Originally Broadcast: July 12, 2025
Spexi
Unknown: Welcome everybody to the decentralized tech live stream. I am Jon Radoff. I am the CEO and co founder of a company called Bemable. We are a decentralized physical infrastructure network.
Jon Radoff: The company is a network infrastructure network deep in project for the game industry where game servers back end live services, all the things you need to build a game. And today I'm talking about decentralized tech again this time with Graham from a really cool company called Spexy. They are also a deep in company doing something absolutely completely different with drones and it's super cool. I can't wait for you to learn about this. You know when I was with the founders of Spexy at I think it was it wasn't Dubai. Yeah I wasn't taking 49 in Dubai we met each other. The cool thing I thought about it is when I looked at the interface to Spexy it had all these cool cigars and I was like that is like a that's like an old school like map faced game. And I saw like an almost gamified aspect to using this but I don't want to jump the gun too much. We're going to learn about Spexy and what's cool about it. And everybody here by the way like if you're curious about deep in if you're curious about drones because we're going to be talking about that. Anything around decentralized tech. The reason we do this live is so that you get to ask your questions so please ask those questions put them in chat and we bring those right on here. We had so many questions. And ask you feedback going on. Just 1000 today apologies. We got a few technical difficulties but that's that's live TV and that's the internet and on top of it we're on X which is not always the most stable platform. So we're do we're taking all the risks and we're talking about web 3. So we're doing everything that's cool and different and right on not the leading edge the bleeding edge because if you aren't doing this and you're not you don't have a few wounds. Then I will be doing in web 3. Let's take a step back here Graham let's meet you. What have you done before you got to this point in time where you were working on Spexy like what got you excited about web 3 drones all of this tech and then we'll we'll dive into Spexy after people get to know you a bit.
Guest: Yeah, absolutely. So I'm Graham. Hop's one called now the ops manager at Spexy. Now one of the core founders but it was a little bit of a match made in heaven sort of for my my hobbies and interests along with my career and so I come at it from two different paths 2017 I found through a through a friend through a random Facebook post.
Unknown: I found web 3 crypto currency as it was called more more locally at the time.
Guest: Literally a Facebook post from a high school friend saying he had quit his job and bought a van and it was going to be driving across Canada and if you knew anyone on his route that he should they should drop a comment and he would stop it and say hi that's the this is my story into into web 3 crypto deep in blockchain all space. And I was living in Toronto area at the time. He had quit his job because he had cashed out after a Bitcoin run and he had made some money in fairly early crypto and he was going on an adventure with a with an old Chevy camper van and I dropped a comment and said hey if you're if you're coming through the GTA stop in and say hi and so a couple weeks later he did just that. And at the time we started talking he you know he did a tech head and we started talking yeah he actually ran a kind of podcast video podcast similar to this for low alcohol and needs and talking about web 3 back this back 2012 2013 2014 time. I don't know the exact dates honestly. And as he stopped in and we started talking about different sort of application ideas and things like that I had a small business at the time and and but he was obviously you know big into crypto and he started sharing that a little bit of that with me he continued on his trip after a couple of days stand with us and then we had this idea percolating and I had a basement sweet he said well if it's all right with you could I come back and stay in your basement sweet and will sort of dabble in this idea. Yeah sure yeah that sounds great and when he came back he offered to pay a portion of the rent in in Bitcoin right and I knew nothing about it but I was like sure this worked out well for one of you and not the other. Yeah we were still very good for as well. I said yeah you can you can pay a little bit of the rent in this cryptocurrency and teach me about it and whatever so got my first wall it set up and you know this is 2017 it's Bitcoin's going from. I missed Bitcoin I thought because it had gone from 22 to $2800 coin and so that's that's out of my price range is too expensive maybe I can catch a theory I'm right at time 180 or 200 bucks and so you know he spends the time we did start working on this little project but then the sort of the whole ICO craze happened and we both started doing that we never really got our project off the ground we just started you know. Doing all sorts of of cryptocurrency and Bitcoin stuff and so that was my my introduction into that and wrote some waves up and down and a couple of years past I actually moved to South America for a bit with my wife from where I met a different friend new friend who had a DJ I spark a small drone and you know we were off he showed me how to fly on the and I thought I was just the coolest thing I had ever seen I had no idea that the. Advances drone technologies have made sort of from 2013 14 when they were still very unstable and it was like a GoPro mounted on a on a home built type of thing to like 2018 where you had these consumer units you could buy off the shelf that were very stable and and control with phone and I just went wow this is this incredible and again I said to him like we should start some sort of little business training or education you can see my my entrepreneurial like ideas. You know type of personality coming out in the story because everywhere I see a cool technology I jump into it like how do you do that is that work so you know we started looking you know yeah. Whatever this the space had for drones at the time and I still had kind of cryptocurrency going as a hobby I was trying to do some you know arbitrage trading with Bitcoin and things like that mostly after the sort of 2018 crash I had let that be a side bar and I was doing other stuff. And so I got into drones and at the end of that year is moving back to to Vancouver and so I started applying to some jobs at drone companies and I got a job as a business development lead at a small drone business you guy ran it seven rate drones and he would do construction imaging and real estate photos and videos and he was just getting into to mapping and LiDAR and stuff like that sort of fire and survey work with drones but he didn't really know the space and I neither did I but I came out.
Unknown: This is sales and business development back on it so I'll help you all learn all about it all help you grow the space and it's really cool I think it's got a lot in front of it so I did joined him did spent kind of for six months just like crash course on like GIS and drones and imaging and ortho imagery and and mapping and 3D point clouds and all the things that you could do.
Guest: I just sort of taught myself and called everybody that I knew that might have a you know an inkling about what the space could offer or what you could do with the drone and and we were building that business up for him there was kind of four or five of us with different you know a lot of hats and flew a lot of drones and took on a lot of pilot projects so it was like while we think a drone would be able to do this so let's let's try it out and sound some different companies and organizations that were that we're willing to do that with us and take those those journeys with us. And that small company was acquired in 2021 by a larger sort of drone firm was it was doing through mergers and acquisitions was was building a kind of a drone conglomerate across North America so moved into a BD role with that bigger firm and around 2022 I saw The announcement for the city the seed round fundraise for what specs he was doing and I had since in 2020 2021 gotten back into crypto with DeFi summer and within fts and with all these things so I have this thing might you know I go to work every day and I sell drones and I talk drones and I'm trying to make a drone you know business work and as soon as I shut off work every day. I'm in crypto and I'm in deep in DeFi and I'm on the finance market and I'm trading in fts and I see this this announcement out of Vancouver where I'm living Vancouver Canada that specsy has just raised five and a half million dollars to create a decentralized deep in drone network and speculative time is like 10 or 12 employees they have sort of a legacy drone software platform but they're trying to figure out how to scale this thing globally and create standardized imagery and I just went.
Jon Radoff: Like that's everything that I do but that's that's John you're muted then you're doing crypto you were doing drones and spexy put it all together and you met and you met some of these folks originally or some of the things you were doing.
Guest: I knew Alec and Bill who are two of the founders of spexy I knew them digitally could make a drone space was very small spexy was another drone company you know that drone software company. Alec also had a training business who was supposed to join us today I'm not sure it'll be able to be my drop in in a little bit. I literally was a LinkedIn message I said Alec I said we have to talk I said Alec a LinkedIn message because we were LinkedIn connections and I said he kind of went you know okay why we knew we had friends of friends and new people in the industry. He said okay well let's you know give me a call is it there's does it have full people in the world that no like I look at what you're doing and I get it I get it from start to finish I understand the complexities of the drone space and of operations and the difficulty and regulations and of getting pilots and trying to collect all this data but I also understand cryptocurrency and NFTs and deep in and why you might want to do this in a decentralized way and what the block chain can bring to your you know your idea to the business idea. So that was the like I say I'm not making heaven a little bit like you don't very often see. That wasn't even a job I just said you have to hire me because because I have to work with you guys.
Jon Radoff: We have to talk about spexy and what you do now and by the way again I want to repeat the message out to our audience here so we've got over 600 people that have tuned in live so this is your opportunity to ask questions about drones about crypto about a really interesting deep in project that brings those two things. Together that you're about to learn about so definitely get those questions ready ready I see one from third and a major coming in already will get your question after the intro that we're going to go through here but also I was inspired by this road trip intersection that you had in in Canada I'm going to be travel it I'm going to be doing a Canadian road trip for the first couple weeks of August with my.
Guest: So so maybe I'll meet a wayward crypto person in that process that'll that'll change my whole future trajectory it honestly did change the trajectory of my life just this just one Facebook comment comment on a Facebook post really I always go.
Unknown: That's better butterfly effect right there okay Graham let's talk about spexy explain what you guys do.
Guest: Yeah so we are developing have developed and continue to grow a deep in for lack of a better sort of niche or space decentralized network of drone pilots we develop the software and where a marketplace software so much like an Airbnb would match a guest to a host or doober would match a rider to a driver. We sort of match a drone pilot to a collection task except we we do it at this broad scale so we using a what's called an h3 grid which was actually developed by uber something might be familiar and you just what you noticed in the application these hexagons that lay over a map we can. Add a drone automated drone mission to these hexes that grid out over the entire area of interest geography call it you know city of might we started in Vancouver city of Vancouver so the whole city we layout these hexagons are about 25 acres and in them we load our software loads automated flight mission.
Unknown: And then any drone pilot using the specific drones that we integrate with so anybody we don't have to know them like an uber driving you know a little bit about them but anyone can sign up to the platform take their their off the shelf consumer grade drone approach that mission zone collect the data pans off like the flight software automates the flight and data collection. And then when they upload it they get a reward right there reward so it is very much gamified and that is lots of opportunities for the collection portion being a game you know like a like a Pokemon go right you go in the world you move around you take your drone you capture some images you share those images back with our network and you're rewarded for it you know currently in cash but moving towards this this blockchain and this token generation and.
Jon Radoff: And maybe it's worth kind of talking about that business model for moments of people are familiar with I think for the web three people who are on this talk which is probably just about everybody given the topic of decentralized technology a lot of people have been exposed to play to earn walk to earn run to earn like all these to earn things yeah the fundamental issue with a lot of these things is where is the sustainability come from from play to earn kind of has played out as almost a ponsinomic. So that's the kind of structure and a lot of games that said it could work it's not just say that it can't work in the right kind of games like if you look at even online which is not a web three game like they probably could build that kind of economy because there's so much like work that actually happens that fuels this massive economy that they've got there so in a big enough economy I think played or earn could actually work and in fact the founders of even line are trying to build a game that I think is going to do that. But then like walk to earn like but you guys have a business model like there's a commercial market for this data that you're collecting I think is what it comes down to could you just maybe talk about like what that market is what they need how your project kind of fills that.
Guest: Yeah I mean these absolutely and I totally agree I mentioned before we went live that I still have some axes rotting in a wallet somewhere. You know I haven't opened in a couple of years and you know paid way too much for at the time when that was the big rage and so I know you know I went through the space and again this is part of the value that I think that I bring to the speccy team who some of whom are more crypto sort of need of than others but it's really this understanding of the space and that you know as much as it is magical internet money it's not magical internet money you need a business model you need money flowing in and money. It has to be a real economy and so that's one of the things that I really liked about what I saw with speccy is this is a multi you know hundreds of millions of not billions of dollars business already right there's already an industry around earth imaging right geospatial imaging aerial observation from satellites airplanes and helicopters and the challenge the thing that isn't nobody has solved for yet and something that I learned in the drone space is very difficult to solve for is. And realizing a layer of imagery at the drone level so the easiest way that people will understand is when you open Google Earth or Google Maps and you see that world in the middle of it and you zoom in into the world you start in I mean you actually start when you're way zoomed out it's like a rendering it's not even a real picture and then you get to a satellite level and so you can move around the earth and you can see the whole world that has been captured by satellites and Google has aggregated all that and rendered the earth through a bunch of satellite imagery from. Their own satellites and imagery that they purchased from other other organizations and as you continue to zoom in you'll actually see that you go to airplane imagery at a certain level you go to this this imagery captured by airplanes and if you notice you go to a city if you go to Los Angeles you're going to get really pretty good resolution but if you go out into the mountains right near Los Angeles you're going to get much worse resolution and that's because airplanes don't fly nearly as much nearly as high as a resolution out there because there's less value but all of that is a huge industry and economy where you can see the world is going to be a real big deal of the world. Google is buyer number one but then Google selling that through APIs to all sorts of entities and there's a whole bunch of other buyers Google's not the only player in that space right these are big players that are buying to a spatial data earth imagery data for a whole host of of reasons from you know planning and emergency management to city city planning and infrastructure mapping and management to you know well you can't even imagine how many industries use imagery data of the earth. They just need to know where something is where they want to know where something is and as you go below you know the best resolution that airplane scrap is maybe like to set seven to 10 centimeters companies you know ultimately million dollar companies. Vexel and and ego mapping and you know this whole bunch of organizations that move airplanes around the country and around the world and they fly these these imagery flights and they collect this imagery and they sell it to Google and they sell it to cities and they sell it to these different organizations that they buy industry. And that's the best imagery that exists in the marketplace right now is this 10 to 10 centimeter imagery that costs tens of thousands to hundreds of thousands of dollars to collect and demand aircraft to show up in your city or maybe it's from near your city and it flies the script that over your city and captures much of images and they put it all together and they give you this map and contrast that with the resolution you can get you with a drone. Yeah, you could get to we fly at three centimeters so a little twice as fine you can go lower with the drone we we fly at an altitude that doesn't allow for you run into privacy issues essentially at some point you run into like you're peeking in someone's window right so so you can fly as low as you want with the drone. But we do this balance where we fly at 80 meters we get about a three centimeter resolution which isn't it's actually dumbed down enough to not be able to recognize a base run facial recognition not be able to read a license plate or a home address or things like that because that sort of abstracts us from some privacy entanglements like Google's needing to blur people and needing to do certain things like that so you can get really good imagery of the drone. What's hard to get with a drone because of limitations in technology and battery life and regulations and a whole bunch of things was hard to get is the same imagery over a big scale right because I might own a drone with it with a you know a one centimeter camera that captures at 48 megapixels and you might own a drone with a three quarter inch sensor that captures at 24 megapixels an Oscar has a different drone so. So all of that aggregated data doesn't doesn't stick together to talk to each other well so what's vexie is solving for is that standardization of imagery across a broad scale so now you can look at a whole city and all of the data looks the same. All captured by sensors that are the same or similar it's all aggregated and now you have this smooth seamless layer of data over over an entire city that a buyer like a Google or like an apple maps might be interested in purchasing because previously they didn't have a way to buy from all these small local operators and aggregate that data in the way that made any sense financial.
Jon Radoff: So we've got a lot of questions coming in about fly to earn as as the model so as the drone pilot can you maybe explain how that works for them how do they get paid what do they get paid for.
Guest: Yeah so actually currently we're one of the few deep bins and we're we're racing towards a TGE here at Co-Congeneration event but we currently paid cash so we're North America only right now Canada and USA when we we essentially launch an area of interest we lay this. This grid map over an area of interest local pilots or in some cases these pilots have started traveling around and sort of chasing our launches. They the missions open they enter the zone they become live and they haven't associated cash reward to them so it usually to $12 per mission which takes about seven to 10 minutes to fly a mission so active pilots on the network are. Are earning like 40 to $60 an hour while they're while they're flying the toe. In count and that's to strike like a digital. Like the digital payment services provider is integrated into our app currently we are you know having announced a date yet but before the end of the year looking to do it token generation event at which point the cash payments will will disappear will move to a blockchain payment rail token based payment rail and that will allow you for. Among other things easier expansion and international expansion and facilitate some large.
Jon Radoff: So I'm really curious about like where the token where the whole web three aspect fits in will come back to that that another question coming in from Nancy here is what is the spec of the drone for this and I kind of have almost an associated question so not only what is the spec of the drone what what are your what is my understanding from our conversation a couple months back is. You actually have to be a licensed pilot drone pilot to be able to do this as well this isn't for. People who just buy the drone from Amazon and are screwing around probably breaking a lot of laws in the process but no one really seems to monitor this stuff a little bit about so interesting you know it was really interesting unlock for this.
Guest: The base of this project was the the creation of the microgrown class to the sub 250 gram drone so it's like the best toy drone you can buy. But it is in Canada and this is you know we started in Canada and we hope this would be worldwide it didn't get picked up everywhere in Canada there there are not no regulations but very very few regulations on microdose because they're considered a sexually a toy. In Canada anyone can order a drone from Amazon or pick it up at Best Buy and start flying that same day there are no legal licensing requirements we recommend that you go through some some basic air space training we offer some some quick like safety tips like you know just things to be aware of obviously we are operating in air space with airplanes and helicopters and mandate aviation and buildings and things that you're craines. But in Canada you can buy a drone you can sign up and you can be flying the same day if we have missions available in your area we just launched missions this yesterday this week in in Burnaby, British Columbia a refresh there with our submissions in Boulder, Colorado this week so we have pilots flying there and I can share the map in a minute but if anyone would like to look themselves the link is explorer as you calm.
Jon Radoff: Yeah let's look at the map live because I think that would be cool to take a look at so I didn't realize the Canada was basically the Wild West when it came to drones so that's kind of fun.
Guest: When you go above 250 grams of licensing the difference and it's actually the same in the US the interesting distinction that the US made was the purpose of the flight so if the purpose of the flight is commercial in nature so if you're being paid or trading anything in kind you have to have a license regardless of the price. So in the US it's actually the same you can go to Best Buy by a micro drone fly that's fine recreational hobbyists if you want to go you know just just pick up a drone you control airplane it's just it's a toy as far as the F.A. is concerned right there. Right but as soon as you are paid or soon as you are being you know remunerated for your efforts then you have to have a license and so you know there was some debate internally about how much of that responsibility and liability do we take on ourselves how much of that do we put on the pilot to confirm that they they are meeting the legal requirements and we did decide to in the US to verify. If they had a part one of seven it's called a part one of seven certificate from the F.A. which is the licensing so in the US to fly with specksy you do need to have a part one of seven license is a process there you know there's some studying exam you have to do to get that license so you you would if you were starting from scratch and nothing but a drones you at least a couple of weeks to go through the studying and the licensing to get yourself set up with a mini drone to fly but the cost of those drones.
Unknown: We work right now with the DJI mini two mini three or many four they sort of range from like about 400 to $1,200 depending on the unit you get of the set up kit so it's very in line with what you might expect from like a geodonet minor or from a you know wing bits minor or even you know a little bit more maybe the helium platform but similar to a weather exam station you're looking at that like five $800 range to get your kit set up to start participating in the network.
Guest: So I see we've got the explore up here all the pink areas are areas that we've launched some missions as you zoom in you'll see the experts start shopping if you had to bolder Colorado it's the most live right now and really exciting was yesterday was the first time we launched on the network missions that are for points only. And you know we get immediately into sticky territory when we start talking about points with the TGE coming up but we now have missions that are cash and missions that are points so you go north of there is bolder you see a little bit color up top that's all locks of the purple color means those are done we flew all of them.
Jon Radoff: And if you zoom into the yeah there you go the hex map I mean I just love this map because it reminds me of playing any number of war games.
Guest: So what you're seeing there is purple is locked somebody's flown it gold is available open and there's a reward associated so if you click on a gold not the brown the gold gold gold yeah gold chasing money so that flight. That would have an eight dollar associated US dollar reward it would take about seven or eight minutes to collect that data for the drone to run its flight plan those brown ones that means some pilot has flown them recently but they haven't given us the data yet. So if you click on that brown one you'll see a flight time July 11th that's today and if you hover over the July 11th yeah what time did you fly that 11 25 am so about an hour and a half ago that was flown by a pilot so we can see the pilots flying and collecting the data and then when they uploaded turns purple and it blocks right so now we have that imagery. So as a network we know which and then the pink is a reservation so pilot has said I'm going to fly this so we have about five or six pilots flying right now today in Boulder we don't necessarily know who they are some of them are in our discord server they talk to see you know we have a community we have conversations but we might know we might not know but there are five or six pilots flying in Boulder collecting that data and uploading that imagery for the rewards.
Jon Radoff: This model I mean it's really obvious why you do this as a drone photography project basically but there's a lot of deep in projects that depend on a certain amount of geographic distribution of their nodes particular that maybe actual hard I'm thinking right now like beam you know don't quite add this to the roadmap anybody yet but thinking out loud like a geographic distribution map where maybe we could create a token in the world. So for example where if you get your node out into like a heavily utilized gaming region we probably need a checker node there that's you know low latency and available to check things locally so maybe maybe there's an idea in there for us in terms of our network roll out.
Guest: Yeah I mean the gaming tie in that we've found interesting enough and I learned this where probably more recently I might have maybe should have realized it was we're all likely familiar with Pokemon go and so you go out into this you know augmented reality world looking through your phone and you see these Pokemon and you capture them. What I didn't know was that the company that made Pokemon go in the antique was using that information to create mapping data right that the whole game is designed to get people to go out and essentially scan the world with their phones in areas of interest so you place a rare Pokemon if you have an area where you have a desert of information of imagery.
Unknown: And a military basis.
Guest: You get this data right it's a really really smart business model and they did it the I gave so they essentially were were paying nothing to have all of this data collected and all of this information brought into their systems by by you know the word tricking is not right they were providing a value in terms of entertainment and fun and things like that to the user but then user was providing a value back sort of you know unwittingly in a way. And allowing them to sort map all of the urban environments of the world you know like really really smart on on their behalf and so similar you know similar ideas and applications obviously could exist in a future state of what we're building right you can you can start to see how. Instead of $8 cash all these areas have a variety of different reward mechanisms whether they be crypto so they're whether they be you know you complete these tasks to achieve these levels to unlock these things and you get that real game of light aspect well the gears in my mind are turning thinking about additional ways we can gamify our own.
Jon Radoff: Network here at beamable which for our audience with a lot of our audience are gamers because they really came to learning about it through gaming even though. Technologically we are like a B2B technology sport games studios but. The token audience I think are excited about the gaming aspects I see some other questions coming in back to it's kind of like business model but it's the market you kind of commented a little bit earlier around the many kind of agencies and businesses that by this kind of imagery data and. To do maybe drill into that a little bit more like for the particular kinds of scanning that you're doing with drones in the fly to earn system. Who do you see as the most immediate use cases in terms of the buyers of your.
Guest: Yeah I'll very first go to market and can also can I steal the share for a second yeah of course and it'll it'll make a nice connection for the the users the viewers and the listeners. Our first go to market was we call it a city strategy so cities are very large consumers of imagery data of their own openness about the city wants to know what's happening within the city so very direct use case and I'll share in a moment. You want to know the right now with all the sg stuff you want to know the urban canopy coverage of trees in your city and you have a target for 2030 you're looking to increase your urban can at canopy coverage to improve livability for your citizens and you want to increase it by by 20% in the next seven years of five years.
Unknown: So you need a baseline first what is our current urban canopy coverage how many trees do we have now and then we're looking to increase and improve that urban county coverage over the next five or 10 or 20 years so we want to monitor that what we need to we need to see the whole city we need to see every tree in the city and sort of measure its canopy coverage area to to know if we're improving or trees are dying or growing or the trees we planted have.
Guest: To have taken and are starting to shade sidewalks and parking lots, etc so that's a really like one specific use case that this type of data would be for for municipalities to monitor that over time so they can refresh the data every year it's low cost it's high enough resolution now that might be able to be done with with existing airplane or satellite data. But may not depending on the specifications but another one is is like a new development so i'm a permitting department and just permitted a developer to build a new development on the edge of city to put in a hundred new homes right but i want to see how how that's progressing and i want to see it once a month just monitor the progress of that development over the next 18 months. Well it's going to be very cost it's not going to be cost effective at all so we're going to be very cost to try bringing an airplane back and fly that every month i just never get the cost for tens of thousands of dollars. it's expensive to send the inspectors out to look and then they're like taking phone cameras pictures with phones from different places and and i don't have a good like i don't get a good sense sort of of how things are moving along there may be writing a filling out a report telling me like oh this is changing so that's been done but i don't really know. Or or what you can now do with something like spexie and let's see if we can pull up the screen shows any sense give me some warnings I have two monitors i'm going to share that. And so if this pops up you you ask her we're looking at boulder these are the images so as they upload the images.
Jon Radoff: All right.
Guest: Processing pipeline and then we get it a high resolution aerial image of the city of boulder wow that is really cool.
Jon Radoff: Is that is that the maximum resolution that you're looking at like is that the.
Guest: Hello do you guys that's the current high resolution yeah. So imagine compared to Google Earth where you would you would lose resolution much much sooner so when a pilot uploads data. Goes automatically into a you know through a processing pipeline that we have it stitches these these panels so if i want to take a look at this building but actually i want to see it from. You know the other side so i'm going to come up here and look at it from the other side and i'm doing some sort of little inspection or me maybe this neighbor is new to this neighbor is under construction maybe i want to see that now. On a bookmark i don't have a time series yeah.
Jon Radoff: Graham is we're looking at this imagery first of all this is just really cool like tack looking at like. groans have flown out into the world you've got a mission assignment system you've got a hex grid that almost gamifies the mission process you collect it then we see these. Incredibly sharp photos of the surface of these areas relative to what everybody is seeing at this point like the satellite imagery out of google earth or Google maps. But now why the web three aspect of it you explained earlier you have you can pay people in cash if you're a US person flying US mission probably US dollars is fine if you're in Canada Canadian dollars probably fine for most people there. You mentioned earlier that it kind of cross border payments is a good use case for crypto is cross border payments really the issue here though what like.
Guest: Where why web three I guess is the question for every deep in project sure there's there's a few few things that it facilitates and we're for believers that if you don't have a web to business you won't have a web. Right like web three is isn't additive it's you if you don't have what I mean by business if you don't have a. Like a revenue generating like an income stream like like blockchain is a technology it helps me is not creating the need for drone photography no yeah and you know and in the most for the most part for me the real additives of right the real additives of the internet or social media where you could reach a bigger audience with your business. But the internet didn't make you a business you had to have a business and then leverage the internet you build a web page so people could find your business you had a social media so you could do marketing and outreach new people so those technologies were building blocks that allowed you to do new things and expand your business but the internet was not a business in and of itself and and blockchain I think is similar it's not a business in and of itself if facilitates certain things that you can now do as a business that maybe you couldn't do before and so one of the things that you know we have to two major aspects from the from the pilot side you know we think being a part of the network being invested in the network. I add stickiness it means that people get to to make some decisions around how the network grows and expands so there's a there's a governance aspect that we're looking at there's then as you mentioned this payment rails this becomes exponentially more valuable geospatial data becomes more valuable to more people as you expand right as you get more and more places and so it was a big lift for us even just to move to the US right to set up all of the entities to create the the corporation that was then able to interact with stripe and do banking and pay these pilots.
Jon Radoff: If we were to think about doing that 200 more times across the world I see yeah as you expand this is a global need so as you expand globally having essentially the payment rails that are inherent in crypto will streamline that whole process in addition to the governance and the stakeholder alignment aspects of having a token that that applies.
Guest: Yeah and there's a third aspect so one of the things when you're offering any type of reward or incentive to decentralize users is spoofing at every deep in project runs into this this issue of spoofing so for anyone that doesn't get it right away someone that's trying to upload a bunch of black pictures and pretend that they're the right pictures by changing the exit data or you're coming the exit data and so what we actually do at the time of capture is we have basically a system that we call proof of capture. So at the time that the pilots on site capturing we grab some data from the drone and from the device and from the images and we meant an NFT. Team minted that's a proof of capture NFT so now that pilot can fly multiple missions as you saw in the previous map the pilots were flying a bunch of missions in a row back to back to back before they upload the data so they might fly 20 missions. I didn't like to $8 $8 $8 $8 $8 $20 times to make 160 bucks they go back to their computer and they upload that data. How do we that that's you know that's 2500 images coming into our pipeline right what we do is we take all those images and we mash them back in the NFTs that were minted on site. Now we know that yes that image is the image that was created when the pilot flew that mission it's got to correct all the correct immutable information from the NFT and now we can approve the payment to the pilot.
Jon Radoff: So you effectively have a cryptographic proof that the correct work was actually performed where it was supposed to be performed. Yes because you wouldn't be the first time that crypto users tried to you bought the system generate rewards without actually doing the work we've never seen that it being a bit never that of a tactor quest leader board. So we got a million questions coming in but I have to take a moment here to release the code a lot of people show up speaking. Yeah we give a code out every live stream because we have a leader board where people are competing for points that are going to translate later into an air drop. They were going to release post tge and we have 10,000 people per season that can qualify for that so it's super competitive. We come from a gaming background so we gamified the leader board a bit ourselves and we're about to give that code out. I know we don't have the drum roll do we? Can you do a beatbox like what can we do here?
Unknown: I don't think Oscars are mused.
Jon Radoff: Okay we're going to give out the code though. The code today is transform tr a n s f o r m. So we'll pop that into the chat. And hopefully all of you get that but it's transform tr a n s f o r m looks like Oscars lagging a little bit and can in can be in it timing there when when we went live with the code but I did post it in chat so hopefully you'll all have a chance to use that code and redeem it. If you don't know about the code yet you got to be registered on our discord it's pretty simple just be registered there in discord don't be a bot type slash claim transform tr a n s f o r m. You got two minutes to do that and that is proof of participation we have we have a similar idea that's what we call proof of participation here.
Guest: Yeah that's your your people's proof of capture. In the meantime while people are doing that I just see an interesting question in the chat that I'll get to. But one of it is it just gives a powerful idea of how this can be leveraged really quickly because one thing we can do very quickly is spin up the pilots again because everyone wants to. So if we add more missions to the map the pilots go and apply them and we can vary those incentives that the pilots will receive so this was St Lewis and anyone that you follow the news will know that a tornado pretty big tornado went through there fairly recently so that St Lewis on April 23rd and that's the same area St Lewis on May 25th. And so you can quickly spin up a pilots post disaster you do obviously have to be very careful around disaster management because you have entities going in there doing different types of you know something like the Texas flooding that's happened right now you have lots of helicopter rescue you have lots of flow flying aircraft so we don't mess around with an emergency if it's effective because there are a whole bunch of f a things you can get into there. But a couple days after it's passed you can get in and now you can so now a cleanup crew can you know not even a specific cleanup crew but from a city perspective I'm a city planner I want to know how cleanup is going overall I can start to have an idea of what's happening in terms of repair and reparations you can imagine applications insurance right I've made insurance claim I can quickly come in and see yeah okay your house was very definitely affected by this tornado that just went through. I'm an insurance underwriter now I don't need to send an inspector out to the site to to verify or ensure that that happened I can see very clearly happened. So these are the types of applications from from different industries that you could imagine you know we've flown several iterations now with with Ellie County post policy fire so doing a bunch of cleanup and and remediation in the specific callisates north of Los Angeles. So there are lots of both emergency preparedness emergency measures a pre we get a base map now now when we have something like a flutter or a tornado or a hail storm we can get a post and we can do time series comparison and that's useful to a whole bunch of people and so that's much less fun much less less gamified. Yeah experience obviously still has an opportunity to be to be gamified and engaging in that way.
Jon Radoff: crypto has it has this question is your system a middleman or does it verify everything's done correctly it kind of relates to the question or the topic that you're just covering around proof of capture and making sure that that there is a legitimate upload but it seems like even beyond just the upload and making sure that the exit is what it was at the time it was there's still like a quality aspect to this too right like you could fly your drone and in theory capture data that's not.
Unknown: useful around the address.
Guest: So as I mentioned off the top if we jump back to the the fights are all automated the capture is all automated the pilot is hands off so you have a specific type of drone right with the me two me three me four which have very similar sensor sizes very similar pixel resolution and then so the pilot approaches this is the area they connect the drone to the phone through our mobile application and they essentially hit fly and then their hands off and so the drone they can launch the drone themselves if they like but essentially the drone will let's say. And to the specified altitude it will transit to the way point it will collect the imagery itself and the pilot is watching the clouds there for airspace deconfliction there for safety and regulatory compliance and take back control the joining time stop the mission essentially so if you had a low flying aircraft or helicopter coming first on like that they can quickly stop the mission but otherwise all of the data capture is happening autonomous so the summer settings are done autonomously the altitude is done autonomously the terrain follows done autonomously so if you're on a slope it's going to capture any meters here and it's going to transit upwards to make sure it's capturing also 80 meters here or vice versa so the the pilot doesn't have any essentially any control over the imagery data we take all that sort of responsibility. The interesting thing about that is we can adjust and vary the mission type that we load into the zone so we can fly these what we would I shared with these panoramas to these panel imagery where you can zoom you know rotate pan tilt zoom we're just starting launching some some media mapping missions which would create and for the Joseque or photo and and if I. Can actually I'm going to refresh this page and then Oscar I'm going to steal back from you and we're going to get back to the gaming use cases which are a little bit farther out but not not too much so I can steal this or did you give it back to me yep I'm going to switch what I'm sharing so I present something else.
Jon Radoff: While you're booting that up there's a there's another question I'll just tee up for you after you get that King seven up is asking well what if there's no pilot available nearby but I think the the broader question that that kind of provokes is how time sensitive is a lot of this data like do you need to get a pilot on it fast if a mission opens up it seems like some stuff is time sensitive like if you want to like see how a disaster recovery is going in a region seems like you're going to need very current information there other stuff maybe it isn't quite so but is it just come down to scaling out the network and having a lot of pilots competing for the space.
Guest: Correct yeah but the what's answer that no jump into what this is yeah a big part of my job is the operations manager is trying to ensure we have a robust enough community pilots across North America to be able to respond relatively quickly to any sort of request that we might get from a client emergency for very emergency response and disaster response is one of hardest for these variety of reasons this is obviously you know people are going through a difficult time. There are a meeting other emergency responders there so there's often power outages and things like that because problems for the cellular outages and cost by the drone. Mostly what we fly and create is a base map and then that's refreshable on on any iteration that that is designed so we could refresh that monthly accordingly or annually depending on demand and we we think that over time that requirement for new data for a variety of reasons and sources will will probably be meet a balancing point in terms of economic. viability around quarterly refreshes for large urban environments for for more rural or remote environments baby. But then if you know we have these sort of recurring role in collections. Like Uber might have the problem with finding a driver and a ride at the same time right to balance out your two me drivers sitting in their cars waiting for a ride you don't want too many riders waiting around for 20 minutes half an hour more to get a ride so you're constantly playing this game of. How many drivers will have how many riders we will have a similar challenge throughout our life you know we'll have this need to have enough pilots available and active that if something comes up they will get it in an appropriate period of time while not having so many pilots that they sort of get bored and turn out of the network because it's not enough to do and so that's balance that will always be playing.
Jon Radoff: What what I pulled up here is that you did you say it's going to be a gaming application maybe.
Guest: Yeah well so what I pulled up here is we're working with a large a large gaming entity this is actually for geospatial AI is what they're developing it for but you can see that this is a Gaussian splat collected you know imagery was collected by many drones in London so this is like. And the resolution is such that you can see what movie was playing at the theater you got the problem for wicked. Yeah yeah. Overround in here and they're working on improving these these flats more more there to some ground based collection where we do aerial collection they merge it all together it's really really incredible. You know the example I like to use with gamers and young kids is like imagine playing colonduty but it's on your street in your neighborhood and it's like the car that was parked at yesterday is parked there like you're almost playing it like like live in in augmented reality. Right but you're virtual reality like you're in a real metaverse and that's the gaming application but the next level application on top of that is what they're called spatial AI right so right now all our. All our ice chat you can see Gemini whatever it is they they've trained on text for the best majority some imagery they're doing video to do text speech now. But they don't understand the world in a spatial sense like how big something is like how do I get around this thing if it's now suddenly in my way. I'd imagine a delivery robot it's one of these food delivery robots that you see in some of the cities now and so if it knows the route and it knows has GPS coordinates and supposed to go from here to here great it can do that. Do it a hundred times in a row now if there's suddenly a pile of trash in its way right this is it doesn't it doesn't know how big is a pile of trash what does it mean if I have to go around it if I'm going to go out on the street so mostly just like stops I don't know if you ever saw that the people who are putting the cones on the WEMO cars because the WEMO car just like that I don't know what to do there's a cone on me right and so the idea that the number of organizations now working on is this spatial AI so basically training an AI. I model millions and millions and millions of constantly updated data points from the real world so it understands like a tree looks like this building looks like that if I want to go around it I can make this decision. Based on this information because I now understand the world in a way that I couldn't for from from say text. So that's these applications that we're looking at is.
Jon Radoff: Yeah and down he's kind of asking this question around web three based AR VR applications that could utilize drone data but I mean you kind of just answered it which is if you have a lot of photography can utilize and it wouldn't be limited I would think certainly wouldn't be limited to web three like I think that's one of the really critical things for absolutely everyone understand about deep in by the way who's watching this right now so. Deepin is a market which is crypto native in terms of the operator side like the people who run nodes whether it's a drone a node for compute like we have on beamable a weather station you tend to be somewhat crypto native on that side because you're receiving token. In exchange for providing that hardware but the customer side of deep in the customer side is not a web three market it's the global market for whatever this is if it's imagery data weather data compute for games to scale up. AI inference for all the AI applications out there like the customers are not limited by being web three native and that's what's very transformative about deep in. Deepin is a market for the first time really in crypto where the end user the customer buying it. Is everybody it is not a web three market on the customer side so everyone needs to understand that that's why deep in is such an important category within crypto it's the tentacles out into the non crypto world right from from the crypto world is these tentacles into every industry every vertical.
Guest: Of the non crypto world and actually you know we recently at at in the buy token 2049 announced the launch of layer drone which is actually our. Our crypto entity right so layer drone will be collecting all of the data and specs he becomes what we call a value added service provider so it purchases the data in in token from layer drone it. Transforms it in some way processes it you know makes an ortho mosaic map makes 3d model does a guising splat and it sells that onwards for cash into the into the marketplace of the world it creates a SaaS product it does and we envision lots of value added service providers pulling from the data to create different. You know applications different you know value add products that they will then sell into the marketplace for cash right so layer drone. Cells in token and all of these value added service providers that are buying that data and transforming into gaming you know they're using it in unity to create a gaming world so that buying the data they're creating their world and then they're selling their game.
Jon Radoff: Right getting into the whole other aspect of web 3 which is the inherent composability that's built into it which is what i'm trying to beat the drum on as well because it's something that I think the general crypto market doesn't actually understand how powerful. Composibility is so when you have protocols talking to each other where here's for example the image capture on your side and then other applications that can be built on top of that the transform the data. Make it more accessible or useful for other markets that's incredibly powerful of protocols talking to each other and it's another it's like another fourth or fifth piece of the use cases under your why web 3. Discussion we had earlier which is when you have protocols that talk about this kind of data and automate the utilization of hardware to capture data and store it and transform it now you're you've actually opened up a whole software development ecosystem. Where the value exchange is built in to the way these protocols talk to each other as opposed to the way we used to do things which is just this home mosaic of different sass applications with intermediate. Software between everything with their own separate payment rails that are incompatible of KYC like all the different issues around that so.
Unknown: Composibility is a huge game changer that is going to then be the next 10 or 100 X on all these dim deep in projects after you know after the initial use cases delivered because also again like we we deliver actual customer value now it's not one of these things where.
Jon Radoff: Like to often in crypto it's like derivative of a derivative of a derivative and then suddenly it changes the whole world right like no there's an initial. Market that is served that can pay now that needs this stuff and on top of that we will then build all kinds of other really compelling applications to.
Guest: Yeah we end others you know that's one nice things about decentralization and some of that stuff becoming your being and becoming more open source is that other people with great ideas and valuable ideas can can build on top of that as well.
Jon Radoff: Graham really compelling conversation wish we had more than an hour to cover this because there is so much here I think first of all. Just drones in general and I hope this opens everyone's eyes to like how important this technology is already and is becoming how we can decentralize it the market for this it's deep is a great application a deep in we started talking about protocols we could spend. A whole bunch of time just talking about protocols and composability we didn't even explain to people what Gaussian splats are I have an AI live stream that I do as well. And spots have been a recurring subject just search it online if you want to learn about Gaussian splats but basically with a very small number of photos that you take. You can reconstruct a 3D ray traced scene from it it's super super interesting in terms of all kinds of applications if you can imagine being able to populate a 3d world from just a few photos. We're looking at here. Yeah.
Guest: So. And many amazing thank you John for having me I love these conversations it's it's just fun when it you know it takes so root and we explore and experience some of these things and you can. I hope to tell it I'm passionate about it thank you to everyone who's tuned in and listen.
Jon Radoff: Well thanks very much to our audience we got 1400 you who have tuned in here live and I know that this is going to be a popular one probably after the fact as well. So a lot of people are going to view this and replay if you were one of those join us next time so you can be one of the folks who asked a question and be part of the live conversation. This this just this whole conversation inspires me to set up some panels on some topics as well like we could be talking about the possibility and some of these topics as well and this and do a real deep dive across multiple projects so maybe we'll do that next but. Graham thank you so much for taking part in this this is it for now so everybody tune in next time we're going to keep having these conversations.
Guest: Yeah can I get a discord link out to people I mean you can hit the the X or whatever but i'm not sure if I ended up sharing that with you Oscar but please feel free to join the discord whether you fly drones or not what I do is my.
Jon Radoff: I invite Mike you still copy did my thing i'm going to give it to the private check the specie dot com right which we're putting on the bottom there and find your discord from there but I think that's your call to action right so if you're curious about this project go to specie dot com go into the discord and yeah we're right down that crazy alpha you know alpha numeric code.
Unknown: I'm going to get a good one.
Guest: So what do you expect to do on X I think I was expecting yeah anywhere you Google it. Google it thanks guys.
Jon Radoff: Alright thanks a lot Graham thanks everybody for showing up today loved the questions super interesting project until next time Graham this has been fun. See you.