Latest / The Paul Barron Crypto Show / Tesla A.I. MEGA DePIN Coming🚀OVR x NATIX CEO INTERVIEW

Transcript

Tesla A.I. MEGA DePIN Coming🚀OVR x NATIX CEO INTERVIEW · The Paul Barron Crypto Show

0:05
Speaker 1: Today we're going to dive in to deepen, go a
0:07
Speaker 1: little bit deeper for you in terms of what this
0:09
Speaker 1: might mean around physical AI.
0:11
Speaker 2: You guys like it.
0:12
Speaker 1: My name is Paul Barren, Welcome back into tech Path.
0:14
Speaker 1: We are going to have a couple of people on
0:16
Speaker 1: to really dive in deep, both from Over and also
0:19
Speaker 1: from Natick. So stick around for that and we'll get started.
0:22
Speaker 2: I want to kind of open up here on the
0:24
Speaker 2: front end to just show you some of what Over
0:27
Speaker 2: has been able to do.
0:28
Speaker 1: And you guys have heard Diego on our show before,
0:31
Speaker 1: but where they are if you look at the Dune analytics,
0:33
Speaker 1: they continue to climb here on their cumulative three D maps,
0:37
Speaker 1: which is just exploding right now. So not only that,
0:41
Speaker 1: if you look at what has been building within the ecosystem.
0:45
Speaker 1: When you think spatial computing in general, Niantic has kind
0:49
Speaker 1: of been the i won't say the big leader, but
0:51
Speaker 1: they definitely are one that is recognized in the space.
0:55
Speaker 1: And I have a feeling that we may be seeing
0:57
Speaker 1: a little bit of a changing of the guard happening
1:00
Speaker 1: as well. The partnership that came about that really caught
1:04
Speaker 1: our attention was this one right here, Natix Network Our
1:07
Speaker 1: latest partnership with over makes it possible to reconstruct three
1:11
Speaker 1: D models of the space around us. This opens up
1:14
Speaker 1: the door to countless new use cases and you kind
1:17
Speaker 1: of see a little bit of what we're going to
1:19
Speaker 1: see in terms of three D mapping. That kind of
1:22
Speaker 1: technology is really interesting and that's where we're going to
1:24
Speaker 1: be going with a lot of this. Now, I want
1:25
Speaker 1: to jump to a clip because this is the smartphone
1:30
Speaker 1: of the naticks smartphone app that will give you a
1:32
Speaker 1: little bit more of a framework before we bring our
1:34
Speaker 1: guests in.
1:35
Speaker 3: Take a look, download our app, place the phone on
1:37
Speaker 3: the dashboard of your car, turn on the app, and
1:40
Speaker 3: that's it. Our ai technology turns your phone camera into
1:43
Speaker 3: a smart center that scans the road into text valuable events,
1:47
Speaker 3: and you earn points for every detection and every kilometer driven.
1:51
Speaker 3: You can redeem your points for the latest electronics health products,
1:54
Speaker 3: planet saving initiatives and Natix cryptocurrency. And with existing and
1:59
Speaker 3: upcome features like dash cam navigation and parking availability, Natex
2:04
Speaker 3: Drive and is your perfect driving companion. So why wait?
2:07
Speaker 3: Download natex drive and and earn with.
2:10
Speaker 1: Every drive All right, So I want to bring in
2:12
Speaker 1: our guests and then of course is Diego coming in
2:14
Speaker 1: from over and also Ali Raza coming in from Natix.
2:17
Speaker 2: Great to have you guys on the show today. How
2:19
Speaker 2: are you?
2:19
Speaker 4: Hi? Everybody?
2:20
Speaker 5: Hi everyone, Thank you Paul for having us.
2:22
Speaker 2: Yeah, good to see you.
2:22
Speaker 1: So, first of all, congratulations Diego continued growth here on overmapping.
2:29
Speaker 1: When I started to see some of the data, you
2:31
Speaker 1: guys are just flying up in terms of the engagement,
2:34
Speaker 1: I am assuming that this partnership is only going to
2:37
Speaker 1: explode that maybe even further.
2:39
Speaker 6: First of all, I mean, we've been knowing Alirisa since
2:41
Speaker 6: long time now, and so I'm really happy that we
2:43
Speaker 6: lended this partnership. And to us, this really makes a
2:47
Speaker 6: lot of sense because our data is really complimentary, it
2:50
Speaker 6: would say, because what we do is actually having super
2:53
Speaker 6: detailed maps on specific locasions, for example a mole or
2:58
Speaker 6: a square, and so I'm very profite. So we have
3:01
Speaker 6: between three hundred and one thousand pitchers every three hundred
3:05
Speaker 6: sclarmeters that allow us to have very precise they are
3:08
Speaker 6: in that location. While with Natis they are mapping streets
3:12
Speaker 6: with dash cams and basically they have a more extensive map,
3:16
Speaker 6: less detailed than ours, but still I mean with a
3:19
Speaker 6: much more bigger coverage. So actually combining these two data,
3:23
Speaker 6: we can actually have i mean a value added very
3:26
Speaker 6: proposition that combines the two. So for example, we can
3:30
Speaker 6: use the data for netics to actually relocate in space
3:33
Speaker 6: in cities in all the city with less precision than
3:38
Speaker 6: our maps, but still much more better than with GPS.
3:41
Speaker 6: But in our case we can go ultra precise in
3:44
Speaker 6: the location subcentimeters. So the two data I mean can
3:48
Speaker 6: look similar, but actually are very different and very complementary.
3:51
Speaker 4: I would say, when you.
3:52
Speaker 1: Look at the blockchains right now, Ali Reza, for instance, what.
3:56
Speaker 2: Chains are you on right now? Where is naticks living today.
4:00
Speaker 5: The moment on Solana?
4:01
Speaker 1: Is there a plan to go multichain? Is that coming
4:04
Speaker 1: down the road?
4:06
Speaker 5: Yeah, we definitely are looking into that as well, because
4:09
Speaker 5: I believe, you know, the future is multichain, but it
4:13
Speaker 5: comes with its own complexities. To be honest with you, you
4:15
Speaker 5: need to divide liquidity. It's something that we're looking at.
4:19
Speaker 5: There are some deep in specific chains for example like Peak,
4:22
Speaker 5: which we have a good relationship with as well, so
4:24
Speaker 5: we're exploring that as well to go multichain. But at
4:27
Speaker 5: the moment, the token is live on Solana blockchain, which
4:32
Speaker 5: I think is the best chain at the moment to
4:35
Speaker 5: build on, especially for deepen projects and deep in products.
4:39
Speaker 4: Yeah.
4:40
Speaker 1: Well, I mean early on, we did see a lot
4:42
Speaker 1: of activity. If you go back a couple of years
4:44
Speaker 1: to break Point, I would say Solana kind of ushered
4:47
Speaker 1: in a lot of the deepened projects that we saw.
4:49
Speaker 1: Even if you look at high Mapper, you also look
4:52
Speaker 1: at Teleport I think it was. I know Solana is
4:54
Speaker 1: getting ready to drop another event here in the United States,
4:58
Speaker 1: so I'm excited to maybe see more deepen projects coming
5:01
Speaker 1: out within that.
5:02
Speaker 2: Speaking of that, I want.
5:04
Speaker 1: To go over to your website. Ali Resident, this is Natix.
5:08
Speaker 1: You guys are dropping this VX three sixty. I'm very
5:11
Speaker 1: interested in this. I drive a Tesla. I've got a
5:13
Speaker 1: car that has one of these front cameras right now.
5:16
Speaker 1: So explain to me how this works, because I'm confused.
5:20
Speaker 1: I've actually point blanked other CEOs on this very topic
5:25
Speaker 1: of why not use cameras from all these cars that
5:28
Speaker 1: are out there, not only Tesla's, but there's other cars
5:31
Speaker 1: out there as well.
5:32
Speaker 2: But explain to me how you guys did this.
5:34
Speaker 5: Quite simply said, it's a functionality that the car manufacturer
5:39
Speaker 5: in this case Tesla allows the user to tap into
5:43
Speaker 5: the cameras or simply said, have dashcan footages and store them.
5:48
Speaker 5: And this is a functionality that is not just Tesla.
5:52
Speaker 5: It's a couple of other car manufacturers that are also
5:54
Speaker 5: having it. And this allows us to also collect you know,
5:58
Speaker 5: video data surrounding video. So it's not actually just your
6:01
Speaker 5: front camera. It's four cameras from the Tesla, one front,
6:05
Speaker 5: two sides and one rare camera exactly as you can
6:08
Speaker 5: see here that you're collecting the data on as well
6:11
Speaker 5: as certain other metadata because that adds value to the
6:14
Speaker 5: data for example gps I. This kind of data is
6:18
Speaker 5: very important when you want to kind of analyze the
6:21
Speaker 5: data and we get into use cases for autonomous driving,
6:23
Speaker 5: for map making. And this box is very simple. You
6:26
Speaker 5: just plug it into the glove box of your Tesla.
6:29
Speaker 5: There's a USPA port you basically there is already a
6:33
Speaker 5: USB inside that you take that out, you plug our
6:36
Speaker 5: box in. It has a very large storage unit. It
6:39
Speaker 5: records three sixty video of your entire trip and as
6:43
Speaker 5: soon as you are connected to Wi Fi, it uploads
6:46
Speaker 5: this data. So this can be where you park your
6:48
Speaker 5: car at home or at work, or if you don't
6:51
Speaker 5: have connectivity, you can just take the box. It has
6:54
Speaker 5: a very large storage unit, so it can be enough
6:56
Speaker 5: for one week of driving pretty much. You can just
6:58
Speaker 5: take it home.
6:59
Speaker 4: You have to.
6:59
Speaker 5: Justnnected to the electricity plug. It connects to your home
7:02
Speaker 5: Wi Fi uploads and the next day you actually get
7:04
Speaker 5: a notification on your phone to remember to take it
7:07
Speaker 5: with you when you're leaving home. So as simple as that,
7:11
Speaker 5: I still call it tap into existing hardware because if
7:14
Speaker 5: you wanted to build a three sixty camera, it would
7:16
Speaker 5: have costed us build up material three thousand dollars week,
7:19
Speaker 5: and with a three seventy nine dollars bucks, we're tapping
7:23
Speaker 5: into something that worth three thousand dollars and collects very
7:26
Speaker 5: rich data, and we basically use this data. One of
7:30
Speaker 5: the very easy use cases is map making, which is
7:32
Speaker 5: very similar to hivemapper or what we have been doing
7:34
Speaker 5: with our smartphone app. But the second interesting use case,
7:37
Speaker 5: which nobody has been really doing in the deepness space,
7:40
Speaker 5: is autonomous driving and is also part of the relationship
7:44
Speaker 5: with OvR. Is also physical AI and autonomous driving. So
7:47
Speaker 5: one of the pain points in autonomous driving is a
7:50
Speaker 5: lack of real world data. At this point, the problem
7:54
Speaker 5: is not AI. The problem is not compute because the
7:58
Speaker 5: NVIDIA is on the cap table of most of these
8:00
Speaker 5: you know, huge autonomous driving companies. The problem is real
8:03
Speaker 5: world data to be able to train their AI model.
8:07
Speaker 5: The AI will get better the more data you essentially
8:10
Speaker 5: feed into it, and then you also need data to
8:13
Speaker 5: test and validate. So once I train my AI model,
8:16
Speaker 5: before I deploy it into the roads and potentially let's
8:19
Speaker 5: see how it behaves, I need to make sure that
8:22
Speaker 5: it can navigate crazy scenarios, right like animal being on
8:27
Speaker 5: the curb side of the road and suddenly, you know,
8:28
Speaker 5: jumping in the car needs to be able to behave
8:32
Speaker 5: correctly essentially navigate through this situation. So this is why
8:35
Speaker 5: we need a lot of real world data and there
8:38
Speaker 5: is a real lack of it. And then the other
8:40
Speaker 5: thing is what the partnership with OvR is. We can
8:44
Speaker 5: do three D reconstruction of the space of the roads,
8:47
Speaker 5: and robots or autonomous cars in this case, can use
8:50
Speaker 5: this three D reconstruction to very precisely centimeter accuracy localize
8:55
Speaker 5: themselves in a three D space and be able to
8:58
Speaker 5: navigate and intract them.
8:59
Speaker 1: Well, that's going to be the significance of overs involvement
9:02
Speaker 1: there in terms of going out and so I see
9:05
Speaker 1: the partnership now of why that makes sense. A couple
9:08
Speaker 1: of questions I have for you, Alareza is when you
9:10
Speaker 1: look at Tesla, I mean, because if most Tesla's have
9:14
Speaker 1: premium connectivity, can you utilize the premium connectivity capability to
9:20
Speaker 1: upload the data as well or is it primarily wireless
9:23
Speaker 1: a Wi Fi network only.
9:25
Speaker 5: We could The thing is that if you want to
9:28
Speaker 5: record the entire video and upload it, which we actually
9:31
Speaker 5: needed because we have customers that will pay for this video,
9:34
Speaker 5: data is quite heavy. So prefably we want Wi Fi
9:37
Speaker 5: connectivity and we don't need this data in real time,
9:40
Speaker 5: to be honest with you, So there is no need
9:42
Speaker 5: to use a cellular connection especially, but we're working on
9:46
Speaker 5: smart ways where we're collecting some of the data that
9:50
Speaker 5: is very interesting and we need it in a more
9:53
Speaker 5: real time way using cellular connectivity. For example, if you
9:57
Speaker 5: have mapped the road already couple of times, uh, you know,
10:02
Speaker 5: and it's already the third time or fourth time the
10:04
Speaker 5: week that that this road is being mapped, and we
10:07
Speaker 5: don't need cellular connectivity. But if you have a blind
10:09
Speaker 5: spot that hasn't been mapped before and we really want
10:12
Speaker 5: the data collection to be as soon as possible, then
10:15
Speaker 5: just for that bit of the road that the data
10:18
Speaker 5: is collected and we didn't have any previous data, let's
10:20
Speaker 5: say for the past months, then we can use cellular connectivity.
10:23
Speaker 5: So these are kind of techniques that we are using
10:26
Speaker 5: to go hybrid between Wi Fi as well as cellular.
10:30
Speaker 5: So this is going to be a higher quality footage
10:33
Speaker 5: than you can see from the you know, Tesla app itself,
10:36
Speaker 5: because we don't you know, compress the video data. It's
10:39
Speaker 5: going to be for your entire trip, not just event based.
10:41
Speaker 5: You know, there are a couple of advantages that you're
10:43
Speaker 5: going to get on the On the let's say dashcam
10:45
Speaker 5: footage side, you get access to our marketplace. I think
10:48
Speaker 5: you saw it in the animation. You can as well
10:51
Speaker 5: as crypto let's say reward, you can also access other
10:54
Speaker 5: types of let's say, noncrypto reward, so that's another added value.
10:58
Speaker 5: And you can also get your name stoke in and
11:00
Speaker 5: be able to essentially get it to your wallet. The
11:04
Speaker 5: other app that we have, which is already live, this
11:08
Speaker 5: is if you don't have a Tesla, you can still
11:11
Speaker 5: download our app. You can put it on the dashboard
11:13
Speaker 5: of your car, drive around and also collect data. So
11:15
Speaker 5: this is these are two different products essentially that we're using.
11:19
Speaker 5: But Tesla will come with its own separate let's say
11:22
Speaker 5: driver Companion app, and the other app, which you don't
11:25
Speaker 5: require to have any specific sort of car, is called
11:28
Speaker 5: drive in. Essentially, you just plug it into the dashboard
11:30
Speaker 5: of your car and you drive around.
11:32
Speaker 1: So Aliver, I have to ask this question in the
11:34
Speaker 1: sense that I mean, why wouldn't Elon do this? I mean,
11:37
Speaker 1: he's got access to the cameras. Now he's got the
11:41
Speaker 1: device capability of plugging into the USB drive and pourting
11:44
Speaker 1: it out.
11:45
Speaker 2: It probably would be a software update.
11:46
Speaker 1: I would think, are you concerned about that that Tesla
11:50
Speaker 1: would just hyper leap over this and say, hey, we're
11:52
Speaker 1: going to pay Tesla tokens for your data because they're
11:55
Speaker 1: going to need this for AI training.
11:57
Speaker 5: They are doing it, Paul, and that's why Tesla Tesla,
12:00
Speaker 5: they are the best, in my opinion, crowdsource company for
12:05
Speaker 5: autonomous driving data. And that's why their FSD is better
12:09
Speaker 5: than any other product out there at the moment because
12:13
Speaker 5: other companies do not have access to you know, this
12:17
Speaker 5: much data. They actually do it one level even more advance,
12:21
Speaker 5: pre process the data because they have a big compute
12:23
Speaker 5: power as well in the car, which is quite actually
12:26
Speaker 5: crazy the amount of compute that the car itself can do.
12:30
Speaker 5: So they're doing it. But at the end of the day,
12:33
Speaker 5: we're not scared that let's say, Tesla will will block
12:35
Speaker 5: this because I think the direction that Tesla is going
12:38
Speaker 5: in general is let's turn the vehicle from a device
12:43
Speaker 5: to a platform where people can use the censors of
12:46
Speaker 5: the vehicle within a certain framework, and they can build
12:49
Speaker 5: applications on top of that. They can you know a
12:51
Speaker 5: way that Apple turned a normal phone that Ericston was
12:55
Speaker 5: building with its own application into a platform with an
12:57
Speaker 5: Apple Store, and this revolutionized. It was not straightforward back
13:01
Speaker 5: in the time, right people thought, Okay, why should I
13:03
Speaker 5: share kind of this access of building other apps and
13:07
Speaker 5: give it to other companies. I'm going to build the
13:09
Speaker 5: apps and I'm going to own them, and I'm going
13:10
Speaker 5: to charge the customer or the user all of the
13:13
Speaker 5: money that they have to pay for those apps. But
13:15
Speaker 5: Apple kind of opened the gate and then suddenly they
13:17
Speaker 5: became a platform. All of the developers came in, they
13:20
Speaker 5: tried to build different business models and products and use cases,
13:24
Speaker 5: and it was unbeatable. The only other company that could
13:27
Speaker 5: compete with them was Android. Google with Android, which was
13:31
Speaker 5: the same kind of platform. A lot of companies like
13:33
Speaker 5: Nokia and ericson and so on went bankrupt. So I
13:36
Speaker 5: see Tesla doing the same to cars that Apple did
13:39
Speaker 5: to mobile devices.
13:41
Speaker 1: Yeah sure, Well okay, so the other thing is going
13:43
Speaker 1: to be uptake in terms of the market because I know,
13:46
Speaker 1: with you Diego, you've had the benefit of you know,
13:50
Speaker 1: device users bringing on board data. Now we're going to
13:53
Speaker 1: see this mergence of now traffic data, trip data, road
13:57
Speaker 1: data starting to flow into this They go, how quickly
14:00
Speaker 1: would do you think we'll see an impact on what
14:03
Speaker 1: over is doing with this partnership and this data now
14:06
Speaker 1: flowing into what you guys are doing.
14:08
Speaker 6: Actually very soon, because actually the next month we're going
14:11
Speaker 6: to release our new proposition that is called OVER Business
14:15
Speaker 6: or basically we're going to start targeting Web two companies,
14:19
Speaker 6: not only more confine only Web three. The first two
14:22
Speaker 6: case will be actually to target with the value for
14:24
Speaker 6: predision that already have marketing agencies. But actually the second
14:29
Speaker 6: and maybe we'll be also together a target that we
14:32
Speaker 6: actually is right sharing companies. So right sharing companies for
14:35
Speaker 6: micromobility have a big problem because they need to relocate
14:39
Speaker 6: their assets, their scooters and their e bikes very precise
14:42
Speaker 6: in space to actually check the parkings, because if they
14:46
Speaker 6: don't park it correctly, they're going to be fined by
14:48
Speaker 6: the city. So the problem is that, of course they
14:51
Speaker 6: are limited with GPS. GPS doesn't doesn't have the precision
14:55
Speaker 6: to actually check the parking. So our data will be
14:59
Speaker 6: used to actually relocate in space to check if the
15:02
Speaker 6: scooter is part correctly, while the data from natives can
15:06
Speaker 6: be used to actually understand where they are in space
15:09
Speaker 6: when the GPS is not good enough. So, for example,
15:12
Speaker 6: the combination of these two data together will build a
15:15
Speaker 6: value proposition that we can actually monetize very quickly. So
15:20
Speaker 6: and not only this, I mean this is really the
15:22
Speaker 6: beginning of the propositions. But another thing that we're working
15:25
Speaker 6: on is what they are called larger special models. Now
15:30
Speaker 6: is something that we still need to fully announce, but
15:33
Speaker 6: basically these models are basically the counterpart to special intelligence
15:38
Speaker 6: to llms, so llms basically they just crunch texts, while
15:44
Speaker 6: these basically models are trained with data, with the visual
15:48
Speaker 6: data to understand the world. And there are many downstream
15:51
Speaker 6: applications to these models with words. In our case would
15:55
Speaker 6: be to have better mapping because these models build.
15:58
Speaker 4: Priors on the struct the world.
16:00
Speaker 6: But are's many use cases also for robots to understand,
16:04
Speaker 6: to segment the world around them and so on, and
16:06
Speaker 6: this is made possible only by data.
16:09
Speaker 1: One thing I wanted to hit you on Ali Raza
16:12
Speaker 1: was that when you look at Tesla, the front the
16:15
Speaker 1: front camera is I think one of the key things here.
16:18
Speaker 1: You've got the cyber truck already has it out there,
16:21
Speaker 1: the new model why which is not yet released here
16:24
Speaker 1: in the US. I know they're starting to sell in
16:25
Speaker 1: China and that's the number one selling truck and the
16:28
Speaker 1: number one selling car that would have this new data input,
16:33
Speaker 1: so that immediately is available. Are you going to meet
16:36
Speaker 1: this March shipping date?
16:37
Speaker 2: I think you guys had on your website. I think
16:39
Speaker 2: it was March. Was it coming to March?
16:41
Speaker 1: Yeah, March, just so next month you're going to hit this, Drey.
16:45
Speaker 5: I think the development is going quite well. So let's
16:49
Speaker 5: say end of March early April. We're going to able
16:53
Speaker 5: to ship, and it's going to come in batches. Of course,
16:56
Speaker 5: we're going to slowly start shipping. Usually with the deep
16:59
Speaker 5: end projects. Hard is the issue for us. Hardware is ready.
17:02
Speaker 5: We're going to announce as well a couple of partnerships
17:05
Speaker 5: that we have done with Bigger two let's say players,
17:09
Speaker 5: both as customers for Natix and partners let's say technological partners.
17:14
Speaker 5: But you know, a lot of this thing is already ready.
17:16
Speaker 5: Hardware is ready. There are like proper testings that we
17:19
Speaker 5: are doing to make sure that when we, you know,
17:21
Speaker 5: ship the product out everything is smooth. Because you know,
17:25
Speaker 5: I always believe you have this like one shot when
17:28
Speaker 5: you ship the product out right. If you ship it
17:30
Speaker 5: out and it's not good, it's it's just bad for everyone.
17:34
Speaker 5: So so we'd rather take maybe you know, one or
17:36
Speaker 5: two weeks more on the software side to test everything
17:39
Speaker 5: to figure things out. But on the hardware, everything is ready. Actually,
17:42
Speaker 5: we're going to get the hardware you know, in bulk
17:45
Speaker 5: into our warehouse I think even in the matter of
17:47
Speaker 5: a few weeks.
17:48
Speaker 1: Okay, all right, well good, A couple of points I
17:51
Speaker 1: want to hit on here, I'm looking at your Natis map,
17:55
Speaker 1: and so I'm in Miami here and I'm looking at
17:57
Speaker 1: the roads.
17:57
Speaker 2: When you zoom in on these, you'll see.
18:00
Speaker 1: See that there are a few spots that are kind
18:02
Speaker 1: of missing, like that little spot right there. And do
18:05
Speaker 1: you think this is going to be solved with this
18:07
Speaker 1: Tesla integration.
18:08
Speaker 5: By the way, this is one of the beauties of crypto.
18:10
Speaker 5: When they ask me, Okay, why are you using crypto?
18:13
Speaker 5: Is it just for the settlement, because you can also
18:16
Speaker 5: use fiat, I say no. One of the powers of
18:18
Speaker 5: crypto and Deepen is that we can mobilize our network
18:23
Speaker 5: extremely precise and efficient. And in this case, we have
18:28
Speaker 5: something called area specific missions or jobs. So we define
18:32
Speaker 5: an area and new work a block and then we
18:34
Speaker 5: put a bounty on it, and then exactly you can
18:37
Speaker 5: see missions here, and there are some missions that are
18:39
Speaker 5: area specific and you can essentially you go do a
18:43
Speaker 5: mission that is a certain area of New York and
18:45
Speaker 5: you get extra reward for that. So we can do
18:48
Speaker 5: targeted mapping based on what our clients have done, and
18:51
Speaker 5: actually in our focus cities like London, New York and
18:53
Speaker 5: so on, we can pretty much map any area within
18:57
Speaker 5: the city and two kilometer padding within forty eight hours
19:01
Speaker 5: from the point that we put the mission. So if
19:03
Speaker 5: that area has not been mapped yet, it's because we
19:07
Speaker 5: either didn't have the need for it or there was nobody.
19:10
Speaker 5: By the way, the data that you see on our
19:12
Speaker 5: coverage map is only from eighth of November. The project
19:15
Speaker 5: is live since April twenty twenty three, so you can
19:19
Speaker 5: imagine there's a lot of data that we're backfalling and
19:22
Speaker 5: it is still not there from coverage point of view,
19:24
Speaker 5: this is just the past few months of data and
19:27
Speaker 5: it will get covered in the next few months. Is
19:29
Speaker 5: if you ask me and to let's say your last part,
19:32
Speaker 5: definitely Tesla will hope although smartphone is much easier to scale.
19:37
Speaker 5: To be honest with you, Tesla definitely filled the gap
19:40
Speaker 5: in terms of coverage. But Tesla is not here for
19:44
Speaker 5: us to fill in those gaps because Tesla produces a
19:47
Speaker 5: type of data that smartphone cannot. It's a three sixty data,
19:51
Speaker 5: it's video data, which is not just used for map making,
19:54
Speaker 5: also for autonomous driving. You cannot use dashcam video or
19:57
Speaker 5: footage or smartph for autonomous driving application because you don't
20:02
Speaker 5: have the surrounding well.
20:03
Speaker 2: And it would.
20:04
Speaker 1: I think to Diego's point is is when you lose
20:07
Speaker 1: that three sixty dynamic, you're probably going to be able
20:10
Speaker 1: to train robots either. So that's a problem if you're
20:13
Speaker 1: going to look at that market in the future where
20:15
Speaker 1: all this data is going to be available. I have
20:17
Speaker 1: a feeling both these the combined data set that you
20:19
Speaker 1: guys are bringing to market is going to be very
20:22
Speaker 1: valuable to a lot of companies, especially as we start
20:26
Speaker 1: to evolve in these hardware components. Now I want to
20:30
Speaker 1: go and flip over to Niantic real quick, and I
20:32
Speaker 1: have a couple of other questions on the auto side.
20:34
Speaker 1: But when you look at Niantic, they've had some layoffs here.
20:38
Speaker 1: They've also canceled the NBA Marvel thing. My question, Diego,
20:43
Speaker 1: are we secretly seeing Niantic start to shrink in size
20:49
Speaker 1: because you've got Sony coming in with what they've been
20:52
Speaker 1: able to do, and possibly Sonium maybe being the new
20:56
Speaker 1: competitor here with their x yn Motion studio. All that
21:00
Speaker 1: now available. What is your thought about where the market
21:03
Speaker 1: is right now? It seems like there's a little bit
21:05
Speaker 1: of power struggle going on.
21:06
Speaker 6: Well, it's always difficult to totally understand what is the
21:10
Speaker 6: situation with ientic, because I mean they don't have the
21:13
Speaker 6: same transparency that we have in what tree I would say,
21:16
Speaker 6: but definitely, I mean there are quite a few moves
21:18
Speaker 6: on that direction, also the potential sale of Pokemon Goal brand.
21:23
Speaker 6: So what I can say, what I can see from
21:25
Speaker 6: the outside is that basically they have a very strong
21:28
Speaker 6: focus on the mapping side. They're still keeping that. Especially
21:32
Speaker 6: their acquisition of Canniverse goes in the direction. What they
21:35
Speaker 6: say in the website to create the largest reread map
21:37
Speaker 6: of the world is going in the direction. And also
21:39
Speaker 6: regarding using these data. So also the training of lgms
21:43
Speaker 6: is also something that they are building on. So I
21:45
Speaker 6: think they are probably concentrating their rectorts in that direction.
21:49
Speaker 6: And I think again robotics is a huge driver over that.
21:53
Speaker 6: I think that the next big wave, I mean, the
21:55
Speaker 6: next big technological way would be robots. And as you yes,
22:00
Speaker 6: and as you mentioned, I mean robots need to understand
22:02
Speaker 6: where they are in space. I mean, there is no
22:04
Speaker 6: doubt on the fact that this data and this technology
22:07
Speaker 6: would come more and more become more and more strategic
22:11
Speaker 6: once we have robots out there. So I think that
22:13
Speaker 6: they are concentrating in that direction.
22:15
Speaker 1: Do you think that x y N their spatial capture solution.
22:21
Speaker 1: You look at that they announced this, do you think
22:23
Speaker 1: they're going to leverage this into things more like game IP?
22:27
Speaker 1: You think Sonium in terms of you know, when you
22:30
Speaker 1: look at the blockchain integration and the potential here for
22:32
Speaker 1: Web three, do you think that's kind of the direction
22:35
Speaker 1: here or is there another strategy from Sony that could
22:38
Speaker 1: impede on deep In?
22:39
Speaker 6: It looks like they are really concentrated, I mean on
22:42
Speaker 6: the gaming part, but I mean also it looks like
22:45
Speaker 6: they are.
22:46
Speaker 4: Like a natural player in the deep in space. So
22:50
Speaker 4: probably we will see some muscles in that direction.
22:53
Speaker 6: So I do not be surprised if we could see
22:55
Speaker 6: more diseasive move also in that direction.
22:58
Speaker 1: Yeah, Sonium's daily transaction continue to climb here, so it's
23:02
Speaker 1: quite possible we'll see some more of that. There's only
23:05
Speaker 1: a handful. You've got Rivian out there Portia, which is
23:08
Speaker 1: getting into the point where they've got a lot more
23:10
Speaker 1: data coming into two vehicles now a lot of the higher,
23:13
Speaker 1: higher end vehicles to your point earlier is because these
23:16
Speaker 1: cameras are expensive, especially when you get into three sixty.
23:19
Speaker 1: Why not other brands, you know, other than Tesla. Is
23:21
Speaker 1: it just that Tesla is more of a platform or
23:24
Speaker 1: is it just you haven't done the deals yet.
23:26
Speaker 5: No, we are going to do other brands this year.
23:31
Speaker 5: So Rivian actually you just mentioned them. They also have
23:34
Speaker 5: this functionality and we're testing stuff. A couple of other
23:38
Speaker 5: brands are also are there, you know, BMW is there there,
23:42
Speaker 5: There are you know, quite a few other brands that
23:45
Speaker 5: we have either done successful testing or on the process
23:47
Speaker 5: of testing. And good news is that it will be
23:50
Speaker 5: the same box. Actually, so the way that we designed
23:54
Speaker 5: this box is to be able to integrate with these
23:57
Speaker 5: other vehicle manufacturers as well. Yeah, you just have to
24:01
Speaker 5: probably change the cable because Tesla has USB A right,
24:04
Speaker 5: you know, for another vehicle might be USB C or
24:07
Speaker 5: you know, another kind of connector. So the hardware is
24:10
Speaker 5: the same, the connector would be different. And we will
24:12
Speaker 5: definitely expand into other vehicle manufacturers.
24:17
Speaker 1: Yeah, listen, the growth of the market is really starting
24:20
Speaker 1: to accelerate. I had a chance to see Linus on
24:24
Speaker 1: one of the talk shows. He was showing the new
24:27
Speaker 1: you know, dual camera glasses that he was talking about.
24:29
Speaker 1: You've got x reel, which now has these dual cameras available.
24:34
Speaker 1: This seems to be coming to market pretty quickly and
24:37
Speaker 1: a fairly approachable price.
24:38
Speaker 2: That's a six hundred seven hundred dollars.
24:40
Speaker 1: And then if you look at the metaglasses on ray band,
24:42
Speaker 1: most likely we're going to see the next level of
24:45
Speaker 1: that right now. I think those are two three hundred bucks.
24:47
Speaker 1: You know, Diego, do you think we are close to
24:49
Speaker 1: mass adoption?
24:50
Speaker 6: Absolutely? And I think for real this time. Of course,
24:53
Speaker 6: I've always been super boulish on this space. That's why
24:55
Speaker 6: we've been building over but we're seeing so many signals
24:58
Speaker 6: that I mean, we have gone have some affordable smart
25:01
Speaker 6: glasses on the market and actually something that you want
25:05
Speaker 6: to wear, so something that you can wear all the day.
25:08
Speaker 6: And this really lets will make this space explodes, because
25:13
Speaker 6: once you can actually project things on your visual field,
25:16
Speaker 6: I mean, world game changes. And not only that, I
25:19
Speaker 6: think there is an interesting thing we're going to announce
25:22
Speaker 6: it actually quite soon. You know, there's also a big
25:25
Speaker 6: challenge over there. When you have all these people going
25:28
Speaker 6: around with s mar glasses on their faces, you will
25:31
Speaker 6: have thousands of cameras that basically look around you all
25:35
Speaker 6: day long. So it's like begging a Panopticon. So it's
25:39
Speaker 6: going to be very very important that we have some
25:42
Speaker 6: privacy by default solution to actually run DPS. Actually we
25:47
Speaker 6: are we are filing a patent right now and there
25:49
Speaker 6: so I cannot totally disclose out these words, but this
25:52
Speaker 6: is something that we're going to work on and we
25:54
Speaker 6: really believe will be very very important because of course,
25:58
Speaker 6: the privacy violation potential for these technologies is just incredible.
26:02
Speaker 6: It's just could have your camera so you every day.
26:05
Speaker 6: But taking away the backside, I mean, on the upside,
26:09
Speaker 6: really the adaptation of the XAR technology special computer and technologies,
26:14
Speaker 6: and we really explode. Thanks to these devices, we'll be
26:16
Speaker 6: able to generate more data and actually we will be
26:19
Speaker 6: able actually to finally use all the infrastructure that we've
26:22
Speaker 6: been built up to date.
26:24
Speaker 1: What about the AI agent side of the OOMI. Here's
26:27
Speaker 1: a good example just to show you possibly the integration
26:31
Speaker 1: of AI agents that could be incorporated into this.
26:34
Speaker 2: Do you see any future there?
26:36
Speaker 6: Absolutely? I mean, of course, you know we are. I
26:40
Speaker 6: would say that we're quickly moving to from a framewer
26:43
Speaker 6: ware LLMS and AI in general or just chatbots, we
26:48
Speaker 6: are moving towards the future where these ais are not
26:51
Speaker 6: any more chatbots are agents that basically add goals and
26:54
Speaker 6: try to achieve these goals by themselves. So of course
26:58
Speaker 6: these agents, again as I mentioned at the beginning, they
27:01
Speaker 6: will need data and so I mean all of the
27:04
Speaker 6: deeping platforms that can provide this data will be valuable
27:08
Speaker 6: also for these agents. So I see a future where
27:11
Speaker 6: we don't have only humans as I mean potential clients
27:14
Speaker 6: for these data, but also agents. So yeah, absolutely's that's
27:19
Speaker 6: a big integration potential.
27:20
Speaker 1: Over there, I'm looking at your statement or your point
27:24
Speaker 1: right here from the over Twitter account, and this is
27:28
Speaker 1: showing the AI agents as real estate you know, kind
27:32
Speaker 1: of mapped in with it. So that in itself is
27:35
Speaker 1: kind of intriguing when you look at, you know, projects
27:37
Speaker 1: like Propy, which we've talked about here on the show.
27:40
Speaker 1: We formately bring up Propy Propy of course kind of
27:43
Speaker 1: the real estate play. The potential here if you look
27:45
Speaker 1: at the growth of matter port, which has been in
27:47
Speaker 1: the market for a decade. Now, I guess to you, Alaraza,
27:51
Speaker 1: would this be something that would compete with matterport if
27:53
Speaker 1: you were to figure out a solution for real estate
27:57
Speaker 1: AI integration with what over is doing. I would see
28:00
Speaker 1: this potentially as another area beyond just roadmapping.
28:05
Speaker 5: Definitely. I think for us as NATIX, we have more
28:09
Speaker 5: focus on outdoor mapping, but if you ask me, I
28:13
Speaker 5: think we need both indoor mapping and outdoor mapping. Aerial
28:18
Speaker 5: mapping is also very important. There are a couple of
28:20
Speaker 5: deep in projects that are doing that. Extremely important to
28:24
Speaker 5: have satellite or aerial using drones, there are good projects
28:27
Speaker 5: that are doing that. And then when we go to
28:29
Speaker 5: indoor mapping, as you're showing, actually extremely important. And in
28:33
Speaker 5: the world that we have physical AI, a robot will
28:37
Speaker 5: go like a human being from the car to the pavement,
28:41
Speaker 5: to the sidewalk to inside the building. Needs to be
28:44
Speaker 5: able to navigate inside the building and then comes out
28:46
Speaker 5: and then sits in a car and then moves right.
28:49
Speaker 5: So we need to have this end to end. Let's say,
28:53
Speaker 5: I would say three D reconstruction of the world to
28:55
Speaker 5: be able to enable robots to interact seamlessly. So yeah,
29:00
Speaker 5: I would say even if Natix wouldn't do it, it's needed.
29:04
Speaker 5: It's one hundred percent needed. Somebody will do it, and
29:07
Speaker 5: it's complementary to what we're doing.
29:11
Speaker 1: All right, this is kind of mind blowing because I
29:14
Speaker 1: think we're at a very interesting state right now, a
29:17
Speaker 1: lot of new advancements here, definitely on the edge of
29:20
Speaker 1: really seeing what spatial computing can do. Let's talk about
29:23
Speaker 1: the roadmap real quick. I'll start off with you, Diego.
29:27
Speaker 1: Can you give me some insights to where over is
29:29
Speaker 1: heading for the rest of the year.
29:31
Speaker 4: Yeah, as I was anticipating.
29:33
Speaker 6: Basically, the big release that we have next month is
29:36
Speaker 6: basically business portal, so basically expanding what is our reach
29:41
Speaker 6: and basically reinforce me what we call the deep in flywheels,
29:45
Speaker 6: so not only revenues from web three anymore, but actually
29:48
Speaker 6: also revenues from web to rangjected in the web pless
29:50
Speaker 6: space and aligning to get more data and process more data.
29:54
Speaker 6: The other big thing that we are working on I
29:56
Speaker 6: also anticipated is the largest passion models right now, as
30:01
Speaker 6: we mentioned, we are right the two one hundred thousand mats,
30:04
Speaker 6: but what is the next target? And the next target
30:06
Speaker 6: will be half half a million mats? How we get there?
30:10
Speaker 6: Of course, we're going to keep mapping like we do now,
30:13
Speaker 6: but to actually scale even faster, we need to actually
30:17
Speaker 6: require less data to reconstruct with the same quality in
30:22
Speaker 6: order to do that, we need to use AI models.
30:25
Speaker 6: So AI models have larger passion models. What they can do,
30:28
Speaker 6: thanks to the priors that they have of the structure
30:31
Speaker 6: of the world, take less pictures, less data and fill
30:35
Speaker 6: the gaps with AI basically generating understanding what is in
30:39
Speaker 6: the angles that they didn't see and actually fill in
30:41
Speaker 6: the gaps.
30:42
Speaker 4: So this technology we will be able to actually.
30:46
Speaker 6: Have faster mapping, and not only that, actually will be
30:49
Speaker 6: also able to integrate like sparse data sets like for example,
30:53
Speaker 6: there are some open source data sets that are not
30:55
Speaker 6: super high quality like Mapillary for example, but still left
30:59
Speaker 6: data basically really being able to train these models. We'll
31:03
Speaker 6: allow us to actually do even more with less and
31:07
Speaker 6: so this is something that actually will bring us to
31:10
Speaker 6: the next step.
31:10
Speaker 1: Well, I'm looking forward to it. I mean, we've been
31:12
Speaker 1: following over all the way since twenty twenty one. I
31:14
Speaker 1: think was really the last time we identified over as
31:18
Speaker 1: a potential future star in terms of where compute was
31:21
Speaker 1: going for deepend before deepend was cool. So kind of
31:25
Speaker 1: interesting to see you guys evolving. Al Raza, how about
31:28
Speaker 1: you give me a couple of points that you guys
31:30
Speaker 1: are going to hit other than shipping in March.
31:32
Speaker 2: What else can you give me?
31:33
Speaker 5: Yeah, so I think. So we're going to work with
31:36
Speaker 5: a lot of fleet owners as well on the tesla
31:39
Speaker 5: side to really scale, you know, to tens of thousands
31:42
Speaker 5: of units globally. Live. Yeah, our focus essentially this year
31:46
Speaker 5: is on revenue generation, revenue. I think Deepen without protocol
31:50
Speaker 5: revenue is just got to do it, talking, Yeah, you
31:54
Speaker 5: got to do it, You got to do it. That's
31:56
Speaker 5: what differentiates Deepened with mean coins and a lot of
31:59
Speaker 5: other empty project in my opinion. As a matter of fact,
32:02
Speaker 5: you know, as a founder as well, we learned a
32:05
Speaker 5: thing or two. So with VX three sixty, before it's
32:08
Speaker 5: even live, we're closing customers. So at the point that
32:11
Speaker 5: it's live, we already close one customer exactly, and at
32:14
Speaker 5: the point that VX three sixty is live, the data
32:17
Speaker 5: will directly feed into the customer and they're paying for it.
32:20
Speaker 5: These avenues that are just explained, we would like to
32:22
Speaker 5: expand on these. Soil was just we started as just mapping,
32:26
Speaker 5: so the use case was only mapping using cameras for
32:28
Speaker 5: map making. Now autonomous driving is much more interesting because
32:34
Speaker 5: it's much more lucrative in terms of what they did
32:36
Speaker 5: for the data and the pain point. And then we
32:39
Speaker 5: are also going towards this visual positioning system three, the
32:42
Speaker 5: reconstruction of the space, which is something that we're exploring
32:45
Speaker 5: with you know OvR. So that's one thing, and then
32:48
Speaker 5: we're definitely also expanding to other types of cameras. I
32:51
Speaker 5: have always said it. We started with smartphone cameras because
32:54
Speaker 5: it's scalable. We then moved to Tesla cameras, and then
32:58
Speaker 5: there are a couple of other cameras that we were
33:00
Speaker 5: working on that we're going to add into our stack.
33:03
Speaker 5: So the goal is to have as many cameras that
33:06
Speaker 5: the vision. Of course, the vision of a company is
33:08
Speaker 5: something that is very difficult to achieve, but the vision
33:11
Speaker 5: is to have forty five billion cameras, which is the
33:13
Speaker 5: total number of cameras globally available as part of the
33:16
Speaker 5: extent work.
33:17
Speaker 2: That's crazy billion cameras. Forty five billion cameras.
33:22
Speaker 1: Jesus Lord, Okay, well that's a that's a lot of cameras, man.
33:27
Speaker 2: All right, So a lot happening over there.
33:29
Speaker 1: We'll definitely we're going to try to get one of
33:31
Speaker 1: these devices in and we're going to test it here
33:33
Speaker 1: in studio with our tesla's here and run it around
33:36
Speaker 1: and see what we can do with the new Natix text.
33:39
Speaker 1: So we'll be on the lookout for that. Guys, thanks
33:41
Speaker 1: a lot for coming in today. We really appreciate your time.
33:44
Speaker 4: Thanks for great sea.
33:45
Speaker 2: Paul, you've bet all right.
33:46
Speaker 1: So you guys, if you're not in the Diamond Circle
33:48
Speaker 1: right now, make sure and get in. I know the
33:50
Speaker 1: markets are really crazy right now, so this is the
33:52
Speaker 1: time to really kind of connectionist protection, that kind of component,
33:56
Speaker 1: which is getting with other like minded people. You can
33:58
Speaker 1: do that in the Diamond Circle. And if you're not
34:00
Speaker 1: in our telegram group, we leave a link for that.
34:02
Speaker 2: It's very easy to join.
34:03
Speaker 1: I'm in there every day, of course, follow me out
34:06
Speaker 1: there on x at Paul Baron.
34:07
Speaker 2: We'll catch you next time right here on Techpath