Local-First AI For Video Libraries, with Dr. Henry Kang의 전체 녹취록

00:05:27.759 --> 00:05:33.919 So when we think about this, that's the first of all, that's the hardest media, that's the hardest problem to solve. 00:05:33.919 --> 00:05:42.479 Second is what's the large like most like fast growing media format? 00:05:42.479 --> 00:05:42.719 Right. 00:05:42.719 --> 00:05:45.679 When we think about that, it's also video, right? 00:05:45.679 --> 00:05:52.239 Everyone is consuming video, everyone is generating or creating videos. 00:05:52.239 --> 00:05:56.959 So that that also means there's a huge market out there. 00:05:56.959 --> 00:06:01.279 Yeah, uh, there are so many creators working with with videos right now. 00:06:01.279 --> 00:06:06.239 So that's how we actually laying down our problem to start with videos. 00:06:06.639 --> 00:06:16.560 Because that's one of the challenges you have with video, is yes, everyone's consuming it, but you have to be able to access it, know where it is, put it in context, and do something with it. 00:06:16.560 --> 00:06:19.359 And that's what your system helps you do, right? 00:06:20.319 --> 00:06:21.359 Yeah, exactly. 00:06:21.359 --> 00:06:22.159 Yeah. 00:06:22.159 --> 00:06:25.439 Because that's the funny thing, right? 00:06:25.439 --> 00:06:27.759 When we are like 2026, right? 00:06:27.759 --> 00:06:31.119 And AI has has been so much advanced. 00:06:31.119 --> 00:06:46.639 But when you look at video files, when you look at computers, computer file systems, if you if you shoot a video and you import it to your computer, all the files are named as DJI003. 00:06:47.359 --> 00:06:48.000 Right, exactly. 00:06:48.239 --> 00:06:50.639 Or did this or yeah, exactly? 00:06:50.639 --> 00:06:51.519 Yeah, yeah. 00:06:51.519 --> 00:06:52.239 Yeah. 00:06:52.239 --> 00:07:01.039 So even though AI has been so much advanced, but when we look at how people are handling and processing those video files, right? 00:07:01.039 --> 00:07:03.759 It's like like the stone age, right? 00:07:03.759 --> 00:07:10.799 So uh nothing been nothing has been uh progressed since like early 2000, right? 00:07:10.799 --> 00:07:22.959 So and when we when people look at, but when when we as human, when we think about content, when we think about media, when we think about video, we're thinking about like visually, right? 00:07:22.959 --> 00:07:29.199 So the the like my background, that's the sunset at the Golden Gate Bridge, right? 00:07:29.199 --> 00:07:33.199 So it's never it's never like a DJI 00.3, right? 00:07:33.199 --> 00:07:39.839 So it's it's when we when people think about those content, we're always thinking about visually. 00:07:39.839 --> 00:07:45.199 We're thinking about it in a way we put it in the natural language description, right? 00:07:45.199 --> 00:08:03.039 So that's where we see the gap, and uh that's where really the product comes in to bridge that gap so that people can when they think about something, they can describe it in natural language and the technology, our technology comes in to bridge that gap. 00:08:03.039 --> 00:08:21.119 So it turns all those misnames, like file names, into uh really people understandable visual descriptions, and so people can find every clip, every moment they're looking for just by describing it. 00:08:21.519 --> 00:08:41.120 Because I'm looking at like a there's a lot of happening, for example, in the professional photographer space, where there's yeah, they're taking still images and they're using various cloud platforms, and there are some localized ones to call the images and whatever, but a lot of them are also doing a lot with video now, like you said. 00:08:41.120 --> 00:08:48.240 And again, I don't think there's a lot of people who are able to manage the video piece accepted. 00:08:48.240 --> 00:08:58.879 So this is sort of an application for that sort of thing, is where you could you know accentuate the video assets you're creating because right now it's almost an add-on to a photographer. 00:08:58.879 --> 00:09:09.360 Yeah, it's like you know, I'm gonna have a drone, I'm gonna shoot some video, I'm gonna have a B-roll with somebody shooting some video, but it's not really the feature of the event because I want to focus on the skills. 00:09:09.360 --> 00:09:18.720 But I can see where something like this could really be very advantageous to the storytelling if there was some context to do it efficiently. 00:09:18.720 --> 00:09:20.080 Because I think that's the challenge. 00:09:20.080 --> 00:09:25.040 You can do all the stuff you're saying, but it's it's terribly time consuming. 00:09:25.360 --> 00:09:26.639 Yeah, exactly. 00:09:26.639 --> 00:09:38.480 Like when we we have been talking with so many creators of photographers in a space, and uh, we are we're seeing quite a few of those uh challenges that people are facing. 00:09:38.480 --> 00:09:51.280 Like when we especially like when you talk about like when they're working on a project, when they deliver that project, and even nowadays, people are actually manually writing down like footage log, right? 00:09:51.280 --> 00:09:56.800 Yeah, writing down all what's happening in each of those files manually. 00:09:56.800 --> 00:09:57.360 Right. 00:09:57.360 --> 00:09:59.920 It's unimaginable. 00:10:01.360 --> 00:10:08.879 So talk a little bit about specifically about what your company does as a product or service for that. 00:10:08.879 --> 00:10:12.720 I mean, if somebody's interested, what how do they what do they do? 00:10:12.720 --> 00:10:16.000 I mean, are they buying hardware and installing your software? 00:10:16.000 --> 00:10:17.360 Are they having a subscription? 00:10:17.360 --> 00:10:18.800 Do you sell hardware? 00:10:18.800 --> 00:10:21.280 What is it that you guys actually do? 00:10:22.720 --> 00:10:33.200 Yeah, so what we believe is nowadays everyone is having a super powerful AI computer on their laptop already. 00:10:33.200 --> 00:10:49.600 When so what we're trying to do is uh what we observe is that they already have the power

of computation, it's just the lag that application that can help them unleash that AI power power on their computer. 00:10:49.600 --> 00:10:55.680 So so we what we deliver is the software, is in it's basically an application. 00:10:55.680 --> 00:11:00.160 So people download from our website and onto their computer. 00:11:00.160 --> 00:11:12.480 They can it supports both Mac and Windows platform, and they just install it as a as an app that as they as they download like a Zoom or any other software. 00:11:12.480 --> 00:11:12.960 Right. 00:11:12.960 --> 00:11:29.360 So they just install the software and we we actually have already set up all the computational pipeline, all the backend system, and without them knowing or worrying anything about the technology part. 00:11:29.360 --> 00:11:43.680 So we yeah, so that's after they install the app, they just add or give us access to the selected files or folders or directories or the archive, wherever it is, right? 00:11:43.680 --> 00:12:18.480 Archives, where wherever it is, and and we will process those data using their existing AI chips and help them basically we call index all the files and understanding all the visuals and helping them catalog them, and also if there's any like known people, we will be we'll be able to recognize the people in the footage in the video and public figures, but people can also define their own people, the people they work with. 00:12:18.480 --> 00:12:25.759 So basically, everything will be indexed and cataloged right on their computer. 00:12:25.759 --> 00:12:35.680 And then when they want to say, I'm working on a project and I'm looking for this, the shot of the sunset golden gate bridge. 00:12:35.680 --> 00:12:42.639 I just type in uh sunset golden bridge, golden gate bridge, and then Victor will help them uh find that clip. 00:12:42.720 --> 00:12:46.480 Uh, I mean, can I mean can does it have to be that finite? 00:12:46.480 --> 00:12:53.280 Could it just say bridge or sunset and it would give you an array of those, or do you is it more specific, the better? 00:12:54.240 --> 00:13:03.440 Well, whichever way they describe, like okay, say I just if I just type in bridge, I may find golden gray bridge, I may find my hand and bridge, right? 00:13:03.440 --> 00:13:06.639 So all the all the ones that match bridge will come up. 00:13:06.639 --> 00:13:14.399 And then the more fine-grained, the more descriptive you you put in, the the the better match it will be. 00:13:14.720 --> 00:13:19.360 Because I think when you know what's interesting is that when you see people do, and I know this is in generative, right? 00:13:19.360 --> 00:13:28.639 But when people do generative stuff, they start with a very simple prompt and they become more and more descriptive as they go, right? 00:13:28.639 --> 00:13:33.200 And I can see where learning that skill would be very useful for what you're doing, right? 00:13:33.200 --> 00:13:34.639 Because show me bridge pictures. 00:13:34.639 --> 00:13:39.920 No, I want sunset bridge pictures, I want California sunset bridge pictures, I want sunset bridge, you know. 00:13:39.920 --> 00:13:47.519 So I mean, you could be the I think when people learn those skills, they'll actually be able to use AI much more than they can transfer it, yeah. 00:13:47.920 --> 00:13:50.080 Yeah, yeah, they can transfer it. 00:13:50.160 --> 00:13:50.560 Yeah, yeah. 00:13:50.560 --> 00:14:08.560 So what about something like can you do something like let's say I'm in a corporate environment and there's a corporate archive, because that's a big business right now where all these companies have these, you know, let's say it's it's a big company and they've got an archive of content, and maybe they want to do a historical retrospective, right? 00:14:08.560 --> 00:14:14.000 Can they create that stuff within your platform or just find this stuff? 00:14:14.960 --> 00:14:28.879 Yeah, yeah, it's so interesting you brought that up because initially we're just talking about creator space for this audience, but really our user base actually goes much broader. 00:14:28.879 --> 00:14:59.519 And and actually the actually one number, one number is that the amount of Fortune 500 companies, actually, over half of the Fortune 5 companies' employees use our platform, use crypto to help them archive their uh business uh content, their business like meetings or their marketing footage, marketing assets, data. 00:14:59.519 --> 00:15:12.879 So just like you said, yeah, so it actually goes beyond just creator film uh filmmaker, photographer, and also extending to general business professionals. 00:15:13.280 --> 00:15:23.519 I can see where that would be a big advantage to have that local, as opposed to putting it on the cloud, where like you said, you may have corporate things you don't want to be on the open internet, right? 00:15:23.519 --> 00:15:25.759 For various reasons, security reasons. 00:15:25.759 --> 00:15:29.120 So that's exactly where it's like that would be a good now. 00:15:29.120 --> 00:15:38.000 If you're also talking about a photographer or videographer who's maybe an independent, again, they don't want their content getting out there either, right? 00:15:38.000 --> 00:15:46.080 If you're let's say you're taking, you know, well, like for example, one of the markets, yeah, one of the markets that we work, you know, that we cover is a school photography market, right? 00:15:46.080 --> 00:15:49.759 Taking pictures of school photos and yearbooks. 00:15:49.759 --> 00:15:58.399 And the challenge there is many of those states have very strict privacy laws on using this kind of stuff. 00:15:58.399 --> 00:16:06.480 So that's so if you care about your client's content from a privacy standpoint, you definitely want to look at an on-premise solution. 00:16:07.680 --> 00:16:08.560 Yeah, exactly. 00:16:08.560 --> 00:16:15.600 We believe that technology should really give that option back to the users, to the customers, so they can

choose. 00:16:15.600 --> 00:16:20.960 I want to keep the data local, I don't want to share with shared it with anyone. 00:16:20.960 --> 00:16:34.800 Actually, we serve quite a few of those uh industries segments that are very uh sensitive and uh data about the data privacy. 00:16:34.800 --> 00:16:42.639 Actually, some of the law firms use our product to archive and to process their internal data. 00:16:42.639 --> 00:16:54.480 Uh and also some of the doctors and uh finished finance industry, they use the crypto to index and uh process their internal data as well. 00:16:55.040 --> 00:17:05.680 Yeah, that makes a lot of sense because you know, even if you're using, you know, I have heard of some companies using things like you know, Amazon photos or Google photos for archiving. 00:17:05.680 --> 00:17:10.240 Thinking you're letting them read all of your content. 00:17:10.240 --> 00:17:48.009 You're opting in to letting them, you know, and then you're trusting that they're not gonna let your content be exposed to the world, which in a case of a company or if you've got privacy concerns with your content at all, that's not a great idea. 00:17:49.130 --> 00:17:50.490 Yeah, yeah, that's true. 00:17:50.490 --> 00:17:54.089 I think technology really should give people options. 00:17:56.650 --> 00:18:00.410 So what's the learning curve on something like this, right? 00:18:00.410 --> 00:18:06.490 So I mean, because I I'm I'm thinking of you know, a lot of my audience are you know small independent businesses, right? 00:18:06.490 --> 00:18:07.529 They're not Fortune 500. 00:18:07.529 --> 00:18:10.730 I've got a few Fortune 500 company listeners, but not that many. 00:18:10.730 --> 00:18:14.809 But you know, so so the learning curve on getting involved in something like this. 00:18:14.809 --> 00:18:18.410 Let's say I've got a Mac mini sitting there, fairly recent. 00:18:18.410 --> 00:18:24.009 It's got the M3 or M4 chip or whatever, so I can do some AI stuff on it. 00:18:24.009 --> 00:18:26.490 I throw your product on it. 00:18:26.490 --> 00:18:35.849 You know, what's how long is it going to take me to get up to speed on something like this if I'm not, you know, a an AI expert? 00:18:35.849 --> 00:18:39.289 If I'm just a regular business person who wants to find my stuff. 00:18:42.089 --> 00:18:51.450 We believe we are probably one of the easiest AI products, AI apps out there that's that's most user-friendly. 00:18:51.450 --> 00:19:01.049 I think our standard is we internally we put a AI like Apple's standard as our product standard. 00:19:01.049 --> 00:19:10.650 So we really want to make a product that's user-friendly and also easy for people to use, and they don't have to struggle too much to understand. 00:19:10.650 --> 00:19:22.329 Because, like when you think about it, basically there's zero need, zero requirement on the users about their technology, understanding of AI. 00:19:22.329 --> 00:19:35.930 They don't have to deal with command line, they don't

have to deal with like any like writing any code, like like working with any like Python or any of that. 00:19:35.930 --> 00:19:36.329 Right. 00:19:36.329 --> 00:19:41.930 For them, they they just they think they're just downloading an app. 00:19:41.930 --> 00:19:42.809 That's it. 00:19:42.809 --> 00:19:48.490 Just like they're downloading a Adobe Premiere or Final Cut, and that's it. 00:19:48.490 --> 00:19:55.049 When they and they don't have to worry about, oh, in the background, actually there are so many AI models running, right? 00:19:55.049 --> 00:20:00.170 So anything like that are all transparent from the user. 00:20:00.170 --> 00:20:12.089 So they don't for them, they're just working with an application and they download it and launch it, open it, and all the user interfaces we believe are quite intuitive. 00:20:12.730 --> 00:20:13.049 Okay. 00:20:13.049 --> 00:20:16.809 So what are some of the customers' feedback you've gotten? 00:20:16.809 --> 00:20:28.089 Because you've been in the marketplace for a while, and if you're you know, you've got some good attention from some of the sites on you know, recognizing recognizing your product. 00:20:28.089 --> 00:20:31.849 So, you know, what directions has that taken you? 00:20:31.849 --> 00:20:35.130 What surprises have you had when you hear from customers? 00:20:35.130 --> 00:20:36.890 Hey, I'm using it this way. 00:20:37.370 --> 00:20:41.370 Yeah, there has been quite a few surprises along the way. 00:20:41.370 --> 00:20:56.890 The first, like we already have mentioned, we already already talked about one point, is when we started, we really thought that crypto is in the product for the content creator market. 00:20:56.890 --> 00:20:57.609 Right. 00:20:57.609 --> 00:21:08.970 And then soon after we launched, we start to realize actually there are more and more people coming from many, many different industries that are starting to pick it up. 00:21:08.970 --> 00:21:18.089 And including lawyers, including professors, students, um doctors from every business. 00:21:18.089 --> 00:21:29.289 So that's that's one surprise we actually found because we thought, oh, crypto initially is for photographer to organize their video video footage, right? 00:21:29.289 --> 00:21:30.089 So yeah, yeah. 00:21:30.089 --> 00:21:39.370 And then then people are actually, for example, lawyers are using it to search for like like evidences in their law cases. 00:21:39.370 --> 00:21:40.170 Yeah, right. 00:21:40.170 --> 00:21:48.170 So that's really how surprising for us how broadly people are using it. 00:21:48.170 --> 00:21:48.490 Right. 00:21:48.490 --> 00:22:01.370 And also that's then we came to the realization that because getting access to data, getting access to information is so generic, right? 00:22:01.370 --> 00:22:03.210 It's so universal. 00:22:03.210 --> 00:22:06.089 Everyone wants that, and everyone needs that. 00:22:06.089 --> 00:22:14.970 And nowadays, when AI is becoming so so much powerful, it's easy to generate more and more content. 00:22:14.970 --> 00:22:17.370 And content is just exploding. 00:22:17.370 --> 00:22:25.930 But there's really no way to efficiently finding and efficiently managing those content. 00:22:25.930 --> 00:22:26.329 Yeah. 00:22:26.329 --> 00:22:26.650 Right. 00:22:26.650 --> 00:22:27.210 That's one. 00:22:27.210 --> 00:22:29.849 So that's first that's the first surprise. 00:22:29.849 --> 00:22:32.250 And the su second surprise came. 00:22:32.250 --> 00:22:37.130 When we recently launched our crypto MCP. 00:22:37.130 --> 00:22:44.650 So what crypto MCP does is crypto by itself is an is a standalone application. 00:22:44.650 --> 00:22:52.490 People download crypto and they import data to crypto and they search data, they search and or do other stuff in inside crypto. 00:22:52.490 --> 00:23:01.210 But what crypto MCP does is it connects crypto with all the other AI agents out there. 00:23:01.210 --> 00:23:17.930 So for example, through Crypto MCP, you can you can have your cloud, like right now it's for cloud or chat GPT PC app access your crypto application. 00:23:17.930 --> 00:23:32.650 And through that you can say, I want to find something when I'm chatting with cloud, I can actually use but is the data going back to cloud or is it staying local? 00:23:33.049 --> 00:23:33.529 Staying local. 00:23:33.529 --> 00:23:34.329 Yeah, it's local. 00:23:34.569 --> 00:23:35.450 Yeah, it's local. 00:23:35.450 --> 00:23:36.170 Yeah. 00:23:36.170 --> 00:24:00.490 And what's interesting, what interesting we found is that people are because before we we were thinking about one-way workflow is I have an idea and about a about a video or a project I'm working on, and I just and looking for a clip or looking for a snippet to go with that project. 00:24:00.490 --> 00:24:05.210 But then when we have the MCP, we find it the other way. 00:24:05.210 --> 00:24:16.809 Actually, people are using cloud and using clip combination of cloud and crypto in a way that I have those data, all those footage. 00:24:16.809 --> 00:24:21.130 What kind of video, what kind of content I can create? 00:24:22.410 --> 00:24:28.250 So it's kind of it's actually coming from the raw content side instead of starting with an idea and starting with that. 00:24:28.250 --> 00:24:36.009 So hey, you know, what can I do with all this content of the San Francisco of the of the Golden Gate Bridge, right? 00:24:36.009 --> 00:24:38.250 What can I create from all of this stuff? 00:24:38.250 --> 00:24:39.529 What can I create? 00:24:39.690 --> 00:24:40.089 Yeah. 00:24:40.089 --> 00:24:59.769 So it's actually when we have all those we call memory, actually, people when we have access to those memory, actually people have generated more ideas rather than they're sticking to one idea. 00:24:59.769 --> 00:25:00.250 Right. 00:25:00.250 --> 00:25:03.370 I think that's another surprise for us. 00:25:03.930 --> 00:25:06.170 That's an interesting way of thinking about it because you're

right. 00:25:06.170 --> 00:25:12.809 I mean, people are creating so much content now, and not even in the corporate space, in the consumer space, right? 00:25:12.809 --> 00:25:23.529 I mean, and what's it what it's done is it's created almost a paralysis where it's almost stressful to look at your photos on your phone now because you have so many. 00:25:23.529 --> 00:25:24.809 And how do you pick the one? 00:25:24.809 --> 00:25:26.250 And what do I do with it, right? 00:25:26.250 --> 00:25:28.490 And wouldn't it be great to just go exactly? 00:25:28.490 --> 00:25:33.289 Hey, listen, here's my trip overseas that I just went on. 00:25:33.289 --> 00:25:35.529 Make me a nice something out of it. 00:25:35.529 --> 00:25:37.930 And maybe you take the videos and you take the still. 00:25:37.930 --> 00:25:38.730 Exactly. 00:25:38.890 --> 00:25:39.450 Yeah. 00:25:39.690 --> 00:25:39.930 Yeah. 00:25:40.490 --> 00:25:41.370 Exactly. 00:25:41.370 --> 00:25:42.730 Yeah, exactly. 00:25:42.730 --> 00:25:47.130 Like we recently also internally we made a funny video. 00:25:47.130 --> 00:25:55.690 We we tried to index uh a thousand of Elon Musk's interview clips. 00:25:55.690 --> 00:26:16.809 So once we have indexed that in crypto, we connect it to cloud and asked it, we have all those Elon Musk uh interviews, and we want to make we want to create a funny video of Elon Musk and maybe a video with some music. 00:26:16.809 --> 00:26:37.609 And what really surprised us is after after the AI looked through all the interviews of Elon Musk, it actually came up with an idea that it saw Elon Musk repeat one phrase in many of those interviews. 00:26:37.609 --> 00:26:37.930 Right. 00:26:37.930 --> 00:26:42.890 Elon Musk, he likes one phrase around the word. 00:26:42.890 --> 00:26:43.690 Okay. 00:26:43.690 --> 00:26:47.450 He repeats he repeats that phrase in many of those interviews. 00:26:47.450 --> 00:26:48.490 What phrase is that? 00:26:48.490 --> 00:26:50.490 Around the world. 00:26:50.490 --> 00:26:51.289 I can type it in. 00:26:51.450 --> 00:26:51.930 Yeah, yeah, yeah. 00:26:51.930 --> 00:26:52.809 No, that's fine. 00:26:53.049 --> 00:26:53.690 Yeah. 00:26:57.930 --> 00:26:58.250 Okay. 00:26:59.849 --> 00:27:13.289 So then so then the AI decides, okay, then we can actually make a funny video of Elon Musk by putting his interview clips into you know Daft Punk, right? 00:27:13.289 --> 00:27:18.809 Daft Punk has the MV titled Around the World. 00:27:18.809 --> 00:27:23.049 So it's just repeating the around the world, around the world, right? 00:27:23.049 --> 00:27:34.009 So yeah, so AI came up came up with that idea and including all those, you can you can actually play that video in the podcast later on. 00:27:34.009 --> 00:27:54.250 So it decides the topic and it decides which music to pick, and then it actually pulls all those tips that you know must repeat the around the word, a phrase, and bug that in in back into that video. 00:27:54.250 --> 00:27:55.210 Oh, that's

crazy. 00:27:55.210 --> 00:28:02.089 I'll send you the the the the so eventually AI created the whole video by itself. 00:28:02.089 --> 00:28:06.170 So it's it's very funny. 00:28:06.809 --> 00:28:14.329 Yeah, that's yeah, that's interesting because I mean, you know, obviously that kind of video has been around a long time, right? 00:28:14.329 --> 00:28:24.170 People have done that, but what you've created there is a really opening that up to somebody who may not have those technical skills to do that or the time or the all that. 00:28:24.170 --> 00:28:42.650 But I mean, people have been doing that sort of thing for a long time, and I think that's really what I think is interesting with sort of the AI revolution that's happening now is people want to do these things, but it's either time prohibitive, cost prohibitive, skills prohibitive, right? 00:28:42.650 --> 00:28:46.970 Because maybe they may not have the video editing skills to make those things work. 00:28:46.970 --> 00:28:51.049 And I think it's gonna open that up to a lot of a lot of people. 00:28:51.049 --> 00:28:53.529 And some people obviously some of it's gonna be crap, right? 00:28:53.529 --> 00:28:56.089 Some of it's gonna be terrible, some of it's gonna be really good. 00:28:56.089 --> 00:28:58.009 It's just like everything else, right? 00:28:58.009 --> 00:29:01.130 And I think it'll be interesting to see what happens. 00:29:01.130 --> 00:29:11.370 Now, with something like that, let's say you were to post that somewhere, let's say on Facebook or Google, would you have you would have to say that's AI generated? 00:29:11.370 --> 00:29:13.609 You would you have to indicate that's AI or not? 00:29:13.609 --> 00:29:15.529 Because you're not generating anything. 00:29:15.529 --> 00:29:17.210 Is it AI editing? 00:29:17.210 --> 00:29:22.650 Is that I'm curious, you know, because there's such a sensitivity to that now. 00:29:22.650 --> 00:29:33.289 With because like you create something in a generative AI, you can't you get we have to say it in the post. 00:29:33.289 --> 00:29:38.329 This is creative, but you you're not generating anything actually, you're just editing. 00:29:39.930 --> 00:29:41.769 Yeah, yeah, that's a good point. 00:29:41.769 --> 00:29:51.289 You I think it doesn't really fall under the standard AI generated context category. 00:29:51.289 --> 00:29:53.849 Yeah, but it's fun. 00:29:54.009 --> 00:30:01.130 I think people yeah, I'm I'm just curious because it is, I mean, AI in that case is making some editorial decisions. 00:30:01.130 --> 00:30:04.809 Yeah, it's assembling something, right? 00:30:04.809 --> 00:30:05.769 It'll be interesting. 00:30:05.769 --> 00:30:07.690 I I don't think it's gonna shake out. 00:30:07.690 --> 00:30:27.769 I think that's but it is interesting because you know, I think people are super sensitive sensitive now to to generative stuff where somebody does something they wouldn't normally do is of course you know not great, but assembling a video a

music video, that's that's fine. 00:30:27.769 --> 00:30:29.210 That's who who cares? 00:30:29.210 --> 00:30:30.730 That's fun stuff. 00:30:30.730 --> 00:30:32.650 I found the link. 00:30:32.650 --> 00:30:33.609 Yeah, I saw that. 00:30:33.609 --> 00:30:33.930 Thank you. 00:30:33.930 --> 00:30:34.650 Appreciate that. 00:30:34.650 --> 00:30:38.650 Yeah, we'll we'll drop that into the final edit, and we'll see what we'll see what we can. 00:30:38.650 --> 00:30:41.130 Cool, cool. 00:30:41.130 --> 00:30:45.930 So, over the course of the company, where do you think you're gonna be heading to, right? 00:30:45.930 --> 00:30:49.529 I mean, you're getting a lot of feedback from customers, you're finding new markets. 00:30:49.529 --> 00:31:02.170 Is there a media type you think you're gonna specialize in, or are you just gonna be the all-for-one type thing, or does it even matter media types anymore? 00:31:02.170 --> 00:31:04.490 Because it's all kind of a mishmash today. 00:31:04.890 --> 00:31:05.370 Exactly. 00:31:05.370 --> 00:31:09.130 I think the that mishmash is the right word. 00:31:09.130 --> 00:31:25.210 I think right now it's it's just we as a human being, we're just so buried and so overwhelmed by the explosion of all those data, all those content. 00:31:25.370 --> 00:31:25.609 Right. 00:31:25.769 --> 00:31:28.890 And including we started with video, right? 00:31:28.890 --> 00:31:40.250 We started with the hardest format, but we we are also, or we have already included like audios, included photos, um, like all those easier content, right? 00:31:40.250 --> 00:31:41.690 Easier format. 00:31:41.690 --> 00:31:49.769 And we are probably also going into PDF, Word, all those like more structured data. 00:31:49.769 --> 00:32:08.410 But really, I think uh regardless of the media, regardless of the content, what's really fascinated and what we are really obsessed with is really to help the humanity solve the memory problem, exactly. 00:32:08.410 --> 00:32:16.250 So I believe that's uh that's going to be even more in the future, right? 00:32:16.250 --> 00:32:33.849 When when we have more rich content, more more and more of our everyday life, uh, everyday work, are interacting with and overwhelmed by those contents, then we're going to have a more memory profile platform in the future. 00:32:33.849 --> 00:32:39.370 So really, I think to summarize, I think that memory problem is really what we're solving. 00:32:39.769 --> 00:32:40.089 Okay. 00:32:40.089 --> 00:32:47.849 So if somebody wanted to experience Clip 2 and play with it and do where can they go for more information? 00:32:47.849 --> 00:32:51.370 Because you can just download it from your website, right? 00:32:51.849 --> 00:32:52.490 Yeah, yeah. 00:32:52.490 --> 00:33:06.089 They can just go to our website, crypto.com, and they will be able to find all the solutions out there, regardless of their if their Mac or Windows or either of the platforms. 00:33:06.410 --> 00:33:13.609 Is there a I mean, is there a pricing based on the size of the archive or how many devices it's on? 00:33:13.609 --> 00:33:15.130 How does that even work? 00:33:15.130 --> 00:33:23.930 You know, I mean, because I would think you've got giant companies and they're not paying all, you know, I'm just curious how that scales, right? 00:33:24.569 --> 00:33:30.089 So right now we are offering it as a more simplified subscription, right? 00:33:30.089 --> 00:33:38.250 So it's 25, I believe it's 2499 a month, regardless of the how much content they put in. 00:33:38.250 --> 00:33:40.089 Yeah. 00:33:40.089 --> 00:33:42.490 So we want to make it as easy as possible. 00:33:42.490 --> 00:33:49.849 So don't they don't have to worry about oh, I have one terabyte data, I have 10 terabytes data. 00:33:49.849 --> 00:33:50.170 Right. 00:33:50.170 --> 00:33:53.049 So it's a it's a simple yeah. 00:33:53.450 --> 00:33:54.490 So what about devices? 00:33:54.490 --> 00:33:56.329 What if you had more than one device? 00:33:56.730 --> 00:33:57.690 More than one computer. 00:33:57.690 --> 00:34:03.930 Right now it's yeah, it's actually right now it's one subscription across the platform. 00:34:03.930 --> 00:34:04.410 Okay. 00:34:04.410 --> 00:34:07.289 Um we yeah. 00:34:07.289 --> 00:34:21.529 The where we haven't really looked more into the multi-device just yet, but uh right now I think people are just genuinely finding it useful to use it. 00:34:21.529 --> 00:34:25.769 Um where we haven't think about like exploiting, exploiting it. 00:34:25.929 --> 00:34:26.089 Yeah. 00:34:26.329 --> 00:34:27.369 We'll see in the future. 00:34:27.369 --> 00:34:27.609 Yeah. 00:34:27.769 --> 00:34:27.929 Yeah. 00:34:27.929 --> 00:34:34.650 Well, I mean, I mean because I can see the point where maybe you'd have depending on the size of your content library, right? 00:34:34.650 --> 00:34:46.570 You may have a device you'd want to, hey, this is my video device, this is my corporate archive device, and maybe they'll talk to each other, but for up for speed reasons, besides the archive, you may want it on separate devices. 00:34:47.210 --> 00:34:48.170 Yeah, yeah, yeah. 00:34:48.170 --> 00:35:01.850 I think I think honestly, we are genuinely just interested to see how people are going to use it, uh, how the product is solving their, like helping them improve their workflow. 00:35:01.850 --> 00:35:14.090 And once people they're going to actually we see more people bring us ideas how they they can update and how they can improve their workflow. 00:35:14.090 --> 00:35:27.050 And it will generate like I we believe it just will bring us more idea about how we can adapt our business model to those, to those workflows, to those use cases. 00:35:27.530 --> 00:35:27.850 Okay. 00:35:27.850 --> 00:35:29.210 Well, thank you, Henry. 00:35:29.210 --> 00:35:31.930 It's been great to talk to you and meet you and learn more about the company. 00:35:31.930 --> 00:35:32.810 I appreciate it. 00:35:32.810 --> 00:35:34.170 Thank you so much for your time. 00:35:34.170 --> 00:35:39.290 And I encourage everyone to go try the software and download it and give it a whirl. 00:35:40.730 --> 00:35:41.530 Thank you, Gary. 00:35:41.770 --> 00:35:42.970 Yeah, thank you, sir.