The AI Warning Nobody In Silicon Valley Wants You To Hear | Ep. 426 with Jamarri J. Founder of Klyno AI

0:00If you're afraid of AI, what you really should be concerned about is how we use quantum. Boom, it's instant. All encryption systems wiped out in the year. This is Jamarri Jones, founder of Clino AI, one system that connects you to every AI you use and remembers all of it. He works a 10-hour shift, then builds Clino until sunrise, and he says the entire industry is building it wrong. There's thousands of them. And every single one that you use, they're either just a wrapper, where all it is just an API and a little chat box. With Frontier models, they seem to sometimes, again, they still forget.

0:39It's great that we're using all these Frontier models. It's great that we're using all these cool tools. But are those tools being built for you or against you? Do you think that we'll ever really own our own data? I'm going to be honest. The mission, the philosophy is that...

1:04you didn't start an ai company for the hype or for the money is what you said you started it because you were frustrated but what are you frustrated with i'm specifically frustrated with the fact that there's so many ad tools i mean there's thousands of them like in every single one that you use they're either just a wrapper where all it is just an api and a little chat box and all they do is they go pay me$90 a month or whatever or$50 a month but you're not actually getting anything it's completely barebone I got tired of it because there's all these tools but then I have to switch tools to this then I have to switch tools to that just to use the way this is really the greatest AI and I was like well you know what why don't I build something that takes all of that, puts it into one system.

1:54And I solve one of the biggest AI issues at the exact same time being the memory and the context. So no more context is being lost across your tools. Can you explain more? Because, and I've experienced this, I've been at events and every time I ask someone, what are you building? And they're like, you're saying it's a wrapper. They're building the same thing. They're not really even solving many problems. Yeah. So Clino, the very big thing that we actually saw and it's kind of crazy because you're still in a kind of doing like one fail swoop essentially memory right ai has an issue with trying to remember things especially when it's large quantities of said things so we did is in clino we actually built a thing called clino brain and what this does is is the same way that big frontier models kind of have their system we kind of create something a little bit different because i noticed that with frontier models they seem to sometimes, again, they still forget, right?

2:51They're still at a lot of that forgetting. So the thing we solved is, like an actual brain, there's nodes, right? And each of these nodes will actually remember very particular contexts. So when you're actually in your chats and you're typing or you're needing to refer to something, it goes to the brain where there's a whole little core in the middle, and then it shoots out to different nodes. So instead of it just remembering chunks of data, how like, for example, whenever you upload like PDFs or any type of file to front-term models, that's how they remember, basically. It's all in chunks, basically.

3:31But instead of us doing like the chunking, we said, okay, we can do chunking, but like we need to actually take it a step further. And so what's happened where there's a node in the middle, that's the general summary, right? so it has all the chunking there but then we need to break it out even further within hey from this node here to then this node and the nodes around it almost like brain neurons basically it's the exact same way so that way it will fire off and then fire back basically to the main core which then will go into your chat or whatever tool you're using within our app so being in your early 20s how do you look at ai and the reason i ask is there's been this big uproar around people that love it and people that hate it right at college universities somebody is saying ai and they get booed on stage you though being in your early 20s who's building with ai how do you look at it i look at it is and this is actually our philosophy it's

4:34technology shouldn't replace people because technology like technology is the representation of humanity so technology that we have represents how we all are whether we like it or not as a species and so the way that i view it is that it's not something we should be afraid of right and it's also not something we should slow down or let pass up right we shouldn't also have it be consolidated to big people either because, well, it's going to affect everybody, right? I'm not in the market of, you know, making people lose their jobs or anything like that with AI. What I'm in the market is, is, hey, here's a system that you own your data, number one, which is the biggest thing, and the privacy that comes included with it, and you can keep this locally right meaning that hey this will run your own machine right so for those who are actually environmentally conscious because that's what a lot of people when they're booing they're like boo ai right but it's funny because most of those kids used ai basically to get through all their classes so it's kind of like a it's like guys i know you hate it but you also kind of use ai to you pass like a lot of your stuff so there's a valid concerning sword but also i think this is something everybody needs to have a voice in and on because currently i feel that ai is just it's all consolidated to very few voices trying to dictate where you go and it's not going in a good direction what do you think would make it a good direction like in in five ten years from now more voices get in more people get to dictate that what is the most positive outcome the most positive outcome i can't bring up video games but i shouldn't bring up video games but the most positive outcome i would say is one which in every city all right which is something we're working on well can't say too much about it but in every city or in every household you own your own almost like fraction of an ai that only works on your data right that has actual controls in place so that way it's not being sent out but you still have a piece of something that's like a frontier right you still have a piece of the actual future that you own that is a very great like outlook that's that's our happy ending where ai isn't there just you know replacing you you have your own assistant that works off your own data that's private to you right that's the number one thing that it's actually private to you.

7:12You're not losing. There's no ownership loss. The biggest and the greatest outcome that we can have from this is that people have control and they have the ability of, hey, I'm like, it's almost like having an AI assistant, right? You know, like kind of from the movie, not movie, but the show, The Jeffersons, right? And they used to have that little robot or whatever. I know, guys, I'm 23. So obviously, yes, I can, you know, I know the show. right but it's like having that little robot that would do everything for you but guess what you own it so there isn't like this long-term pain or anything like that but no it you own it it's in your own home knowing how things are now do you think that this is even possible knowing how the data is currently knowing how most of the software is around our llms around ai how they are currently in this moment and their direction do you think that we'll ever really own our own data yeah it's a it's a really tough question because we i know that obviously because we have very strict like data privacy right where you like you don't remember when you used to go to the early internet and you could actually go to a website and it'd say accept cookies or deny it right yeah so it's okay we have that deny button right and i know that it's probably gonna cost It's just money, right?

8:37But at the same time, we would rather die morally right than morally wrong. And so, like, there is hope to it. The thing is, is that we need to be championing, as, like, people, we need to be championing things that are actually in the interest of us, not against us. And that's the issue, is that it's great that we're using all these frontier models. It's great that we're using all these cool tools. but are those tools being built for you or against you and clino is built for people not against people that's all we you don't see us having anything in there that's going to be putting you out of a job or anything like that we have very simple things we do have some advanced things we even actually have it where there's what we call clino citizens so they're actually because clino is a system within itself it actually has citizens that you control actually so when it's on your own computer you can be like hey well let me make sure i don't have this okay it's not open sorry i have to be careful because i've created it where it actually can listen to me where i can go hey orion open up this app on my computer and then it will open the app on your computer and then navigate to it for you so i have to be careful not to you're in your early 20s you're building at night what's that grind like it's it's very excruciating i'm gonna be honest it's very tiring because there is there's so much to it where it's like oh my gosh i should add this oh my gosh i need to make sure this is working oh my gosh i need to do this so it's it's a very there's a lot of burnout i'm going to be honest but the mission the philosophy is that ai humanism ai humanism ai humanism it needs to be together with people right ai shouldn't get ahead of people or in humans humans shouldn't get too far ahead of ai right we need to be like synchronized so it's it's a tough grime um actually funny enough uh i went to sleep almost two hours ago so i've been up since 8 a.m yesterday because i worked 10 hours a day and then the moment my shift's done i'm straight to working on the client now so you didn't go to an ivy league school you didn't go to some elite ai lab yet do you feel like an outsider very much so yeah very much so all the people i've talked to within the ai like community in space they're they're all like they all have machine learning backgrounds data science like and you know me being me you know in high school I just did cybersecurity and like IT.

11:16And so the moment I was done with that, I was like into automations. I was like, hmm, how can I make this more efficient? Right. That's that's always like how my brain always works is how can I make this more efficient? How can I do this more efficient? And so it's weird because I will present and I will show off the app. I'm like, hey, we're building this. And they'll be like, oh, that's that's cute. That's cute. Yeah, cool, cool. And it's almost like here's a Gen Z term for you. They little boy me, basically. essentially like being like oh that's cool robin like you know or boy wonder for those who know that that's what actually batman called robin boy wonder like that's what they do to me basically oh yeah that's really cool whatever like but what's its actual use case you know like how are you gonna make me from it it's like well it's not really the end goal right i haven't heard that phrase yet so thank you i learned something new today i appreciate that you believe that the future isn't one giant AI model, which you were alluding to before, but many working together.

12:16Why do you think everyone else is getting it wrong? Because the issue is that, for example, when, and this is what I've learned actually, being the outsider within AI, most AI is trained on very particular sets of data, right? And because it's trained on these particular sets of data the thing is that for example a country like brazil has certain cultures in certain ways that it speaks right that their ai would need to be trained for right and so if you only have one system right which you know i'm a very big uh i'm very into like dystopians in like 1984 you know fair night 451 like those are my favorite genres basically um not because i want to like enact that or anything no it's like it serves as a good warning of what happens when you have one system controlling everything basically that's not democracy yeah exactly terminator yeah exactly so it's like that's not democracy right and that's not ai humanism like that's not our philosophy so it's you need to have multiple models you need to have multiple systems working together right almost like a kind of how like our computers work it's an operating system that connects you to multiple different other systems and tools.

13:33You don't want just one because it defeats the purpose, essentially. Have you ever seen the Time Machine movie? I don't think I have. I've seen Hot Tip Time Machine. I don't think I was supposed to watch it as a kid, but I did see that one. That is a good one. I wasn't referencing that one, but that is a good one. You should watch this movie called Time Machine. I don't know if it was Orson Welles I can't remember who wrote it, but you should watch it though. Cause it's very dystopian though. It reminds me of social media in the future, but they weren't referencing social media. Right. But you'll see, you could relate it to how, you know, we, how you'll see, I don't want, I don't want to spoil it.

14:18I think it'll be interesting. What do you think is misunderstood around LLMs and AI currently? Because I think if you right now ask most people about what's an AI company, they'd be like, you know, ChatGPT or whatever. But I don't know if they really even understand what ChatGPT is. Yeah. So I think something that most people have misconstrued when it comes to AI is that it's this like this crazy machine that basically knows everything. Right. Like in reality, though, it's that it's the best way to put it is you ask it a question. question and then what it does is it will go and then find all these different keywords is the this is the layman's way of explaining it find all these different keywords to then find you their correct answer basically so it's like it's essentially like you back in the day having to go to like a dictionary and like you're like what is there's a word that goes with this word right and you go and look at it you're like peculiar peculiar and then you look at it's like oh here's send them forward and then it goes on and on and on most of them don't realize that ai and specifically a certain ai companies there's ai learning labs essentially companies that are all research purpose basically so they're using this as ways to train data on specific things like how one guy quite literally cured his dog's cancer using chadgbt right so it's like and this is happening in australia by the way for those who want to follow the story that like that's ways that AI is solving things.

15:54And it's not like this bad little machine that's like Terminator, right? Funny enough, right? It's not like just a Terminator. It's getting there, I will say. It's getting there. But it's not like a Terminator. There's research labs. There's the frontier ones that you've obviously seen, right? Which is the Chajipiti and all that. There is also ones that are just implementation-based where their only job is they're using AI within certain use cases, like the medical field, right? How they used AI, right, in a particular, its own AI model to train medical data to then figure out like blood and like DNA, basically.

16:32So there's multiple different actual use cases of AI, but within that, because it's such a broad term, right? There's actual like specific, like specific, specificalities, that's the word, like that you actually have basically with an AI. I'm hoping that AI can solve poverty because I feel like it can. I feel like if we all work together right now as a world, we could already solve so many problems. That's something that I hope AI will solve so no one ever goes hungry. What is a big thing that you hope AI can solve? I honestly hope AI can solve the the almost like information gap that people have and i mean it already solely is doing that why do you think like consulting companies are so far down right now because there's no need to go buy or hire somebody you know for ten thousand or a hundred thousand dollars for them to consult with you when you go to an ai model ask the ai model oh hey i need help doing x y and z how can I implement this?

17:39And it just does it. You go, oh, this is how you want to do it. And not only that it tailors it to you, which is an issue within itself, because it's a yes man. But I hope that AI can just, it solves the information gap. I hope it makes people more efficient. That's what I hope. It makes things efficient. Because currently, no offense to the older people, I love you guys, but we have dinosaurs running our world. Sorry. And dinosaurs meaning old people. Would you say someone in their 40s is old? Yeah. Okay. For all the people that are like, oh, my gosh, I'm in my 40s or I'm in my 50s. I could be your dad or your grandpa.

18:18Well, stop it because you're not old. You're not old until you hit 70. Once you hit 70 and above, you're old. How old old kid? I am. 30? Nice. That's why I wanted to have you on the show today. Did I know that? 30? It's pretty close, man. Pretty close. Thank you. That's good. I'm actually 42, so I appreciate that. Oh, cool. I'm getting up there. Hey, I am too. I'm 23, half of 46, you know? I'm getting up there in age. You got an old soul. Yeah. You have an old soul. The kids stopped calling me Unc. They started calling me Pops now, so. Oh, Pops, nice. You got it. See? Some people just have an old soul, and I appreciate that.

18:59Have you done any research on Quantum? Not much. because it's so like a very closely guarded thing when it comes to quantumry and the actual implication of it. But from what I know that it is a, it is going to be one of the most scariest things, I think. Because if you're afraid of AI, what you really should be concerned about is how we use quantum. Because instead of us going from numbers, basically we're going straight to, like it's just straight chemistry. So it's literally, does this atom exist at the same time that this other atom exists? and if they do boom it's instant meaning that all encryption systems wiped out immediately basically so everybody you know like missed us right fable the model that like you know everybody only got to have for three days i don't i used it for 10 minutes it was nice yeah i i used it for the three days that i had it it was like man what so great right imagine that but on insane steroids basically where hey i need you to crack into this system in south korea basically and then it would just it was boom it would just do it because there is no in quantum there is no friction there is no limitation the same way that we have limitations with ai where because because it runs off gpus and hardware right that doesn't exist with quantum and that's something that's really scary because How do we regulate something like that?

20:25We had a guest on before talking about they're trying to solve or they're solving the problem around quantum plus AI hacks. This is like over a year ago when they came on. And now I'm starting to see a lot of things coming out. It's quite scary. But hopefully if more people like Jamari, like yourself, are building more things, there's more representation of people building. The things that we all need will be in a better place. but I appreciate you joining Kleino AI. Hope everyone checks it out. And thanks for being on today. Thank you. I was glad to be on.