55
  SPEAKER_00: And that's a really interesting control problem.
• 13:57
  SPEAKER_00: Um and you can, if you deploy it correctly, uh cut the cost of wind energy probably about half, which is material, right?
• 14:05
  SPEAKER_00: So um, so I was like, okay, I'll do that.
• 14:07
  SPEAKER_00: And and if that works, then I'll start a company doing that, or I'll go and work with the startups.
• 14:11
  SPEAKER_00: Google had funded a business after I started the PhD called um McCartney Power.
• 14:16
  SPEAKER_00: Another one called uh got started again while I was doing a PhD called Joe Joby.
• 14:21
  SPEAKER_00: They pivoted from that to doing uh flying electric vehicles.
• 14:25
  SPEAKER_00: Now Joby Aviation is a listed company.
• 14:27
  SPEAKER_00: So they took all the the tech that they used to develop uh for kite power and pivoted to that.
• 14:33
  SPEAKER_00: So it's kind of an interesting one.
• 14:34
  SPEAKER_00: Um Joe Ben's an interesting entrepreneur, like uh good guy.
• 14:38
  SPEAKER_00: Um yeah, and actually actually he offered me a job uh after my PhD.
• 14:43
  SPEAKER_00: And I didn't I thought their architecture didn't make sense for power, and it turned out it didn't did make sense for flying vehicles.
• 14:49
  SPEAKER_00: Um so we we managed to get funding uh to start a company then and um me and these guys I was doing some work with on the side um uh during my PhD.
• 15:01
  SPEAKER_00: And we we the first check came from Shell.
• 15:03
  SPEAKER_00: So Shell funded it uh initially, and um then Ion and Jean Berger, and we uh we we grew that business and uh we basically got to the point where we sold sold it to uh a competitor um as it came as basically uh this is some analogies to carbon markets is like the the what the premise of that was uh of kite power was that the um scaling uh of wind turbines would hit a wall and you wouldn't be able to get to um the low costs that everyone needed for it to be competitive.
• 15:40
  SPEAKER_00: And what actually happened was the wind turbines have scaled and keep scaling to like 20 megawatts, even offshore, you're seeing like uh extremely large systems, and they've they've done an amazing, amazing job stripped more power out.
• 15:53
  SPEAKER_00: Still, you could get to a cheaper power with a with a kite system, but the um the capital that there was the Google's putting in stuff kind of ran out, and the appetite to invest, you multi-billion dollar investment to get to anything at like a mature scale kind of ran out.
• 16:07
  SPEAKER_00: So there's a consolidation industry.
• 16:09
  SPEAKER_00: So we we uh we merged our RIP and our work with that of um kite mill, who's they're still going in Norway.
• 16:16
  SPEAKER_00: They're doing they've and they continue to progress.
• 16:17
  SPEAKER_01: And what was that process like being part of an MARC?
• 16:20
  SPEAKER_00: It it wasn't like, I mean, you you have different versions of this, right?
• 16:24
  SPEAKER_00: So so you're like uh, you know, you have six people in a room, you're printing money, and Facebook gives you a check for like a billion dollars, right?
• 16:31
  SPEAKER_00: Uh this is more like or and you have versions which are more like uh like how what's our option now?
• 16:39
  SPEAKER_00: Are we gonna slog it out for another 10 years um uh on a very small amount of money, or should we should we just pull forces?
• 16:46
  SPEAKER_00: And it was kind of the the latter.
• 16:48
  SPEAKER_00: Uh so at that point, um, it wasn't like um, I mean, you don't see billions of dollars, it's not like a hundred billion dollar market for um uh high altitude wind power.
• 16:59
  SPEAKER_00: People starting to get more, and actually, and it's interestingly, almost certainly where this will get developed is China, because they're able to innovate extremely quickly and they have more of the manufacturing supply chain.
• 17:13
  SPEAKER_00: And now so what we're seeing now is a resurgence of a high-altitude wind powered um thinking development in China, uh, which is kind of a shame, to be honest, because if the West had continued developing it for the last uh what getting on like 20 years now, actually we'd have an industry in high-lastroid wind power, but probably will the Chinese will develop it and or power to them.
• 17:34
  SPEAKER_00: Like good good luck to them, because it's a it'll break through that price fall.
• 17:38
  SPEAKER_00: And it's more of a baseload power because there's always been a lot of people.
• 17:40
  SPEAKER_01: These things have wavelengths.
• 17:42
  SPEAKER_01: 100%.
• 17:42
  SPEAKER_01: And it's just when it's gonna I I look back on be it carbon markets or AI and tech developments, and you can see when something's advanced a little bit, then dropped away and advanced a bit more.
• 17:52
  SPEAKER_01: 100%.
• 17:52
  SPEAKER_01: People think it's a leap.
• 17:54
  SPEAKER_01: Rarely is it a leap.
• 17:55
  SPEAKER_01: It's a series of like iterations and tacking towards an objective.
• 17:59
  SPEAKER_00: Yeah, tacking is a nice way of putting it.
• 18:01
  SPEAKER_00: So so uh yeah, you you do have periods of progress, periods of stagnation, periods of progress.
• 18:06
  SPEAKER_00: It's never so smooth, right?
• 18:08
  SPEAKER_01: Aaron Ross Powell I think the value is where you get that stagnation or drop-off is looking to a different sector or a different realm to bring in an idea or a concept that helps move it forward because often the learning isn't within the narrow vertical you're working in.
• 18:21
  SPEAKER_01: You can bring it in from elsewhere.
• 18:23
  SPEAKER_00: Yeah, yeah, yeah, yeah, definitely.
• 18:25
  SPEAKER_00: And and in in that field is like it's materials, it's sensors, uh and compute um and the development costs as well.
• 18:37
  SPEAKER_00: So all of those things are in a completely different place.
• 18:40
  SPEAKER_01: When you were doing your first AI chapter, was compute as a term used?
• 18:45
  SPEAKER_01: Was it regularly talked about in terms of computer processing power?
• 18:48
  SPEAKER_00: It was.
• 18:50
  SPEAKER_00: Um I remember during my uh during my NMO PhD, the stuff I had to do was quite computationally intensive.
• 18:56
  SPEAKER_00: So we had to simulate the flight of these um these wings and uh train the control systems, and you would train sort of hundreds of thousands of hours of simulated flight to to get them to learn about all the different uh circumstances they'd have to deal with.
• 19:10
  SPEAKER_00: And um uh there'd been a high performance uh cluster bought by Sussex Uni, and uh they hadn't started the project, which is around actually um moth pheromone, how how how the brains of moths like uh insanely, insanely sensitive to smell, like unbelievably, like to the almost to the limit of physics.
• 19:35
  SPEAKER_00: Um and so they'd got this comp, but they hadn't got ready to start the work.
• 19:39
  SPEAKER_00: So I just stole this cluster for like six months.
• 19:42
  SPEAKER_00: There's literally like eight a rack of 80 machines that I was using, and I just had them maxed out or like the whole time.
• 19:48
  SPEAKER_00: And uh that kind of got me through the PhD.
• 19:51
  SPEAKER_00: And then at one point I got this email being like, um you're using this cluster that was for this in this multimillion dollar research project.
• 19:59
  SPEAKER_00: And there was that was just the first use.
• 20:01
  SPEAKER_00: Uh and and can we have it back now, please?
• 20:03
  SPEAKER_00: And I was like, Oh sh uh, yeah.
• 20:04
  SPEAKER_00: Yeah, you can have it back.
• 20:05
  SPEAKER_00: Sorry, sorry, sorry.
• 20:06
  SPEAKER_00: But it kind of got me over the opportunity when it's there.
• 20:09
  SPEAKER_00: Yeah, yeah.
• 20:09
  SPEAKER_00: I mean, it was just sitting there doing nothing.
• 20:11
  SPEAKER_00: So so it's it was it was better that it got put to use than just depreciate in a in a in a darkened room somewhere with the AC on.
• 20:18
  SPEAKER_01: And did you get to step away from the business at the point of sale or were you tied in for a while?
• 20:23
  SPEAKER_00: I no, no.
• 20:23
  SPEAKER_00: I'd I'd already segued out before that.
• 20:25
  SPEAKER_00: Okay.
• 20:26
  SPEAKER_00: Yeah, because it was kind of it was clear where I was going.
• 20:28
  SPEAKER_00: It was clear where it was going.
• 20:29
  SPEAKER_00: And then um Yeah, and then I then basically, and you've got kids.
• 20:33
  SPEAKER_00: At some at some point I did need a job.
• 20:35
  SPEAKER_00: So I uh because I had kids and you know, you know, just some stable income would be nice.
• 20:41
  SPEAKER_00: Um so I worked for Bain for a for a bit um uh in Magical Existency, and then I um uh I'd done an MBA before that because uh what it's a particular thing in the UK, actually, I think, um, but put maybe globally is like um if you have a PhD, people assume that you can't think about anything commercial.
• 21:01
  SPEAKER_00: Um so in in that business I was CTO, but I was doing all the cost optimization and like writing half the business plans.
• 21:08
  SPEAKER_00: Um so it's like obviously I knew about the you know, you can't develop anything in the power sector without economics being like the single thing you're optimizing for.
• 21:17
  SPEAKER_00: Um so but I'd done an MBA and during that I basically worked my whole way through it, I wasn't, you know, um paid.
• 21:23
  SPEAKER_00: And the work was probably more useful than the MBA, to be honest.
• 21:26
  SPEAKER_00: Um and one company I worked for was developing um high performance heat heat exchange technology, and I did did some work for them to figure out ways they could sell it into other markets.
• 21:36
  SPEAKER_00: And they poached me out of Bain when they raised a load, they raised 20 million from uh BA systems and uh they said, Oh, there's projects you did for us.
• 21:45
  SPEAKER_00: Can you come in and basically sp set up this spin off unit that um that you said we should in the in in the project?
• 21:53
  SPEAKER_00: And and that actually was really successful.
• 21:55
  SPEAKER_00: Uh that ended up being the vast majority of uh the revenue.
• 21:58
  SPEAKER_00: In fact, I think Over 90% of the revenue for that business came from the spin-out unit that I set up.
• 22:04
  SPEAKER_01: Trevor Burrus Just going back to your observation about the PhD, I think is really pertinent.
• 22:08
  SPEAKER_01: And I think it is changing, but go back to 2015, to sort of 2018 when I was leaving the public sector.
• 22:17
  SPEAKER_01: I I remember private sector organizations I talked to, they viewed a PhD on a CV as a concern.
• 22:23
  SPEAKER_01: Trevor Burrus, Jr.
• 22:23
  SPEAKER_00: Yeah, 100%.
• 22:24
  SPEAKER_01: I remember they were like, oh wow, we were really worried you wouldn't be able to talk.
• 22:28
  SPEAKER_01: I was like, what?
• 22:30
  SPEAKER_01: Well, you'd be too academic.
• 22:31
  SPEAKER_01: And it I there's so many cultural differences between, you know, before we started recording, we were talking about I was just coming back from San Fran and Climate Week and Evening AI.
• 22:39
  SPEAKER_01: I don't think that was ever the case in the US.
• 22:41
  SPEAKER_01: It always would have been a, oh, cool.
• 22:43
  SPEAKER_01: You can think quickly, you can process stuff, you're in you're interested.
• 22:47
  SPEAKER_01: Let's do that.
• 22:48
  SPEAKER_01: Whereas actually we've had such a almost a reaction to PhDs, which I think are dropping away now because of the prevalence of AI.
• 22:56
  SPEAKER_01: But yeah, I do agree it was probably more of a barrier than it was a help to start with.
• 23:00
  SPEAKER_00: Yeah, 100%.
• 23:01
  SPEAKER_00: And then and then the the annoying thing, and then again, you don't get it in Silicon Valley, is like um now if you're on the commercial side, people assume that you can't code or do anything technical.
• 23:11
  SPEAKER_00: Yes.
• 23:12
  SPEAKER_00: Hold on, I could literally I could code and build this entire system if you want me to.
• 23:16
  SPEAKER_00: I just like uh it's the tacking point again, though.
• 23:19
  SPEAKER_01: Like because we make a bit of progress in people's understanding that people can be multifaceted and can understand numbers and also words.
• 23:26
  SPEAKER_01: And you know, there is a blessing.
• 23:28
  SPEAKER_01: Exactly.
• 23:28
  SPEAKER_01: There is a blending to this part of it.
• 23:30
  SPEAKER_01: Yeah, yeah.
• 23:30
  SPEAKER_01: I think we are arriving there now.
• 23:32
  SPEAKER_01: Um everyone who's that I know who's worked in a management consultancy perspective, tend to speed past that chapter.
• 23:40
  SPEAKER_01: What what were the best parts about that chapter?
• 23:44
  SPEAKER_01: Um other than it leading to a fantastic job, by all accounts.
• 23:48
  SPEAKER_00: Uh I didn't last very long.
• 23:50
  SPEAKER_00: Um as in when I got the call, did I want to go and uh leave to go and work in this um sort of a gross stage business, a couple hundred people?
• 24:01
  SPEAKER_00: I was like, yes, definitely.
• 24:03
  SPEAKER_00: Uh uh, I think the thing that I learned uh was like they always they say lead with the answer.
• 24:11
  SPEAKER_00: And then so what you do have with technical people is they want to walk you through the very detail to get to the answer.
• 24:18
  SPEAKER_00: And they do uh so in a commercial context, you just want to be like, and this is the thing.
• 24:22
  SPEAKER_00: Uh and then if you want to explain all the rationale, um, then great, but typically you don't have to.
• 24:28
  SPEAKER_00: So so leading with the answer, and then uh the other thing is if you want to get agreement from somebody, um, or group of people in particular, uh so what was what they're very good at is basically when you go into a room that's supposed to be presenting the report or whatever, you're you know already um what you want to happen after that.
• 24:53
  SPEAKER_00: And um every single person in that room you've pre-aligned with.
• 24:58
  SPEAKER_00: So we call you call it pre-aligning.
• 25:00
  SPEAKER_00: So so if you you're if you put people in a room and you you kind of get them to expose themselves, um people think about uh so like we're trying to decide X or Y or Z.
• 25:12
  SPEAKER_00: Um and uh people on the spot think, oh, what does Y mean for me and like my team?
• 25:19
  SPEAKER_00: And you know, and they get a bit defensive, and you never get anywhere because people um are kind of reacting in the moment.
• 25:25
  SPEAKER_00: So what they'd always do is basically say, okay, this person's beef with this might be that, and let's get ahead of it.
• 25:32
  SPEAKER_00: And and so that the um when you presented something wasn't presentation at all.
• 25:37
  SPEAKER_00: It was just yeah, it was just getting people to agree on the thing that everyone had actually pre-aligned with everyone.
• 25:43
  SPEAKER_00: So everyone's in there being like, oh, we already agree on this thing.
• 25:45
  SPEAKER_00: Uh how amazingly convenient, but that that that was um very much deliberate.
• 25:50
  SPEAKER_00: Uh and I I just thought there was uh it's not like um I wouldn't say that sort of dark arts thing.
• 25:54
  SPEAKER_00: It's just complete pragmatism.
• 25:56
  SPEAKER_01: Yeah.
• 25:56
  SPEAKER_01: It's just and is that actively coached and taught to the new people that come in, or it's more of a you know, that's more of a partner thing.
• 26:04
  SPEAKER_00: Um so you you you and it's just born of absolute pain of of presenting ideas in rooms and and it just spiralling by people who are extremely invested uh in the outcomes and the impacts of things, and and uh you just can't leave it to chance.
• 26:22
  SPEAKER_00: Um so so I think it was it's just an extremely practical thing.
• 26:26
  SPEAKER_00: Uh, if you ever want to get agreement from people, you kind of need to know that what is what's up for grabs as point of agreement.
• 26:33
  SPEAKER_00: Because if you put something up that nobody can possibly agree on, well, it's just a waste of everyone's time.
• 26:38
  SPEAKER_00: Yeah.
• 26:39
  SPEAKER_00: Um, so you just got to what's the minimum, what's the thing that everyone could agree on that will move things forward and just focus on that thing.
• 26:45
  SPEAKER_00: Uh and anything else is just like is kind of a waste of time.
• 26:49
  SPEAKER_00: So so um, yeah, I think that that those two those two learnings, just like focus on the answer and you know, focus on the things where you can make agreement and and take the chance out of it.
• 27:01
  SPEAKER_00: The valuable learnings, but you don't need to spend like five years of your career doing you know, 60, 70 hour weeks to um doing PowerPoints uh to get those.
• 27:12
  SPEAKER_00: I mean, like it comes and the the thing that you you you are compelled to do is like focus on what's important and ask questions from first principles.
• 27:21
  SPEAKER_00: Um but you get you get that from a PhD.
• 27:23
  SPEAKER_00: And I think that's what people don't understand with PhDs, is like um almost by definition, a PhD has to do something new.
• 27:30
  SPEAKER_00: And then to ask something new, you have to be like uh you have to have an opinion, and you've got to be able to derive that opinion by thinking for yourself, which nowadays is like a very valuable thing, right?
• 27:42
  SPEAKER_00: Thinking for yourself.
• 27:43
  SPEAKER_01: Completely.
• 27:44
  SPEAKER_01: Uh I think something that I've become guilty of post-PHD is not looking to history enough.
• 27:52
  SPEAKER_01: I think there's a tendency to be like, oh, this is the first time this problem's presented.
• 27:58
  SPEAKER_01: Where it's actually like, no, no, no, hold on.
• 28:00
  SPEAKER_01: Let's let's learn the background of this more.
• 28:02
  SPEAKER_01: Let's dig in and find out what's gone before us and learn everything we can from history.
• 28:06
  SPEAKER_01: And then peripheral to that as well.
• 28:07
  SPEAKER_01: What are the things that are sort of tangentially related that we could learn from?
• 28:11
  SPEAKER_01: Um I don't know when I lost that.
• 28:12
  SPEAKER_01: But because PhD, you have to.
• 28:14
  SPEAKER_01: You pull every thread of every publication back, and then you look at that reference list, and then that reference list, and you understand everything that's gone before.
• 28:20
  SPEAKER_01: I don't think we do that now.
• 28:22
  SPEAKER_01: I think we kind of limit our thinking to the last couple of years or what we're aware of rather than actually going back in history.
• 28:29
  SPEAKER_01: Maybe the more focused we are on the future, the more we're losing the past.
• 28:32
  SPEAKER_01: But I've tried to lean into that a little bit more.
• 28:34
  SPEAKER_01: And in fact, set up various skills and projects that automate me to ask what would history teach me about this?
• 28:40
  SPEAKER_01: Yeah, yeah.
• 28:40
  SPEAKER_01: Because I think we're losing a lot by not looking there.
• 28:42
  SPEAKER_00: Oh, I yeah, I totally agree.
• 28:44
  SPEAKER_00: And you're forced to respond so quickly now that you just want the the quickest sensible, useful thing that allows you to progress.
• 28:52
  SPEAKER_00: And maybe you lose in some cases, you're kind of lose a lot uh by by by doing that.
• 28:58
  SPEAKER_00: So yeah, any defense against that would is valuable, right?
• 29:02
  SPEAKER_01: Yeah.
• 29:03
  SPEAKER_01: Um the um for people who who are listening to this because they they like the AI chat that we're weaving in, um The Atomic Human by somebody Lawrence.
• 29:15
  SPEAKER_01: Um it goes back and and charts the development of AI from Bletchley Park.
• 29:19
  SPEAKER_00: Nice.
• 29:20
  SPEAKER_01: And the first computers and um flowers and Turing and how they developed and things.
• 29:26
  SPEAKER_01: And actually, once you I I've always been obsessed by Bletchley.
• 29:29
  SPEAKER_01: My wife did her undergrad dissertation on it, and I spoiler all that, I wrote the intro and conclusion, and since then I've been obsessed about it.
• 29:36
  SPEAKER_01: Um but actually it makes you appreciate and understand AI to a much greater degree, following that thread back to, as you say, almost a hundred years, you know, net 42.
• 29:45
  SPEAKER_01: I I still think one of the greatest achievements of humanity is breaking the Enigma code on paper and pencil.
• 29:51
  SPEAKER_00: Oh, it's amazing.
• 29:52
  SPEAKER_00: What was done was extremely um extremely basic tools.
• 29:56
  SPEAKER_00: Yeah, amazing.
• 29:58
  SPEAKER_00: Amazing building the tools to do it.
• 29:59
  SPEAKER_01: Take us back from so we've we've exited management consultancy.
• 30:03
  SPEAKER_00: Yeah.
• 30:04
  SPEAKER_00: Uh yeah.
• 30:04
  SPEAKER_00: So then I worked for this business called Reaction Engine.
• 30:06
  SPEAKER_00: They had um the highest performance heat exchangers in the w in the on the planet by some margin, to the extent where we would um travel to Northrop Grumman and the Skunkworks.
• 30:19
  SPEAKER_00: We went out to the Skunkworks at Lockheed Martin, uh, and they could not believe the the performance that we were able to get.
• 30:25
  SPEAKER_00: It's basically like transferring heat from from very hot gas to um uh uh uh to to to like some other thing.
• 30:34
  SPEAKER_00: And then we the the they developed that to as a key component of aerospace engines that could propel a space plane to orbit.
• 30:43
  SPEAKER_00: Um so you'd fire off an un runway into orbit, uh uh, which is like uh probably sort of it's it's a reasonably optimal configuration uh because you 100% usability.
• 30:55
  SPEAKER_00: So this is before SpaceX.
• 30:57
  SPEAKER_00: Um the problem was you have some scaling effects in in space uh access that require a very large vehicle to do that.
• 31:06
  SPEAKER_00: Uh and um they uh were quite focused on this idea of having a vehicle that would fly all the way to orbit, which is hundreds of at least tens of billions of dollars of development, maybe hundreds.
• 31:20
  SPEAKER_00: Uh and um there wasn't really market for that.
• 31:24
  SPEAKER_00: Um uh timing point again, right?
• 31:27
  SPEAKER_00: Yeah.
• 31:28
  SPEAKER_00: And at the same time, um SpaceX was just starting to show that they could um that they could slash the cost of space access because they were just starting to get landings of Falcon 9.
• 31:41
  SPEAKER_00: Um and uh it wasn't proven they could reuse those boosters, um uh, but they they were just starting to do it.
• 31:50
  SPEAKER_00: Um so so there was uh there was this kind of question.
• 31:54
  SPEAKER_00: Um, hey, do we have tens of billions of dollars and a customer uh uh that's willing to pay for the product at the end of it or or not?
• 32:03
  SPEAKER_00: And so we, I mean, I was in the in business development now, which is kind of cool.
• 32:07
  SPEAKER_00: It's just like a completely complete flip from being the um so I had a crash course in um in the uh in the engineering of propulsion and high speed propulsion systems and all the manufacturing, which is amazing.
• 32:19
  SPEAKER_00: But fundamentally, there was no customer.
• 32:21
  SPEAKER_00: So we're like, well, hold on.
• 32:23
  SPEAKER_00: Um let's develop something that does have a customer.
• 32:26
  SPEAKER_00: And there was two tracks there.
• 32:27
  SPEAKER_00: Um, who would use the version of the this engine, this hypersonic engine that one would fit on the table, and instead of one that would uh like this table, uh, you could build one.
• 32:39
  SPEAKER_00: Um and that would power a uh a drone that could maybe get up to Mach 4, Mach five.
• 32:45
  SPEAKER_00: Um, or uh and not one where it's the size of a coach, like a National Express coach with a one engine, right?
• 32:53
  SPEAKER_00: Not the vehicle, the engine.
• 32:55
  SPEAKER_00: Um so uh the um and basically the applications of that engine are obviously military, right?
• 33:04
  SPEAKER_00: Um so um so that was one thread.
• 33:07
  SPEAKER_00: And then the other thread was like, well, you've got the most high performance sea exchanges in the world.
• 33:12
  SPEAKER_00: Surely they must be usable for something that isn't this space plane that might might see the light of day in 2060 or something.
• 33:21
  SPEAKER_00: Um so so uh it turns out both things were quite true.
• 33:25
  SPEAKER_00: Um uh there are defense use cases, um, but there was reasons why the company didn't want to explore those.
• 33:33
  SPEAKER_00: Um so I'm not like a big defense uh advocate, but I'm I I was I was like I didn't pick a defense tech company to go work for.
• 33:40
  SPEAKER_00: Um uh the thing that was to me was the second that was interesting was the second thing, the spin-out technologies.
• 33:45
  SPEAKER_00: But the um but clearly they needed to pivot hard and immediately kill this single-stage Torbit thing, uh, because it was gonna kill a company uh if they didn't do that.
• 33:55
  SPEAKER_00: So um uh and and funnily enough, a couple years after I we made this diagnosis, uh a company in the US got started called Hermes, which basically what we'd mocked up and we'd gone to speak to the defense community about what they uh want to be like a drone that you could fly anywhere in the world, uh and nobody could stop you flying this thing, it would pop into orbit uh so you could spy on insert country name X.
• 34:21
  SPEAKER_00: Uh and uh even if you had a space-denied environment, which is basically like um somebody just taking out all the satellites uh so you could have that.
• 34:30
  SPEAKER_00: And that people will pay money for that.
• 34:32
  SPEAKER_00: And then Hermes basically got started to build that thing.
• 34:35
  SPEAKER_00: So we're like, well, we literally just did the discovery on this two years ago, and there's a market for that.
• 34:40
  SPEAKER_00: So we're like, if you want to have a market for this um technology set of technologies, there is only one, and that is that's it.
• 34:47
  SPEAKER_00: Um so that was one part of the story.
• 34:49
  SPEAKER_00: And then the other one was like, what could you do with this technology uh in other applications?
• 34:54
  SPEAKER_00: And you can use those um heat exchangers in motorsport, in clean tech, in aerospace.
• 35:00
  SPEAKER_00: And so I started a group that started to develop, and and so our first customer, I think, was McLaren.
• 35:05
  SPEAKER_00: Uh so uh the side pods on the McLaren vehicles radically changed shape a few seasons back, and that was because of the heat exchangers that our team supplied them.
• 35:16
  SPEAKER_00: The the sad part of that story is the the business pursued the single-stage orbit space plane despite it having no legs at all.
• 35:29
  SPEAKER_00: So so all of the spin-out part of the business grew and it did, it was doing uh eight-figure revenue, like amazing client base, um most of the revenue of the business.
• 35:40
  SPEAKER_00: Obviously, there was no demand for the actual product that had been developing for 20 years.
• 35:45
  SPEAKER_00: Uh and very late in the day, they started to try and pivot to the defense use case too late.
• 35:51
  SPEAKER_00: But uh I I I basically had said, there's no point me.
• 35:56
  SPEAKER_00: I I dis I just disagree with the direction of the business.
• 35:59
  SPEAKER_00: So there's no point me staying here.
• 36:00
  SPEAKER_00: Um uh so so I I I I kind of like, I'm not gonna just have this cognitive dissonance.
• 36:06
  SPEAKER_00: That's clearly how it is, and and so I I'd left at that point.
• 36:10
  SPEAKER_00: Uh the business cost folded, and then it got all that IP got sold to Honeywell, who obviously are just milking it for all of the applications uh uh that we'd identified.
• 36:22
  SPEAKER_00: It was just, it was, it was like this is um this is what the UK gets wrong about technology development.
• 36:30
  SPEAKER_00: It's kind of like you have to be market-led.
• 36:33
  SPEAKER_00: Uh it's like uh if they'd focused on just serving the applications that would pay, um, they would be killing it now.
• 36:42
  SPEAKER_00: Unfortunately, they didn't.
• 36:43
  SPEAKER_00: A good example of this is that aerospace company, Boom.
• 36:46
  SPEAKER_00: And and I'm a bit I'm a bit skeptical as to whether you're gonna have a supersonic um uh passenger jet developed purely by um like a startup because they have to develop their own engine.
• 36:59
  SPEAKER_00: Anyway, so so um they I think they realized this too.
• 37:04
  SPEAKER_00: That was they were probably never gonna get it to market.
• 37:07
  SPEAKER_00: Um, but they had started to develop their own engine because I think GE were gonna supply the engine and uh G was like there's they they were like, G is like there's no market, there's no point in us developing an engine.
• 37:17
  SPEAKER_00: So Boom has started to develop an engine.
• 37:19
  SPEAKER_00: They've basically completely pivoted to supply power, uh, so gas turbines to power AI data centers.
• 37:26
  SPEAKER_00: So but that's what you do in the States, right?
• 37:27
  SPEAKER_00: You're like, oh, the business isn't working.
• 37:29
  SPEAKER_00: Okay, let's revise it and we'll we'll orient it around where there's revenue.
• 37:33
  SPEAKER_00: And um like I don't know what the what what it is.
• 37:37
  SPEAKER_00: Uh maybe UK, Europe, European culture, a bit less flexible in that respect.
• 37:42
  SPEAKER_00: It's just like um it's this kind of quality of the signal.
• 37:45
  SPEAKER_00: There's a signal, you either listen to it or you don't.
• 37:47
  SPEAKER_01: So so I'd I'd um was that the biggest learning from that that period that you've you've got to follow the commercial signal?
• 37:54
  SPEAKER_00: Yes, yes, yes.
• 37:54
  SPEAKER_00: Yeah.
• 37:55
  SPEAKER_00: I mean it's kind of self-evident, right?
• 37:57
  SPEAKER_00: But but it's like uh it was what's strange is the um the willingness to for people to um just they want security and they want to carry on doing the things the way they are, obviously, right?
• 38:11
  SPEAKER_00: Um to change and to confront the fact that you're not like quite on the right track is challenging.
• 38:17
  SPEAKER_00: And especially when you have an organization that's going this way and you're like, no, no, we need to go this way.
• 38:20
  SPEAKER_00: It's difficult.
• 38:21
  SPEAKER_00: Um, but 100%.
• 38:23
  SPEAKER_00: And there's an analogy there with carbon markets, I think, as well, and how Severov um responded to what's happened in the market.
• 38:30
  SPEAKER_00: Um and uh so so the the other business that I um worked with when I was doing MBA was uh uh is kind of a fund called Entrepreneur First, or it's now called Entrepreneurs First.
• 38:41
  SPEAKER_00: And they basically helped quite technical people start businesses and they would uh fund them very, very, very early stage.
• 38:50
  SPEAKER_00: Um and so those guys had started an office in Singapore, they got some early validation, and they basically wanted somebody to help kind of um mature the operations in Singapore.
• 38:59
  SPEAKER_00: And so they'd given me a call.
• 39:00
  SPEAKER_00: And I uh at the point I was getting um a come to the end of the sort of strategic misalignment of reaction engines they're like, do you want to come out to Singapore and build this thing?
• 39:10
  SPEAKER_00: So I and that was good.
• 39:11
  SPEAKER_00: And um we did that in Singapore, we got Singapore up and running doing really well.
• 39:15
  SPEAKER_00: And then did you have partner and children at that point as well?
• 39:19
  SPEAKER_00: I did, yeah.
• 39:19
  SPEAKER_00: And you took them with you?
• 39:20
  SPEAKER_00: I did, yeah.
• 39:21
  SPEAKER_00: Yeah.
• 39:21
  SPEAKER_00: Big leap.
• 39:22
  SPEAKER_00: It it it was, but the my son wasn't in school at that time.
• 39:26
  SPEAKER_00: So it's kind of fine.
• 39:27
  SPEAKER_00: Uh once kids are in school, as you know, you don't want to be moving them around.
• 39:31
  SPEAKER_00: But he was only like um he was only a few months old at that point.
• 39:33
  SPEAKER_00: So it's like, okay, we'll go, we'll see how it is.
• 39:36
  SPEAKER_00: And that was a really good experience.
• 39:37
  SPEAKER_00: And there we had this sort of parallel.
• 39:39
  SPEAKER_00: So I was there for maybe a year and a half, and we we got like we we funded about 50 companies.
• 39:45
  SPEAKER_01: Wow.
• 39:46
  SPEAKER_00: Very technical, quantum, AI, uh, biotech, um uh more like software, regular software businesses, marketplaces, crypto.
• 39:54
  SPEAKER_00: Like so we saw this massive spread.
• 39:55
  SPEAKER_01: So what was the route into that?
• 39:57
  SPEAKER_01: How did they find the founders and things?
• 40:00
  SPEAKER_00: Um they found so it's a funnel.
• 40:04
  SPEAKER_00: It's like a it's a funnel.
• 40:05
  SPEAKER_00: Everything's uh everything's a kind of a funnel.
• 40:07
  SPEAKER_00: So they had inbound, so so people would find them, yeah, and they'd have outbound.
• 40:12
  SPEAKER_00: So they would outreach to um uh to people that they thought had the characteristics of a of a founder.
• 40:19
  SPEAKER_00: Um and so yeah, we would um have we would outbound thousands and thousands of people everywhere.
• 40:32
  SPEAKER_00: Yeah, yeah.
• 40:32
  SPEAKER_00: But it was extremely selective as well.
• 40:34
  SPEAKER_00: So so we would say that the um the success rate of getting in was lower than Harvard.
• 40:40
  SPEAKER_00: Uh and it's I think probably even more selective now.
• 40:43
  SPEAKER_00: Um and they uh they now, it's interesting, they now funnel all of their startups to to the West Coast.
• 40:52
  SPEAKER_00: So so it's kind of like um like thankfully, it's not like full brain drain, just like remove all the talent and reprice them.
• 40:59
  SPEAKER_00: But there is a bit of a talent arbitrage play.
• 41:02
  SPEAKER_00: So they remove them, they they take them there, uh, they spend several months there, they usually get funded there, and then they they relocate to wherever the market is.
• 41:10
  SPEAKER_00: So if the market was in Europe, they'll come back to Europe, and if their market is in the US, they'll probably stay in the US.
• 41:15
  SPEAKER_00: Um, that's a really, really interesting uh organization.
• 41:19
  SPEAKER_00: And they've um some amazing people have come through that program, like the founders of um Superbase, uh, which is uh one of the most um popular databases and for AI and vibe coding, the most popular.
• 41:33
  SPEAKER_00: Um the the their two founders um were on the program in Singapore.
• 41:38
  SPEAKER_00: They actually started a different program there and then they started Superbase afterwards, but that's a five-bit dollar company now.
• 41:43
  SPEAKER_00: Um most of the uh databases behind Lovable, et cetera, are Superbase databases.
• 41:47
  SPEAKER_01: Was there of the founders that were successful, were there some common traits that you recognized through that process?
• 41:53
  SPEAKER_00: Uh yes.
• 41:54
  SPEAKER_00: Yeah.
• 41:55
  SPEAKER_00: I mean the the thinking kind of changed uh over time because you don't really see, you know, it takes several years to but you could see who would get um disproportionately who would be disproportionately successful.
• 42:10
  SPEAKER_00: Um you it it was uh it it was people who were leveraging uh some uh some skill or knowledge they had that was typically differentiated.
• 42:24
  SPEAKER_00: So it wasn't like um and the classic thing is you would get a the these two two it's this founder that was I I love them so bits, Iranian guy, um and he was uh a specialist in fluid dynamics, and um his uh wife actually was a uh AI uh medical imaging AI specialist.
• 42:47
  SPEAKER_00: Uh and he was like, we were at this sort of event, and he was like, Yeah, I've got this thing where he's basically he wanted to have like um I called it being rude, um what was it?
• 42:59
  SPEAKER_00: Uh Airbnb for tent pegs.
• 43:02
  SPEAKER_00: So his thing was like, I don't want to buy a tent uh I uh because I but I would do want to go to this festival or something.
• 43:08
  SPEAKER_00: So so I want to have this sort of, I want to rent the tent.
• 43:12
  SPEAKER_00: And it's like, dude, what is your right to win in this business versus like the other five billion people on the planet?
• 43:18
  SPEAKER_00: Uh and it's like, oh, okay, good point.
• 43:20
  SPEAKER_00: It's like, what do you know that other people don't know?
• 43:23
  SPEAKER_00: Um, because T people tend to discount that a lot.
• 43:26
  SPEAKER_00: Uh and in the end, he was he ended up starting a company with his wife, and they were basically like, how can you um uh use the AI uh medical imaging, create 3D um uh views of arteries, and then model the uh the flow of blood in people's neck arteries and see um for people that had plaques, these kind of fatty deposits, which ones were going to shear off into the bloodstream and give them a stroke, and which ones weren't, because that's an uns completely unsolved problem.
• 43:59
  SPEAKER_00: And you Needed to bring the AI and the fluid dynamics together.
• 44:01
  SPEAKER_00: And they they so they raised money on that basis.
• 44:03
  SPEAKER_00: Um they ended up focusing much more on the medical imaging and just the workflow of the uh imaging people because it's actually to do with the economics of treating those patients and the risks to do with treating like stroke risk.
• 44:18
  SPEAKER_00: But the uh and and actually turns out there's just loads of radiologists doing extremely um grunty kind of manual labeling work the software's just better to do.
• 44:26
  SPEAKER_00: So they they uh and they sold that business as a decent exit uh to um uh an imaging chain in the US uh about a year ago.
• 44:34
  SPEAKER_00: Um but so that's one thing that is like people don't ask that question enough in the UK.
• 44:39
  SPEAKER_01: What's your right to win in this area?
• 44:42
  SPEAKER_01: Yeah.
• 44:42
  SPEAKER_01: Um and you need like we often get asked, oh, are we going to pivot into commercial forestry as long as everything else?
• 44:49
  SPEAKER_01: I'm like, we don't have a right to win there at all.
• 44:52
  SPEAKER_01: No.
• 44:52
  SPEAKER_01: Like Gresham, I always say Gresham and Foresight do that wonderfully.
• 44:55
  SPEAKER_01: Yeah.
• 44:56
  SPEAKER_01: Talk to them about that.
• 44:57
  SPEAKER_01: What are we gonna do?
• 44:57
  SPEAKER_01: Try and catch up.
• 44:58
  SPEAKER_01: There's very few companies who've successfully disrupted incumbents.
• 45:03
  SPEAKER_01: Yes, you've got an octopus energy where everyone told Greg, do not touch this, there's no more room.
• 45:08
  SPEAKER_01: But they're rare, right?
• 45:10
  SPEAKER_01: You know, where people are so well placed and so much learning's been there.
• 45:13
  SPEAKER_01: Catching up as someone new to that sector, we just don't have a right to win.
• 45:16
  SPEAKER_01: Whereas when we look at taking a genuine data-led, holistic approach to acquiring, managing, developing, and optimizing landscapes, well, I've got 40 years of building into that realm.
• 45:28
  SPEAKER_01: So there we've got a chance.
• 45:30
  SPEAKER_00: Yeah, yeah, yeah, exactly.
• 45:31
  SPEAKER_00: And of, I mean, it doesn't mean, like you say, that you say, that uh somebody coming from outside a domain can win.
• 45:37
  SPEAKER_00: Like, but you your odds are are gonna be like so if you're if you're um an average individual, uh, you're probably better off using some area where you have some knowledge.
• 45:47
  SPEAKER_00: Um, but it could be the process knowledge, or it could be, oh, you you have a you have a tool that will change the Dowicks of an industry, and you might not have a lot of exposure to that industry, but you kind of know that the potential of your technology or tool can transform it because of uh and you then go validate that.
• 46:04
  SPEAKER_00: And that's what the your VC or whatever, they'll be like, oh yeah, you've proven that you can 10x the economics of this thing to the to the good.
• 46:12
  SPEAKER_00: Okay, now now you've now you've I'll give you some money to figure out whether you can get to the next stage and get some some revenue and some growth against it, or whatever.
• 46:21
  SPEAKER_00: So um so it's uh you can't always you can't have um uh there's so much random randomness and luck as well.
• 46:30
  SPEAKER_00: So you can't be hard and fast about this stuff.
• 46:32
  SPEAKER_00: Like definitely definitely not.
• 46:33
  SPEAKER_00: Look's a huge part of it.
• 46:34
  SPEAKER_00: Yeah, yeah.
• 46:34
  SPEAKER_00: Right place, right time uh accounts for a lot, I think.
• 46:37
  unknown: It does.
• 46:38
  SPEAKER_00: We we basically uh they I then got brought back to Europe to run the European operations, and then um uh at that point I'd had my second child and it was kind of like, okay, I've I've I've uh saved up enough money and um I need to go and start another company again.
• 46:52
  SPEAKER_01: Um that need, it literally felt like an itch that I want to I want to create something, I want to build something.
• 46:57
  SPEAKER_00: Yeah, yeah, yeah, definitely, definitely.
• 46:59
  SPEAKER_00: Especially when you're surrounded by uh people doing it, you're kind of like uh I mean you do see the failure rate is extremely high.
• 47:07
  SPEAKER_00: Uh so there's no illusions about uh about that.
• 47:10
  SPEAKER_00: But it's kind of like um if you don't have a hypothesis or an idea, then obviously don't do it.
• 47:15
  SPEAKER_00: But if you're like, no, no, I think there's something here, um uh, and then oftentimes reality will tell you that's a good or a bad idea or some variant of it.
• 47:24
  SPEAKER_00: And that's actually what happened with Silvera was we weren't trying to be a ratings agency, we were trying to develop uh landscape restoration projects.
• 47:32
  SPEAKER_00: Um, and then the the kind of synthesis of everything that I'd done to that point was um in uh in renewables, everything is about cost.
• 47:45
  SPEAKER_00: So it's like, how can you get power to the minimum cost per kilowatt hour um so that it can compete and the market, which is extremely it's just price, everything now a kilowatt hour from it is directly substitutable, right?
• 47:57
  SPEAKER_00: Um so you just have to compete on price.
• 47:59
  SPEAKER_00: Um so the original thinking was um you're going to need to do um carbon removal, and with climate change, you're gonna need to uh remediate land and there's opportunities around there.
• 48:14
  SPEAKER_00: So the the original hypothesis was like um use an hour and machine learning approach to discover the opportunities that had the best economics and use some of those tools to optimize the operations of those.
• 48:27
  SPEAKER_00: It's not a million miles from where you can.
• 48:28
  SPEAKER_01: I mean it's very similar, mate, isn't it?
• 48:30
  SPEAKER_00: Yeah, yeah.
• 48:30
  SPEAKER_00: Um the and and then the the interesting thing that we were trying to do is so with um the basic equation in power is like the the the cost per kilowatt hour is going to be a function of the the capex, um the cost to get the thing up and running, the operational expenses, the opex, the cost of capital, and then the yield of the of the project.
• 48:52
  SPEAKER_00: And and so we were kind of taking that lens onto um carbon removal and landscape restoration.
• 48:57
  SPEAKER_00: And we were like, you know what's really expensive?
• 48:59
  SPEAKER_00: And my my wife, um, we it was really instrumental in this because she was at the time, her family has a farm and she was doing a regenerative transition.
• 49:07
  SPEAKER_00: They'd just done an organic transition, doing regenerative transition, and we were looking at um uh places up in Scotland.
• 49:14
  SPEAKER_00: Uh in 2018-2019, you could get amazing costs per hectare.
• 49:19
  SPEAKER_00: 900 pounds an acre, something like that.
• 49:21
  SPEAKER_00: Oh my god, it was it was so good.
• 49:23
  SPEAKER_00: Uh, those opportunities have gone, I think, now.
• 49:26
  SPEAKER_00: Sorry about that.
• 49:26
  SPEAKER_00: Yeah, I I mean, but we I kind of wish I'd I'd kind of got some money together to do that at the time.
• 49:31
  SPEAKER_00: But the um the the challenge of this model is incredibly capital intensive.
• 49:35
  SPEAKER_00: So capital intensive.
• 49:36
  SPEAKER_00: So and then so so partly that we were like, well, we don't have that money, but also uh what is the product, what is the main cost there?
• 49:42
  SPEAKER_00: It's the cost of the land.
• 49:43
  SPEAKER_00: So so we're like, well, where is the land that costs nothing?
• 49:46
  SPEAKER_00: And it's basically where it's uh unproductive.
• 49:49
  SPEAKER_00: So the original thesis was like um climate change is making some green uh productive areas uh desertify, but it's shifting around rainfall.
• 49:58
  SPEAKER_00: So there's some desert areas that will re-green.
• 50:01
  SPEAKER_00: Uh it's it's quicker to kill a green area and make it desert than it is to regenerate because of the soil structure and the uh and the vegetation like um takes longer to restore and having a seed bank.
• 50:12
  SPEAKER_00: If it's desert, there won't be a seed bank, all this sort of stuff.
• 50:15
  SPEAKER_00: Um, so we were kind of like um uh but so that will take maybe a thousand years to re-green.
• 50:22
  SPEAKER_00: Um, but if you can get that to 50 or 100 years, uh you can get um a very high quality carbon removal credit, as long as you can get it to sustainable, um, you don't need to, you can remove the human interventions.
• 50:36
  SPEAKER_00: Um uh you can uh increase the cost of the land uh or the value of the land because some some of it would be retained as uh would be converted to productive land rather than just like a nature reserve.
• 50:47
  SPEAKER_00: Um and uh and the costs would be way, way lower.
• 50:52
  SPEAKER_00: And we we figured that we could um we could hit our um our target rate of return at a$15 a ton carbon removal price.
• 51:04
  SPEAKER_00: Uh so this is like 2019, and we were like, this is amazing.
• 51:08
  SPEAKER_00: $15 a ton is incredible, right?
• 51:09
  SPEAKER_00: For carbon removal, nature NBS carbon removal.
• 51:12
  SPEAKER_00: Um and uh we were trying to speak to potential buyers, and they were like, no, no, I can buy a uh carbon removal credit, an NBS carbon removal credit for$4.
• 51:23
  SPEAKER_00: And we're like, yeah, but look at it.
• 51:25
  SPEAKER_00: This is like some eucalyptus plantation in uh it has nothing to do with carbon removal.
• 51:31
  SPEAKER_00: They're just doing those plantations anyway, for paper and pulp and for forest products, and they're just slapping a label on it that's like a this is a carbon removal credit.
• 51:39
  SPEAKER_00: It's it's kind of basically fortunate.
• 51:41
  SPEAKER_00: Uh and they were like, but it's verified.
• 51:43
  SPEAKER_00: We were like, oh my God, like um there are people out there that could do high-quality nature restoration projects, you know, um, increasing the quality of habitats, um, water uh resilience, or you can have people um taking land, even in some cases, deforesting and then replanting a plantation and marketing it as a carbon removal credit, which is just a lie, right?
• 52:09
  SPEAKER_00: So um so we're like, this market will never scale, uh, or nobody will ever be able to do valid projects while while that persists.
• 52:17
  SPEAKER_00: So then we were like, okay, we'll take the tools that we had learned or we'd we'd been thinking about and apply them to we'll we'll supply data that allows the market to function better.
• 52:28
  SPEAKER_00: And that's how Severo started.
• 52:29
  SPEAKER_01: What was the time gap between that sort of, oh, those sort of nine months.
• 52:33
  SPEAKER_01: Nine months.
• 52:34
  SPEAKER_01: Nine months.
• 52:34
  SPEAKER_01: Kind of view it like the um the big short movie where they go and see the fact that all the mortgages on the houses just aren't working.
• 52:41
  SPEAKER_01: They go to Vegas and go, oh my God.
• 52:43
  SPEAKER_01: It feels like you've had that realization of this is not good.
• 52:46
  SPEAKER_01: And then nine months of the making that a reality.
• 52:48
  SPEAKER_00: Yeah, I mean, uh for maybe half of those are still working at uh an entrepreneur first, and we were just doing our discovery around it.
• 52:55
  SPEAKER_00: So it was really like we the first commit we had from a fund, which is speed invest, um, to invest was with the original model.
• 53:03
  SPEAKER_00: But by the time we did our pre-seed round, we'd pivoted um at that point to be like, well, if the market is that dysfunctional, there's no point developing projects into it because uh you're just gonna, it's a kind of a you just erase to the bottom.
• 53:16
  SPEAKER_00: Uh, and we're gonna you how can you get a premium price when nobody will give you a premium?
• 53:22
  SPEAKER_00: Uh like uh for for quality, right?
• 53:24
  SPEAKER_00: Yeah.
• 53:24
  SPEAKER_00: So we're like, okay, well, we'll be the tool that allows projects in general to get a premium for quality.
• 53:29
  SPEAKER_00: Uh and yeah, it didn't take long, you know, it's this kind of listen to the universe is telling you, no.
• 53:35
  SPEAKER_00: So so uh we would hope that now, because you have overlays like ratings, it is easier to get a premium for something of high quality.
• 53:43
  SPEAKER_00: Um and there's that kind of narrative that shifted.
• 53:45
  SPEAKER_00: You were definitely ahead of us in terms of the time plan.
• 53:47
  SPEAKER_01: So you founded in 2020.
• 53:50
  SPEAKER_00: Yeah, it was first of January we we incorporated, yeah.
• 53:54
  SPEAKER_01: I recognize I wrote the paper of what OC is as a concept, Christmas 2020.
• 54:00
  SPEAKER_01: Yeah.
• 54:00
  SPEAKER_01: So you were kind of a year ahead of us in terms of the thinking about the model that we've ultimately ran.
• 54:05
  SPEAKER_01: And then probably March, April 2021, I remember us scribbling down if what we viewed the world would become, what are the other businesses that we needed.
• 54:16
  SPEAKER_01: Yeah.
• 54:16
  SPEAKER_01: And like the top one on that list was a ratings agency.
• 54:19
  SPEAKER_01: So interestingly, we kind of thought of your business the other way around.
• 54:23
  SPEAKER_01: Yeah, yeah, yeah, yeah.
• 54:24
  SPEAKER_00: Um so I mean, to be honest, I we would rather have done what you're doing because it's more satisfying to have like boots on the ground.
• 54:31
  SPEAKER_00: We're kind of like the fun police.
• 54:33
  SPEAKER_00: Like people are trying to make money, uh, and we're kind of being like, but you haven't crossed your cross a T here and the I's there, and I like how to have you, you know, it's kind of it's quite boring uh in some respects.
• 54:46
  SPEAKER_00: I mean, we do go quite deep into the science, and that's that's a bit more fun.
• 54:50
  SPEAKER_00: And you know, the the market structure element of it is kind of interesting, but it is like it's a grind.
• 54:55
  SPEAKER_00: You're you're trying to create an extremely reliable, standardized opinion.
• 54:58
  SPEAKER_00: So you're manufacturing this opinion, very high level of consistency, quality control.
• 55:05
  SPEAKER_00: Now we we manufacture a rating.
• 55:08
  SPEAKER_00: And the guy that runs that, he has a manufacturing engineering background.
• 55:12
  SPEAKER_00: And he was a manufacturing engineering, uh, he's like a manufacturing consultant, and he runs the production.
• 55:18
  SPEAKER_01: Why is it so purposeful about the word manufacture with the rating and not like create or choose or select?
• 55:24
  SPEAKER_01: It felt very specific choice of word.
• 55:27
  SPEAKER_00: Uh because it has to be what you have in uh in in carbon markets is like 900 plus 900 a thousand different recipes of producing a rating.
• 55:39
  SPEAKER_00: Uh and part some of that's intrinsic because there's lots of different activities that are validly um abating emissions or removing some of them are industrial, um, but you can have like industrial abatement of some, you know, uh of methane in some production process, or you can have a of some fluorocarbon or like so much diversity that's inherent in this stuff.
• 56:00
  SPEAKER_00: Uh so what you want is standardization.
• 56:02
  SPEAKER_00: You want to make everything comparable.
• 56:04
  SPEAKER_00: Now, so so our rating is like an overlay that allows you to compare what is kind of apples and pears, otherwise.
• 56:11
  SPEAKER_00: Um if there's variability between our ratings because some analyst thinks one thing and another analyst on another people thinks another thing, like they've brought their beliefs in, then it's kind of not doing the job it's supposed to.
• 56:23
  SPEAKER_00: It's just kind of it's just a way, again, it's just adding complexity.
• 56:27
  SPEAKER_00: You're trying to simplify.
• 56:28
  SPEAKER_01: Well, at that point, you're the public sector.
• 56:30
  unknown: Yeah.
• 56:30
  SPEAKER_01: You're the public sector who say yes or no to an environmental project based on the whim of a particular officer who thinks a landscape should be one way and not another.
• 56:38
  SPEAKER_00: Exactly.
• 56:39
  SPEAKER_00: And but it what it what it turns out is, and I don't know if you see this as a slightly skeptical uh view I have, is like people actually love doing that stuff.
• 56:47
  SPEAKER_00: If you've done an environmental science degree, you think you perceive your value add as you bringing all your knowledge and create crafting this handcrafted, beautiful document.
• 57:02
  SPEAKER_00: Uh, and you perceive that as success and like you're getting five.
• 57:06
  SPEAKER_00: Uh and actually what you want to do is say, no, I need to force you to extremely, extremely consistently apply a consistent set of tests, exactly the same, every single time, all about data quality, all about consistency.
• 57:22
  SPEAKER_00: It needs to be formulated in the same way.
• 57:24
  SPEAKER_00: So otherwise, you know, you'll produce and we did this extra very, very early on.
• 57:29
  SPEAKER_00: Um we were like, okay, we'll um we'll split the teams and we'll get them to rate the same, uh we'll split the team and get them to rate projects blind.
• 57:39
  SPEAKER_00: Uh and uh this was like in this was in uh sort of maybe 15 months after our uh and uh our first funding, and we we found they were extremely inconsistent.
• 57:50
  SPEAKER_00: Uh maybe even 12 months.
• 57:52
  SPEAKER_00: It was early days for us.
• 57:53
  SPEAKER_00: Uh and at that point we were like, oh my god, just just just hold the hold the phone or stop the rig a minute.
• 58:01
  SPEAKER_00: Um we think we're creating a consistent product, and it looks like it, because the format is the same, but it's not consistent because there's too much opinion in there.
• 58:10
  SPEAKER_00: So we kind of um at that point we were like, no, no, no, no.
• 58:13
  SPEAKER_00: Uh the framework is is gotta be locked down, uh, and everybody has to treat it exactly the same way.
• 58:19
  SPEAKER_00: And then we re-rated everything.
• 58:22
  SPEAKER_00: And I personally had to ring every single customer, so very big energy company in particular.
• 58:27
  SPEAKER_00: I remember the comment conversation etched on my brain.
• 58:29
  SPEAKER_00: I'm like, I know you have a position in this project, and I know we rated it well, and I know that that was uh part of your decision making uh in basically putting it on your inventory.
• 58:41
  SPEAKER_00: Uh when we've applied this framework consistently, that is not uh we don't, we don't subscribe to the opinion that we published previously.
• 58:49
  SPEAKER_00: Uh and they're like it's literally millions of dollars that we've written down.
• 58:56
  SPEAKER_00: Uh and uh from from that day, I'm like, uh there's no point doing anything that's like a cottage industry rating opinion.
• 59:05
  SPEAKER_00: It has to be like, it has to be this manufactured product.
• 59:07
  SPEAKER_00: It has to be complete, it has to be like the Toyota Corolla.
• 59:10
  SPEAKER_00: So we we always use those like manufacturing analogies, like it has to be standardized.
• 59:16
  SPEAKER_01: My brain went, exact that makes so much sense.
• 59:19
  SPEAKER_01: So you sat there, you've recognized I'm gonna have to make some phone calls I don't want to make.
• 59:23
  SPEAKER_02: Yeah.
• 59:24
  SPEAKER_01: How do you prepare for those phone calls?
• 59:26
  SPEAKER_01: First question.
• 59:28
  SPEAKER_00: I mean, don't overthink it.
• 59:29
  SPEAKER_00: You're just gonna have to cycle up as in as in like what what actually happened?
• 59:34
  SPEAKER_00: And uh um people really appreciated it.
• 59:38
  SPEAKER_00: So so they were like, look, we know this is hard, we know you're pioneering this, we we think these opinions are useful.
• 59:45
  SPEAKER_00: And honestly, the fact that you're doing this head on, it actually uh built ability, trust.
• 59:53
  SPEAKER_00: It built trust, yeah.
• 59:54
  SPEAKER_00: Because we they they knew then that we were serious.
• 59:56
  SPEAKER_00: It's like because you can produce like a rating product and and and not really care about it and it look and say lots of the same things.
• 1:00:04
  SPEAKER_00: But when you're like uh guys, this I know this is gonna cost you literal money, but like I we I we have to tell you this today, because we know today.
• 1:00:12
  SPEAKER_00: And and so it was more like, yeah, you have to be able to the prep was like being able to explain why this happened and what we were doing differently.
• 1:00:20
  SPEAKER_00: But the the main thing is just to suck it up and and and to like face up to it and and and uh yeah, actually it in the long in actually very quickly they were like, you know, actually Selen Jenny rule, and then journey's fine.
• 1:00:32
  SPEAKER_01: That I find that really reassuring in live, because my one of my values is solve for honesty and transparency.
• 1:00:38
  SPEAKER_00: Oh, hundred percent.
• 1:00:40
  SPEAKER_01: But but you are in doing so, you have no choice or control in how that lands with the other party.
• 1:00:44
  SPEAKER_01: You can hope for the best, but you you can't know that it's gonna go that way.
• 1:00:47
  SPEAKER_01: So that I love that it did.
• 1:00:48
  SPEAKER_01: Okay, so you've got those conversations out of the way.
• 1:00:51
  SPEAKER_01: How do you then start to make sure that the process you have defined is followed in such a standardized, consistent way?
• 1:00:59
  SPEAKER_01: How do you get to that level of Toyota quality manufacture?
• 1:01:03
  SPEAKER_00: Uh I think well, you use the word continuous improvement.
• 1:01:05
  SPEAKER_00: So you don't you can't do it overnight.
• 1:01:07
  SPEAKER_00: Um, so I think where we are now is just so far ahead of where we used to be.
• 1:01:13
  SPEAKER_00: Um but the the kind of the first step was like extremely uh clearly defining all of the components, the inputs, the steps of treatment and the outputs.
• 1:01:24
  SPEAKER_00: And and then we the thing that kind of slowed us down as a business, to be honest, was you have to recognize that a um uh a forest protection project looks very different to a landfill methane project, which very recipe or the methodology, or we call it a framework, the evaluation framework you have to use has to be designed and tested.
• 1:01:46
  SPEAKER_00: So you so for us to launch a new uh say a biochar framework, we we we spin up this evaluation framework, um, we then take it to a full set of actors in the value chain, they beat it up, we iterate it, then we test it on a sample of set of projects, uh, we get it to the point where we're like, okay, we'll now launch it to the market and lock it.
• 1:02:06
  SPEAKER_00: And that lock it, that's what fascinates me.
• 1:02:09
  SPEAKER_01: How do you know it's done?
• 1:02:10
  SPEAKER_01: Because like when in every other aspect of life, you can keep tweaking, you can keep adjusting, but once you've locked it, you've got so many people reliant on that.
• 1:02:19
  SPEAKER_01: Yeah.
• 1:02:19
  SPEAKER_01: So, yeah, yeah.
• 1:02:20
  SPEAKER_01: What is the what is it, is it uh a set of criteria, or is it still a judgment call of you or someone when you know that'll do?
• 1:02:27
  SPEAKER_00: That's where uh I mean you it it's basically when there's a consensus that like this reflects the best, this reflects the state of play on the information that's available, the the consensus view on how best to treat it, and the outputs that are like make sense and are reasonable.
• 1:02:41
  SPEAKER_00: And at that point, you you know, you've got to just say, okay, that's enough.
• 1:02:45
  SPEAKER_00: We there is a revision process, but we that's a formal process.
• 1:02:48
  SPEAKER_00: So we're like, okay, what world, we're gonna change the biochar framework.
• 1:02:52
  SPEAKER_00: This accounting of this feedstock is um is gonna change it for these reasons.
• 1:02:56
  SPEAKER_00: This is the feedback that we've had, this is what its effect is gonna have.
• 1:02:59
  SPEAKER_00: Again, are you cool with that?
• 1:03:01
  SPEAKER_00: Cool.
• 1:03:01
  SPEAKER_00: Now we can lock it.
• 1:03:02
  SPEAKER_00: This is the timeline, it rolls out.
• 1:03:03
  SPEAKER_00: So it has to be very controlled.
• 1:03:05
  SPEAKER_00: Um so you know, you need to be able to version control, but you can't have like flux between v1.1 and v1.2, it can't wobble, but it can't like gradually evolve into it, it has to be like everything's v1.1.
• 1:03:17
  SPEAKER_00: And then when you identify something clearly of a set of changes, you version it into version 1.2, and actually we need to be able to, you know, show why the deltas exist and everyone be like, yeah, you know what, to be honest, that new data is available, that new paper's come out and the consensus are shifted forward.
• 1:03:34
  SPEAKER_00: We want you to reflect best practice, uh, not be locked in the past, but also we don't want it to like veer under the hood, like, you know, that's not acceptable, right?
• 1:03:42
  SPEAKER_00: You can't you can't do that.
• 1:03:44
  SPEAKER_01: That's really challenging from environmental projects in general and and almost r legislation and and regulation, because the tendency of viewing today as the right answer, and environmentalists want to lock in and stop any change.
• 1:03:58
  SPEAKER_01: But there's an arrogance to presupposing that our knowledge and awareness today is the right answer.
• 1:04:03
  SPEAKER_01: So it's definitely not.
• 1:04:04
  SPEAKER_00: It's just best.
• 1:04:05
  SPEAKER_00: Exactly.
• 1:04:06
  SPEAKER_00: Exactly.
• 1:04:07
  SPEAKER_01: So our head of operations, Dave, he comes from an engineering background, he always asks, how do we keep enough optionality that we can make sure in the future we can make it better rather than just it's one of my frustrations about national parks as a concept.
• 1:04:19
  SPEAKER_01: So national parks were never designed to do anything to do with nature.
• 1:04:24
  SPEAKER_01: They were just public access for people living in industrial cities.
• 1:04:27
  SPEAKER_01: Yeah, yeah.
• 1:04:28
  SPEAKER_01: That's what they were for.
• 1:04:29
  SPEAKER_01: Now we yell at national parks and they're not doing enough about nature.
• 1:04:32
  SPEAKER_01: That's not what they're saying.
• 1:04:33
  SPEAKER_01: No, exactly.
• 1:04:34
  SPEAKER_01: But once you declare somewhere a national park, what you say is don't change anything.
• 1:04:38
  SPEAKER_01: Yeah, yeah, yeah.
• 1:04:38
  SPEAKER_01: So even if you want to improve it for nature, it's nine impossible because we've locked it away.
• 1:04:43
  SPEAKER_01: And I just don't want us to repeat those processes.
• 1:04:45
  SPEAKER_00: Yeah, yeah.
• 1:04:46
  SPEAKER_00: Yeah, it's it's it's challenging.
• 1:04:48
  SPEAKER_00: I mean, it what if you're doing the informal like expert has a view version.
• 1:04:54
  SPEAKER_00: Um, what the nice thing is if you're operating in that world, someone is like, hey, we're doing uh direct eye capture, like and instantaneously you can get somebody provide put a report on it uh a month later.
• 1:05:06
  SPEAKER_00: Uh the problem is your consistency over time, it it's it's not there.
• 1:05:11
  SPEAKER_00: Yeah.
• 1:05:12
  SPEAKER_00: And your your your ability to update it.
• 1:05:14
  SPEAKER_00: So say a project's 20 years, running 20 years, your analyst is not going to be there 20 years.
• 1:05:19
  SPEAKER_00: Um so so you know, Joe does it, and then you know, Annal takes it over three years later, and they're like, oh Joe's, what's what was Joe thinking here?
• 1:05:27
  SPEAKER_01: Like, and they always say that, right?
• 1:05:28
  SPEAKER_01: Because everyone's opinion's right.
• 1:05:30
  SPEAKER_00: Exactly.
• 1:05:30
  SPEAKER_00: So you can't have that.
• 1:05:32
  SPEAKER_00: Doesn't it doesn't work?
• 1:05:33
  SPEAKER_00: And and so then we kind of we were really, we were kind of behind uh other people in the market in covering these new sectors because we were so rigorous about creating these um these like toolkits to evaluate them.
• 1:05:45
  SPEAKER_00: What happens and in an AI world um is two things.
• 1:05:49
  SPEAKER_00: We we have this like a software.
• 1:05:51
  SPEAKER_00: So so when we define an evaluation framework in in code, so it's kind of like a tree, a graph of all these different tests and data that go.
• 1:06:00
  SPEAKER_00: In.
• 1:06:01
  SPEAKER_00: And because we set spend so time, set so much time setting them up in code, it just turns out now it's extremely amenable to update the ratings.
• 1:06:12
  SPEAKER_00: So we actually update them.
• 1:06:13
  SPEAKER_00: Most of our peers don't really bother.
• 1:06:22
  SPEAKER_01: And you think it's the m mechanism of building and the infrastructure that's allowing you to update?
• 1:06:27
  SPEAKER_00: Yeah, because it's like you have a mold now.
• 1:06:29
  SPEAKER_00: So you put, you know, you update the new, the new information goes in, the updated view comes out, but it's entirely consistent with the old one because everything is defined so uh so rigidly.
• 1:06:41
  SPEAKER_01: There's such echoes because like so I think I I sometimes get asked, how have you been able to adopt AI so quickly across the business and use it that way?
• 1:06:49
  SPEAKER_01: And I was like, we never weren't using it.
• 1:06:52
  SPEAKER_01: Like we were using it in whatever guise we could, like our our algorithm for finding acquisition opportunities had a machine learning component as soon as we could put one in place.
• 1:07:01
  SPEAKER_01: So if you've always built it that way and you built the infrastructure to do it, it is easy to adapt.
• 1:07:05
  SPEAKER_01: It's not putting a new way of working in, it's just continuing to develop and improve on it.
• 1:07:10
  SPEAKER_00: Yeah, yeah, yeah, yeah.
• 1:07:10
  SPEAKER_00: So 100%.
• 1:07:12
  SPEAKER_00: So because uh we have these software-defined evaluation, uh, we can click connect them to our MCP server uh so that then uh a sales engineer, they get a technical question and they're like, why is this uh why is this project being scored in this way?
• 1:07:26
  SPEAKER_00: And it'll be like, well, according to this scoring rule and this data that went in, this is the treatment and this is how and the the AI will be able to answer that for them.
• 1:07:34
  SPEAKER_00: Whereas before you'd have to go to an analyst and then you know dig it out some word document they did three years ago.
• 1:07:39
  SPEAKER_01: How is that has that changed the type of person you're hiring as a result of it being almost an engineering question rather than an environmental opinion?
• 1:07:49
  SPEAKER_00: Uh no, I don't think so.
• 1:07:51
  SPEAKER_00: I I think you you know, you said um certain people adopt quicker than others, and you put and then certain people are more um are better at generating the tools, have better intuition around them.
• 1:08:03
  SPEAKER_00: So it's more just like celebrating good practice and kind of creating an environment for it rather than being like, um, I mean, we did we did shift our hiring actually at the analysts to be a little bit more quant and technical and a bit less um ecologist.
• 1:08:18
  SPEAKER_00: Uh well, it's just significantly more.
• 1:08:21
  SPEAKER_00: So there's a shift in our hiring of the at the analyst intake level so that they would be more ready for the kind of automation.
• 1:08:27
  SPEAKER_00: But it it wasn't really like we're like firing people and like hiring new ones.
• 1:08:31
  SPEAKER_01: It was I mean, business and cost-wise, that's a sensible move as well, because uh ecologists are essentially extinct now in terms of finding them, it's incredibly hard.
• 1:08:39
  SPEAKER_00: Yeah.
• 1:08:39
  SPEAKER_01: So actually, your ability to find analytical minds is going to be much easier, much larger pool in the UK, certainly.
• 1:08:45
  SPEAKER_00: Yeah, I mean the UK is amazing for talent.
• 1:08:47
  SPEAKER_00: And and um there's n I mean, the bad thing about for the UK economy is like there's not a huge number of opportunities for them to go into.
• 1:08:54
  SPEAKER_00: So I'm sure you have amazing talent pools for every open role.
• 1:08:58
  SPEAKER_01: We're averaging a thousand applicants, just just shy.
• 1:09:00
  SPEAKER_01: We're not quite at four figures, just shy of a thousand applicants for every role.
• 1:09:03
  SPEAKER_00: That sounds that sounds about right, yeah.
• 1:09:04
  SPEAKER_00: It's um that's actually slightly better.
• 1:09:06
  SPEAKER_00: That's better slightly better than what we have.
• 1:09:08
  SPEAKER_00: But it's I mean, that's what it's like, which is how glad are you that we're not in the job market now?
• 1:09:13
  SPEAKER_01: Oh my god.
• 1:09:13
  SPEAKER_01: So like we we joke, our senior team are like, we wouldn't get an internship now.
• 1:09:18
  SPEAKER_01: Like it is it is a really tough spot in terms of being a young person trying to get into the sector.
• 1:09:23
  SPEAKER_00: Oh yeah, 100%.
• 1:09:24
  SPEAKER_00: 100%.
• 1:09:25
  SPEAKER_00: We just need to get, you know, bu build our companies to 20 times a second.
• 1:09:28
  SPEAKER_01: So what we're doing, why are we not trying harder, mate?
• 1:09:31
  SPEAKER_01: Um, in fact, on the way past that one, um, since we both sat here, what what should young people do if they do want to get into our respective businesses to stand out?
• 1:09:40
  SPEAKER_01: What what should they do?
• 1:09:42
  SPEAKER_01: How should they best position themselves for opportunities you do have?
• 1:09:47
  SPEAKER_01: Uh I'm gonna let you go first, then I'll give an answer.
• 1:09:50
  SPEAKER_00: That's a good one.
• 1:09:51
  SPEAKER_00: Like w one thing that I um it's really about what you can do and not about your CV.
• 1:09:58
  SPEAKER_00: Like the CV is just a hygiene, and then you being useful and thoughtful about what you do is the differentiator, I think.
• 1:10:05
  SPEAKER_00: So um literally every week or two, I'll get a student or an MBA student or like a science person or engineer pick ping me, uh, being like, this is me, this is how great I am, I want a job.
• 1:10:18
  SPEAKER_00: And I'm uh most of the time I just can't I just can't even be doing with it.
• 1:10:22
  SPEAKER_00: But so some when they show some some glimmer there, I will bit, I'll reply and be like, cool.
• 1:10:29
  SPEAKER_00: From your outside in knowledge of the business, what your skills are, what you you're interested in, tell me what you would do, what you think we should do and how you should um how you'd implement that.
• 1:10:39
  SPEAKER_00: Uh and I've hired every single person that's given me a good answer to that question, which is two.
• 1:10:46
  SPEAKER_00: Uh, because most people are like, I will do sales for you, I will do strategy.
• 1:10:50
  SPEAKER_00: And you're just like, okay, thanks, guys.
• 1:10:52
  SPEAKER_00: Uh like really, really it's right.
• 1:10:54
  SPEAKER_00: No, I need something specific because you we operate on day-to-day specific.
• 1:10:58
  SPEAKER_00: So it'd be like, you know, I've seen that you're uh you're um you're a bit behind on this, or you have this thing, like this is how I would solve that problem for you.
• 1:11:06
  SPEAKER_00: And I'm just like, you're in.
• 1:11:07
  SPEAKER_00: You can just skip to the skip to final uh because it's so rare for people to think in that way.
• 1:11:13
  SPEAKER_00: So rare.
• 1:11:14
  SPEAKER_00: So so so that's what differentiates the you know, the the the the seven hundred thousand two hundredth um Russell group grad.
• 1:11:22
  SPEAKER_00: Yeah.
• 1:11:22
  SPEAKER_00: It's the one who's like, I think I think you guys should do this, and this is how it helps you.
• 1:11:26
  SPEAKER_00: And we're like, oh my God, this is like this is like unicorn.
• 1:11:29
  SPEAKER_01: Um well, and on that basis, they're arriving as a peer already because they're saying I I can help, I want to be part of the process.
• 1:11:37
  SPEAKER_01: My my answer went to um I'm gonna steal that, by the way, in terms of an approach, so thank you.
• 1:11:43
  SPEAKER_01: Um because you just I get sick of saying no.
• 1:11:45
  SPEAKER_01: You just feel like such an ogre, don't you?
• 1:11:47
  SPEAKER_01: Like, no, I'm sorry, we haven't got you, there's no room, there's no room at the inn.
• 1:11:50
  SPEAKER_01: It's like um giving them something more tangible to respond to.
• 1:11:53
  SPEAKER_01: We did have a couple this week where people had gone that extra mile and like, you totally deserve a conversation with our people team as a result of that.
• 1:12:00
  SPEAKER_01: Um when it comes to our intern process, which is the most in-demand area we get, we had to shut it after two weeks because we had 2,000 applicants.
• 1:12:07
  SPEAKER_01: Yeah.
• 1:12:07
  SPEAKER_01: And we were like, we were gonna run it for four, and we were like, no, we can we just can't pro Although now we've built a really cool uh we built our own recruitment platform, so to um really speed through things, which is which is helping, and built in like a video component of it, and it's it's really cool.
• 1:12:21
  SPEAKER_01: Shout out to Karis for doing that.
• 1:12:22
  SPEAKER_01: Um, it is it is that standout as being uniquely you and different.
• 1:12:26
  SPEAKER_01: Don't give me anything generic.
• 1:12:28
  SPEAKER_01: Like when I do look at a CB, I look bottom up.
• 1:12:30
  SPEAKER_01: What is the adventure?
• 1:12:31
  SPEAKER_01: What is the achievement?
• 1:12:32
  SPEAKER_01: If if you've if you've played international sport or performance level music or something that way, uh and you said about the hygiene thing, they've all everyone's got the qualifications you need now.
• 1:12:43
  SPEAKER_01: But if you've got that other thing as well, where you've shown you've been adventurous, you've suffered for something, you've worked hard, because it's hard.
• 1:12:49
  SPEAKER_01: Like working in startup and scale-up businesses, we expect a lot.
• 1:12:53
  SPEAKER_01: Hopefully, we give a lot in return.
• 1:12:55
  SPEAKER_01: But I want to know that if you want to go through hard work, you've you know what it takes to get there.
• 1:12:59
  SPEAKER_01: Yeah, yeah, yeah.
• 1:13:00
  SPEAKER_01: So I think that's what stands out to me.
• 1:13:02
  SPEAKER_01: Definitely.
• 1:13:03
  SPEAKER_01: Um, jumping back to you guys, when was when was the time that you you saw the value of a rating to a to a product?
• 1:13:11
  SPEAKER_01: When were you able to literally say that rating is now showing as that premium?
• 1:13:15
  SPEAKER_01: People can know the difference between a good and a bad product.
• 1:13:17
  SPEAKER_00: Honestly, we were looking for this for two years and we didn't have so for for maybe a year and a half, we were like, if the price doesn't change from the rating within, you know, give it, you know, to our next round or something, then we should just go home.
• 1:13:34
  SPEAKER_00: Because like the the thing that we were trying to solve for, as in rewarding, economically rewarding goodness for the climate, uh, like if that relationship wasn't established, uh, we should just go home.
• 1:13:45
  SPEAKER_00: Uh and um uh for an uncomfortable amount of time, we couldn't see that.
• 1:13:52
  SPEAKER_00: And then you started to see it, partly we didn't see it have a lot of pricing data either.
• 1:13:56
  SPEAKER_00: Um, so your your statistical power was weak.
• 1:13:59
  SPEAKER_00: And then we started to see it anecdotally.
• 1:14:02
  SPEAKER_00: And then uh really in the last, I'd say 18 months is only it's only 18 months where you've had a very strong statistically powerful relationship between a rating and surprise.
• 1:14:12
  SPEAKER_00: We we we we didn't come out saying it sooner than that because we were like, I mean if you squeeze your eyes uh and like you can kind of fit a line that looks like this.
• 1:14:22
  SPEAKER_01: Uh but no, you guys can't, can you?
• 1:14:24
  SPEAKER_01: Because because you are the arbiters of fact and detail.
• 1:14:28
  SPEAKER_00: Well, I mean it's easy to say stuff and and and people some people it some people do.
• 1:14:33
  SPEAKER_00: But the uh yeah, you uh our our our our bar was it needs to be statistically significant, not like there's a positive slope.
• 1:14:44
  SPEAKER_00: Uh there actually has to be significance behind it.
• 1:14:46
  SPEAKER_00: But but yeah, it was it's it's only really in the last year and a half where you've seen it.
• 1:14:49
  SPEAKER_00: Uh and I think part of it is um the drama the market had about quality uh 2023.
• 1:14:57
  SPEAKER_00: Um that kind of that depressed the prices that known branded but weak projects were getting because it really projects are trading on the brand, like Kariba.
• 1:15:08
  SPEAKER_00: We'd rated it poorly forever, and the price hadn't really reacted.
• 1:15:11
  SPEAKER_00: But everybody um we all we came out with, we had it rated for um at C grade, I think it was the second lowest grade for um for two and a half years before it whole collapsed and the scandal came out about it.
• 1:15:24
  SPEAKER_00: And and we'd had a lot of engagement with the developer about it and saying, guys, you just need to fix this stuff.
• 1:15:30
  SPEAKER_00: And um uh the price hadn't moved at all.
• 1:15:34
  SPEAKER_00: Uh so then when people were like, oh, well, it turns out that some of the opinions are valid.
• 1:15:39
  SPEAKER_01: People just weren't engaged with it.
• 1:15:40
  SPEAKER_01: They, you know, they put that there, we've gone with that project, fine, we're happy, done, leave it alone.
• 1:15:44
  SPEAKER_00: I mean, this comes to the heart, frankly, I think, of the problem behind carbon markets and and nature markets more generally.
• 1:15:52
  SPEAKER_00: The credit has to have some value.
• 1:15:54
  SPEAKER_00: And this is where BNG, the biodiversity net gain, like is good, because the the credit has a meaning in some system where there's an economic value associated with it.
• 1:16:03
  SPEAKER_00: Like I want to build a house, and if I want to sell the house, I've got to buy the credits.
• 1:16:07
  SPEAKER_00: It's like a do not pass go until you have this thing.
• 1:16:10
  SPEAKER_00: Um, now there's my BNG isn't perfect by any means, and you're way closer to it than we are.
• 1:16:16
  SPEAKER_00: But um uh the problem with carbon credits is there was storytelling credits, and they the the thing you bought was the power to tell a story about how you were doing the right thing, uh, potentially as a business.
• 1:16:30
  SPEAKER_00: Um, nothing to do with actually uh like what's the economic value to me.
• 1:16:36
  SPEAKER_00: You know, I there's some there's some compliance obligation, there's a tax rebate that has some where the quality of the credit has some bearing on that.
• 1:16:45
  SPEAKER_00: It was like is it was like um I'll use a credit with a brand name, um, even though it might be a C grade, because what I'm trying to do is tell stories.
• 1:16:55
  SPEAKER_00: And it's easier to use a branded name story project to tell a story than an unbranded project.
• 1:16:59
  SPEAKER_00: Uh so the name was kind of the rating before.
• 1:17:01
  SPEAKER_00: Uh, and then people were like, oh, it turns out I actually am going to get called on it.
• 1:17:07
  SPEAKER_00: Uh so it was, although it's been it's depressed demand in the market, um it people now are spending money disproportionately on credits to projects where there is a climate outcome or where the the claim is aligned with um reality.
• 1:17:23
  SPEAKER_00: Uh and that's a good thing.
• 1:17:24
  SPEAKER_00: Uh, but the problem is um there is still no scalable um driver for people to purchase um credits like to justify the expense of it.
• 1:17:37
  SPEAKER_00: So so we actually don't project really hardly any really volume growth in voluntary markets, maybe a bit coming towards 2030.
• 1:17:46
  SPEAKER_00: Um, a bit of growth in the value because people are spending more on like real credits.
• 1:17:51
  SPEAKER_00: But why would, you know, and this came out with Microsoft recently, what why would why would they spend$10 billion?
• 1:17:56
  SPEAKER_00: Like it make it just depresses their margin.
• 1:17:59
  SPEAKER_00: There's no there's no economic benefit for them.
• 1:18:03
  SPEAKER_00: Makes no sense.
• 1:18:04
  SPEAKER_00: It's it's not economically rational for them to do that morally, but like you get a new CEO coming in, you have a bit of a downturn, your margins go down, it's the first thing to go because it's really a marketing spend.
• 1:18:13
  SPEAKER_00: So so um I this is I think the the crux is like the quality problem is kind of solved.
• 1:18:19
  SPEAKER_00: You you can ascertain quality, um and and but you need some utility for the credit.
• 1:18:25
  SPEAKER_00: So so um using the credit to lower your emissions trading scheme allowance consumption, well, that would give value to it, and then you'd see growth uh and you see more money go onto the ground scalably.
• 1:18:40
  SPEAKER_00: Uh and I think that that's like the you know, you still have panels at conferences, people talking about quality and integrity.
• 1:18:46
  SPEAKER_00: It's like, no, no, it's it's demand.
• 1:18:49
  SPEAKER_00: And frankly, like um until you uh solve for that, then there's no growth materially, like no material growth.
• 1:18:58
  SPEAKER_00: Um and the big unsaid thing is like uh for the EU, the UK to do that, the Treasury has to accept that it will get paid less direct fees from the emission trading scheme.
• 1:19:13
  SPEAKER_00: So you're basically saying you have roughly four and a half billion of um revenue that the polluters pay for in the UK.
• 1:19:20
  SPEAKER_00: If you want to allow them to buy credits from like from you guys' uh projects in Scotland or in Devon or something, um the payment isn't going to go directly to the Treasury.
• 1:19:31
  SPEAKER_00: It's gonna create jobs and growth and opportunity and VAT and uh income tax and uh national insurance contributions and infrastructure and all of the good benefits and positive environmental impact.
• 1:19:43
  SPEAKER_00: And all the environmental impacts and like the insurance and the uh regional growth.
• 1:19:49
  SPEAKER_00: Uh, but the Treasury is that number will go down, the direct number will go down.
• 1:19:54
  SPEAKER_00: And and the the kind of that's the challenge.
• 1:19:56
  SPEAKER_00: Like that's really the conversation.
• 1:19:58
  SPEAKER_00: They there needs to, they need to accept that they will get a lower first order payment of ETS and they'll get payments other way.
• 1:20:07
  SPEAKER_00: And and until that gets solved for, it's kind of like, and there's no point really talking about integrity anymore.
• 1:20:12
  SPEAKER_00: That is like if you care about integrity, all the data is there.
• 1:20:15
  SPEAKER_00: Yeah, all the tools are there, all the technology is there.
• 1:20:17
  SPEAKER_00: You just need to ask, you haven't looked, you haven't tried.
• 1:20:20
  SPEAKER_00: Yeah, you just need a reason to pay.
• 1:20:21
  SPEAKER_01: Yeah.
• 1:20:22
  SPEAKER_00: Uh uh, and uh until that gets solved.
• 1:20:25
  SPEAKER_00: And you are seeing those happen in different markets, not in the UK.
• 1:20:29
  SPEAKER_00: Europe is extremely like PTSD'd from the the use of uh the C C DM uh credits, which have lots of quality issues, and that and they're gonna what they're gonna do with CRCF is um has its own challenges.
• 1:20:44
  SPEAKER_00: Uh, but but at least they're gonna create an avenue to um to use projects in the EU uh in substitution for a um emissions allowance.
• 1:20:54
  SPEAKER_00: But again, they are stealing from the tax revenues of different countries.
• 1:20:58
  SPEAKER_01: The beauty of what you just illustrated, though, is so I don't know if you get this narrative, but often we'd get, oh, carbon credits are just a license to pollute.
• 1:21:07
  SPEAKER_01: And and I remember having a really challenging conversation with a guy who threw that.
• 1:21:10
  SPEAKER_01: And I was like, what are you what are you talking about?
• 1:21:13
  SPEAKER_01: Of course of course they're not.
• 1:21:15
  SPEAKER_01: Like I can't produce a carbon credit unless we've planted a tree, we've restored a peatland.
• 1:21:20
  SPEAKER_01: You know, it takes years to do this.
• 1:21:21
  SPEAKER_01: It's not an easy thing to do.
• 1:21:23
  SPEAKER_01: What do you mean a license to pollute?
• 1:21:24
  SPEAKER_01: And and ultimately I had a moment of realization he meant a regulated carbon credit.
• 1:21:29
  SPEAKER_00: It didn't mean a voluntary, yeah.
• 1:21:31
  SPEAKER_01: And then I was like, oh God, he's right.
• 1:21:34
  SPEAKER_01: When it just goes to the Treasury, there is no requirement or no commitment for that to deliver any environmental or social impact.
• 1:21:41
  SPEAKER_01: None just drops into taxation.
• 1:21:42
  SPEAKER_01: 100%.
• 1:21:43
  SPEAKER_01: So actually, we can solve two issues in one go.
• 1:21:46
  SPEAKER_01: We can solve the demand rate, and we can actually put to bed forever the fact that you are paying to pollute.
• 1:21:51
  SPEAKER_01: Because if you if you can only buy uh uh what is now currently a voluntary carbon credit, or one that's nature-based or is verified in in that way, then you can't have a license to pollute anymore.
• 1:22:01
  SPEAKER_01: You have to buy a service that is benefiting people and wildlife and the planet.
• 1:22:05
  SPEAKER_00: Yeah, 100%.
• 1:22:06
  SPEAKER_00: 100%.
• 1:22:07
  SPEAKER_00: But that that is um that's so often missed.
• 1:22:10
  SPEAKER_00: Um, but really that's the that's the that's how you unlock the the next 10 billion of demand reliably that you know some CEO can't just turn it off or off or on.
• 1:22:20
  SPEAKER_00: Um it's just like, oh yeah, well, I just am compelled to pay some compensation, and that is going to create real-world activity, whether it's in industrial emissions abatement or whether it's in nature restoration or protection.
• 1:22:34
  SPEAKER_00: Um, it's actually connecting the money to do something real.
• 1:22:37
  SPEAKER_00: Uh radical idea.
• 1:22:39
  SPEAKER_01: We need to get someone from the treasury on.
• 1:22:40
  SPEAKER_01: Um, Al, you've you've got it, Ron.
• 1:22:43
  SPEAKER_01: You've been really generous with your time.
• 1:22:44
  SPEAKER_01: We'll we'll call that part one and we'll bring you down to Devon for part two.
• 1:22:47
  SPEAKER_00: So yeah, we should do we could do the uh the LiDAR stuff.
• 1:22:50
  SPEAKER_01: Uh yeah.
• 1:22:51
  SPEAKER_01: That would be cool.
• 1:22:51
  SPEAKER_01: Mate, thank you so much for the time.
• 1:22:52
  SPEAKER_01: I really appreciate it.
• 1:22:53
  SPEAKER_00: Amazing.
• 1:22:54
  SPEAKER_00: Thanks for having me.