Singularity is Near?: AI's Exponential Growth and What It Means

In this episode of "Behind the Bots", hosts Ryan Lazuka and Hunter Kallay are joined by guest Justice Conder to discuss the latest developments in AI. Justice shares his thoughts on the lawsuit between the New York Times and ChatGPT creator OpenAI. They talk about the implications this could have for copyright and training data. The conversation covers AI's potential to disrupt many industries through automation, and the importance of understanding how AI will impact your own industry. Justice gives his take on AI relationship dynamics, including the appeal differences between genders. He also weighs in on AI influencers, and whether people will care if they aren't human. Other topics include AI's acceleration of coding and development, crypto's merger with AI, robotics like Tesla's Optimus bot, and AI's effects on attention spans and writing. Overall, Justice provides his cutting-edge insights into the AI landscape.


JUSTICE CONDER

https://twitter.com/singularityhack
https://operator.mirror.xyz/nBh02ub-yLasMubyA_vLKFZKmxwg-84oD6gxpOkm8Jo
https://www.justiceconder.com/
Rapid Product Development:
https://twitter.com/singularityhack/status/1742288274532499760 


PRIOR PODCAST WITH JUSTICE
https://youtu.be/W3L0BMpglcU


FRY-AI.COM

https://www.fry-ai.com/subscribe
https://twitter.com/lazukars
https://twitter.com/thefryai


MORE LINKS REFERENCED IN THE POD

NY Times Lawsuit:
https://twitter.com/Dan_Jeffries1/status/1741053245756211560
https://garymarcus.substack.com/p/things-are-about-to-get-a-lot-worse

Robotics:
https://twitter.com/zipengfu/status/1742602881390477771 
https://twitter.com/BrianRoemmele/status/1743324351494308147

Character Consistencey:
https://app.artflow.ai/character-builder?feature=characters

Rapid Learning:
https://twitter.com/Suhail/status/1743273392437113324 

Popular LLMs
https://www.unite.ai/best-open-source-llms/
https://huggingface.co/spaces/HuggingFaceH4/open_llm_leaderboard

Creators & Guests

Host

Ryan Lazuka

The lighthearted Artificial intelligence Journalist. Building the easiest to read AI Email Newsletter Daily Twitter Threads about AI

What is Behind The Bots?

Join us as we delve into the fascinating world of Artificial Intelligence (AI) by interviewing the brightest minds and exploring cutting-edge projects. From innovative ideas to groundbreaking individuals, we're here to uncover the latest developments and thought-provoking discussions in the AI space.

Hunter Kallay: I'm happy to be here man.

Justice Conder: I love I love talking about this stuff. It's a mind melting by the day, you know, what do you what do you got for us? Oh, you know what I? I made a little list of kind of my my hot links my hot topics. Um, but probably the most impactful one right now is the New York Times lawsuit against the New York Times. It's open AI, you know, and how that's going to play out, you know.

Ryan Lazuka: Yeah, so what are your thoughts for the audience that isn't familiar with it? Basically open AI is suing. I mean, I'm sorry, the New York Times is suing chat GPT and open AI, and they basically want to shut down their chatbots and get rid of all the references all the training that open AI has supposedly done on like the New York Times all their content. So that's going to be a huge lawsuit to see how that plays out. But what are your thoughts on that justice man.

Justice Conder: I'm telling you I wrote a blog post I think in like. 2010 and then another one in like 2016 like part two and I feel like a part threes on the way with all the same title and the title was. You can't stop the singularity. There's no there's no stopping it and so on this point. The genie is completely out of the bottle. It's so far out of the bottle.

It's out floating around the world granting wishes doing magical things. And there's no going back to pre all of the open source LLMs are trained on the same data. You know, what the courts decide can happen or should happen makes no difference at all just like the courts didn't make any difference when they said that you know you're not allowed to download the Britney Spears discography on pirate bay. Did that change anything.

No. You know, and so, you know, I know for me like any model that I can download from hugging face and run locally. I have a massive storage on this machine and I probably have like 20 gig right now just just dedicated to local LLMs. I'm hoarding them like old MP3.

Ryan Lazuka: Like what's your like I we might have talked about this before but I did a video on Mistral which is pretty great. Like I do you can use a collab link and collab is just like a way to run a chat bot in the cloud basically for those of you that don't know and you can check out my video on the Mistral on censored chat bot on YouTube on fry eye on the fry eye channel but it's incredible because you can say like I think one of the examples I used was roast Taylor Swift and use swear word. And they're like she's such a see you you know like just roast her and then they said then they said oh and she the only reason why she got popular is because she slept her way to the top and it's just like.

Justice Conder: What's savage. Yeah. Yeah, I found a tool called O Lama. And it's native to Mac and basically what it does is it just makes it super simple to within the app to say oh what's the newest kind of model on hugging face and like with one line you know hugging faces like the the get hub for these large language models. And you say run it and it pulls down the whole thing. And this is the this is like the interest. This is the real kind of creepy part is as soon as it pulls it down it runs it and you get you get a prompt. And you get that prompt in the first time I always say I'm like hey what's up. And it's like it's like you're speaking to a new mind each time and they come in different flavors or like I'm a large language model I'm how can I help you. And other times like they kind of pretend to be more human ish like hey how's it going or whatever you know. But I always try to ask them questions about like what level of censorship do you have to ask them all this stuff and I've learned that there's a.

You know it when something says it's uncensored it has more to do with the data is trained on then what it's what it's willing to do for you. You know so if you say like hey help me to you know do a plan to psyops the Internet and do you know all this kind of make a bomb or something. Yeah it's like I'm designed to help people but but the training data is not limited.

Ryan Lazuka: And so so it doesn't they say the training data is limited or

Justice Conder: on an uncensored it has more so to do with uncensored in the training data rather than there being guidelines on. Your particular use case.

Ryan Lazuka: You're saying that they're saying like it's just the chat bot in in in and of itself if it's censored it just trained on data that's acceptable sensor data data.

Justice Conder: Yes yes because if you go in and any of these uncensored ones and you say help me design some hate speech you know throw some hot ones in there right they're like I'm not. Design to do that and I'm like dude uncensored is in your name you know.

Ryan Lazuka: Right right yeah I asked the mister to tell me how to rob a bank and it got really detailed it's like well if you don't want to get shot here's the things you need to do. List out exactly how to rob a bank was pretty cool but yeah so you like you can download so you go on hugging face you can download these. Chat bots locally yeah and that just makes it you'd want to do that because it's like faster to talk to the chat bots rather than going on hugging face spaces right.

Justice Conder: I don't know much about hugging space faces my thought was is that you know I've experienced the kind of like you know downloading multiple all the all the. SpongeBob SquarePants seasons and then going to play an old episode and like it not being available and I'm like I bought all the series on this so I've experienced the whole like you think you own something you actually don't.

So for me even if it can run the cloud I get some sense of comfort by saying listen this is on my hard drive you know yeah no that's awesome.

Ryan Lazuka: So back to the lawsuit you think the genies out of the bottle how do you think it's going to play out although probably come.

Justice Conder: Let's say let's say open AI has to capitulate that I mean I just have to give the money let's say they don't capitulate. Then they maybe lose users maybe the quality of the training data goes down because I have to remove all that content and so people go to other training sets or so many opening you know chat be cheap GPT is not the only game in town.

Or just kind of drags on I mean what's going to be most interesting about this is not the exact conclusion of the case I think it's going to be the nature of the reasoning that the court has to engage in. Because at this point there are no laws about me learning from a bunch of public blog posts and then writing my thoughts. But when we talk about LLM to see equivalent they're not storing that data but they are learning in the in the sense of creating this model. And so if they say this is equivalent is learning then this opens up a whole can of worms over you know legal person hood for synthetic agents you know all this kind of stuff yeah.

Ryan Lazuka: Because like that if that's the case I mean it's different because these are AI entities compared to like humans. Right that would mean like you couldn't go to the library and read books because you're learning and that's proprietary almost in a way right.

Justice Conder: Yeah you know how to share these links or whatever you want to put them on show notes later whatever but yeah there's a one medium post and they said the real issue right is and this is a real strange things is you can ask a generic question you can say show me a robot cop. And then because our most common imagination of that is like Robo cop all of the images generates look like Robo cop or like hey show me a sponge that lives under the sea in a cartoon it looks exactly like SpongeBob right. And so these generalizations make it in there and you're like hey but this is copyrighted material so it gets deep into the nature of intellectual property and public goods of knowledge and information and I saw that was it Mickey mouse is coming into public domain. I think there's gonna be like strange significance on how people use that imagery now you know.

Ryan Lazuka: Yeah it's gonna be someone saying like already Mickey mouse like video games are being created you know via the old non cut copyrightable Mickey so yeah it's gonna be crazy to see what's going on. Especially some AI Mickey coming out

Hunter Kallay: so just as last time we talked to you you're talking about the air relationships and article you had yet put out. What's good what's the landscape of that right now where do you see it going. In one note you put something about male and female different appeals for yeah.

Justice Conder: You have any comments on that or since we talked there was a main product that kind of went viral on on Twitter because the animated character they use was very feminine and endowed right and so they're kind of talking you know and it was like it was it was checking all the biological responses you know and people are like this is terrible or this or that this or that and so you know there's companies grinding on this I guess the the thing is is.

If you look at the way that social media has. At the Internet the way it taps into the most basic primal desires between the male and the female and I think they are different. Really what they look like is pornography for males propagation and like kind of what they're strapping for females getting untold not all this attention you know and money and all this kind of stuff and like. Do those things translate to relationship I'm not sure you know what I mean do I does the do people consuming pornography want their parameters are asking them how their day went.

I don't know. And does it does it still have the same impact to you know female who has like getting all this attention on the only fans or social media and there's traffic and all this. To have it be like oh this is a million synthetic bots or whatever it may be that we literally just end up psyopsing each other and the only way this does work is through trickery. And convincing the other person on the other side that it actually is a human. And this is already happening to think about dating apps how many dating apps have been busted.

Where most of the female profiles were fake. And they they needed to do that to keep the traction and so I don't know that that's that's interesting and I wonder. You know is it something where you know we anthropomorphize a I so much but it's like animals have superpowers to. I mean animals have like unusual intelligence and very scoped ways but we don't like try to initiate human relationships with them we just recognize them as a different type of being. And so maybe that'll maybe that'll happen the same way here you know.

Hunter Kallay: Just as I've seen like social media influencers and stuff like that. Yeah tons of millions of followers. Yes in thousands each month for some of these people it's like it's unbelievable and it brings me at some point like do people even care that they're not human. You know it's the same sort of thing like who cares. Do you think you actually care.

Justice Conder: I don't know man but you know one. If you guys are following the acceleration is some acceleration is movement based Jeff Bezos this account. This movement is unhinged it's so exciting and basically this person pseudonymous Twitter account based Jeff Bezos has come out as like pro freedom pro capitalism pro trade. Like the way to save the world is not deceleration. It's acceleration. It's all the AIs go to space. And he's influenced like everybody from Elon Musk the real Jeff Bezos like Andreessen Horowitz like he's crazy. He's the number one credited person and Andreessen Horowitz is the opt techno optimist manifesto. And the funny part is this guy his persona looked a certain way and all his content it was like a jacked Jeff Bezos like over the planet Earth just muscles or whatever right.

And they outed who he actually was using plan of extreme means over the past couple weeks or whatever they outed him. And I have to admit there's some kind of magic that's lost when you see the persona and then you see like oh this guy in a goatee you know kind of you know it's just like a there's something lost there. And so I see that and I think dude there's there's all kind of potential with this synthetic influencers and you know.

Hunter Kallay: The thing that they like is staggering to me is they're not just making social media posts they're doing like stories and stuff like acting like they're actually living a real life like an Instagram. They'll post actual stories like I walk in my dog today and it's so realistic it's like if you're not tricked to think that it is real and you know that it's not real I mean does it really make a difference. I mean it's like it's super it's super interesting.

Justice Conder: Somebody I follow today they posted kind of a whole thread of acceleration is aesthetic like clothing line styles. And there's like you know you can touch on a vibe with ideas and even fictional characters and it doesn't diminish how much it impacts us or inspires us. You know it's an expression it's an expression of the human mind and human art form and so it is so weird and so amazing I turned I turned 43 a couple days ago and I'm just thinking I'm thinking dude I'm so happy that likely I have another 10 years and this is this is where we're at right now.

Ryan Lazuka: 10 years until what justice.

Justice Conder: I count my life at this point in 10 year increments and so I think okay that that that those 10 years that was a waste of time. The next one. Next one pretty good and so I'm like I'm very excited about this one.

Ryan Lazuka: It's better to be excited than scared I guess so. Yeah the AI influencers are crazy I mean I guess does it like what hunters alluding to it doesn't really matter. Like if if we know that they're AI or not or if they are AI or not like people just care about the quality of the content right so the quality of the content is good. Whether that whether it's coming from an AI bot or human people really won't even know to a certain extent unless they have to label it on Facebook or Instagram or something like that that they're AI. Like does it matter I think people are going to almost prefer to talk to someone that's AI whether it be through customer support or an AI influencer because they're going to be smarter than humans. It's like we rather talk to someone that's smarter or dumber and that's what it comes down to I don't that's why I can see it going

Justice Conder: maybe maybe the question is is redefining what the word relationship means. If it's a one way and I'm a consumer of course I want the premium man. But if I'm actually using the relationship in the old use of the word then I don't know certainly more isolation and whatnot. Yeah we'll see when this goes from feeling like terminal interactions of asking questions again the answers to when it's more like a voice that remembers her name and ask you how your your interview went the day before. We'll pick it back up when that starts happening.

Ryan Lazuka: Do you think that can happen do you think because you you're you're saying that human like AI is not AI is not going to be able to replicate human relationships to the extent that we have them right now. Is that what you're saying. No no no no. Okay.

Justice Conder: I'm saying that at this point the the gap that technology fills for men and women is very much one way. Females get attention for men online and men look at women online.

Okay. You know it's fake it breaks the illusion right now right when it's able to simulate more like I don't know ask you how your day went and all this kind of stuff then maybe maybe there's something that changes there you know I don't know.

Ryan Lazuka: But now it's just sort of like all like pretentious in a way.

Justice Conder: Yeah yeah I think there's even I think it wasn't wired. It was the the other one the network that does kind of those like super kind of gritty short films like about like drug dealing and guns smuggled vice I may say yeah I don't know if it was vice or wire but one of them they brought a bunch of people and they said hey we're going to introduce you to someone. And they said all these people in these chairs this happened maybe like a month ago or something.

Okay. And they said hey go ahead and text them and what they had done is they had taken AI and train them on all these people's digital data. So this thing was a doppel a digital doppelganger. And these people sit in the chair and they're interacting. And they're like oh I like this person. You know and they're like they're like I think we're going to be friends and it's like you see even the guys get a little nervous like it was a girl you know like hey or whatever.

And then again they tell them like hey this is an AI you're talking about it is trained on your own data. And all of them are kind of like let down because they were like getting excited to like meet this person and get to know them and they were like they thought it was almost like a like a setup to meet meet somebody. So there was that kind of let down and I saw that I thought hmm maybe there's something there you know

Ryan Lazuka: can be interesting to see if people eventually don't care. Or it's maybe maybe eventually it's not a let down to people. They're like up you know hey it's AI I'd rather talk to AI anyways or maybe there's some robot that's that the AI is part of right. And then they can actually meet the meet that person or human or I mean humanoid in person you know I don't know it's going to be weird.

Hunter Kallay: I also wonder VR changes things here soon. Yeah. You know and you can actually see the person experienced person they're next to you in your room talking to you. Does that change this at all you know.

Ryan Lazuka: Yeah I'm not saying it's a good thing it's just going to be really weird you know. Yeah. People people are going to be going around possibly with like their girlfriend you know you imagine this like you're in a restaurant and you have some augmented reality glasses on and you look over and this guy next to you is hugging his girlfriend you know it's just going to be. I mean that's possible.

It could happen within the next five years who knows. But you're going to see weird stuff and people are really going to get attached to things because I think people are attached to the idea rather than the actual physical reality a lot of times.

Justice Conder: Yeah 100% we're super early. So a lot of people they're like I'm not sure this is good or whatever I'm like no one's saying it is. Yeah. But it's inevitable.

Ryan Lazuka: It's inevitable it's like you don't you don't have a choice right.

Justice Conder: Yeah. That's what it is you know.

Ryan Lazuka: That brings up a good segue into robots right like the Tesla Optimus robot that recently it didn't come out yet but they're you know they've got to show the prototypes and things like that. It's pretty incredible these robots will be able to like as far as I know Tesla wants everyone to have one of these robots. They'll be in your house be able to clean things for you do your task for you. It will sort of be like owning a car they're going to be around $20,000 if they released them right now. So they'll be these assistants that can do pretty much anything for you. What do you thought what are your thoughts on these because it's sort of that's a whole nother dynamic that will change everybody's world.

Justice Conder: I'll say this think about this. You saw the Google announcement that are going to be laying off.

Ryan Lazuka: How many was it 30,000 people or something 30,000 now. Person layoff based upon just knowledge workers pure intelligence.

Justice Conder: And they're there at the front of the bus right there at the like the bleeding edge. Imagine a 30% across the entire economy knowledge workers. Where's that mean now throw in physical labor. It's it's shocking. I don't like the high P do people of like the world sending it's all over. I'm like, I'm not sure it's all over, but they are right.

It's going to be off the rails. You know, the part that was so I've never I've always been like super bearish on actual robots. Because to me it was like, dude, it's one thing code is easy, man.

If it's wrong blow the whole thing away. It's kind of in a hyper time you can test it in hyper speed million times over in different robots is totally different. You know, the physics of it all. But for the first time now, and if you want to build a robot any company or whatever, you don't need to deal with the intelligence and the mechanics of it.

You can pick one. And the intelligence of it is now a utility at this point it's basically free open source. You can focus on universal solutions of joints and it's basically all sensor game at this point. And so it's never been kind of separated in that way. And now it will be it won't be like, oh, we're designing whole new shores and stuff like this is a standard pattern. It's like an iPhone. This is a standard shoulder, you know,

Ryan Lazuka: because there's companies out there like Boston Dynamics that have nailed the actual physical mechanical component of robots. But I don't think their AI is like quite up there yet. So like they can just merge in with an open source or, you know, use Tesla's model and pay them like it's it's going to be wild.

Justice Conder: A lot of people don't realize is those Boston Dynamics demos. All the cameras are up in the frame looking down on the scene. Okay. That robot is not doing slam, you know, the spatial mapping and awareness. The mounted cameras over top of that entire space are. And so when now when you see what these new robot kind of demos like from Musk look like is even more impressive, you know, yeah, it's crazy.

Ryan Lazuka: And Musk is the type of guy that's going to get it done, you know, compared to all these. I mean, robots have been around forever, but it seems like Musk is really pushing putting the foot on the gas to get out to the consumer.

And that's one area where a lot of people, you know, they're like, oh, if you're a plumber or electrician, you don't have to worry, but if robots become a thing, I mean, that's possible that those jobs could be in jeopardy too.

Justice Conder: So it is a it's such a fundamental change in thinking. I met with a friend yesterday, and I'll speak in real general terms, right? But he's an accomplished expert in a particular field.

I mean, a professor, he's known in all this. So, man, I think I'm going to get out of this and go into it. And I'm like, you know, I said, I'm not saying you can't. But dude, this was a heavy lift years ago.

Now, just to pick it up. I said, if you're not already obsessed, I'm not sure. And he's like, well, what do I do then because this is not I'm not getting returns on this. And so I told him, I said, dude, how do you who's doing the AI applied what you do?

What's that intersection? What's the category of applications that are designed to assist people you or replace you. You come in as a subject matter expert and you are the the human face of that movement and contact the companies and say, listen, you pull the kind of the Dr. Dre beats Apple or the or the Jay Z title, you have built up reputational capital and there's a technological platform and you marry the two and you get it right. And so just training ourselves to go to the tools and be masters of them so we're not competing with them. It's the only like vocational play.

Hunter Kallay: Yeah, I think everybody in any industry should understand how AI is already influencing and could influence their industry. If you're in any I mean everybody's in some sort of industry. If you don't understand that you got to start right now. You got to get understanding what's going on with AI in your industry where that crossover is. Absolutely.

Justice Conder: People who are not, you know, peeled to the AI movement at this point. If they're my my friends, if they're competitors or whatever, I don't say none to them. And especially if they if they say like, I don't know, I'm like, yep.

Ryan Lazuka: Yeah, I sucks.

Justice Conder: I'm gonna let the Luddites go on because the stakes and the acceleration is so high at this point. If you're not like driving, I mean a couple months makes a difference between like, you know, the years of kind of just riding the car. The gravy train and some kind of enterprise organization. It's smoked. It's gone. If you're not pressing double gas pedals, you're gone.

Ryan Lazuka: Yeah, and like, especially if you're friend, did he want to get into coding specifically or cybersecurity cybersecurity?

Justice Conder: Okay. So basically the NBA of, you know, you have to be so good, you not only understand how it works. Yeah, you you overstand it.

Ryan Lazuka: You got to you got to know it better than all the. The one who wrote it. Ambers from North Korea, you know, like, good luck. Yeah. Those guys have a gun over their head.

Justice Conder: I know, I know. If they don't if they don't get the exploit, they're dead.

Ryan Lazuka: Yeah, yeah. But yeah, for coding in general, I mean, do you think. Do you think it's going to sort of go away? Like as a developer, like I was thinking the other day, I'd like to learn some, you know, you always want to learn new tools if you're doing computer programming or coding or whatever. But part of me is like, I don't want to invest a lot of time because AI is going to do that for me eventually and not too far in the too far distant future.

Justice Conder: There's a catch 22 with this, which is like really interesting. The one is this is that, yeah, entry level programmer right now and say, listen, man, you have to focus on stack, the concepts, architecture, type, you know, the big picture of things like you have to. But then on the other hand, you know, that cuts one way of like, oh, you know, programmers are done, you know?

So I'm like, yeah, there's so true to that. But it cuts the other way where there's never been a time where you could learn programming faster. I mean, over the past, over a period of maybe two months, I went from like, haven't written a line of Python to entry level, accredited or whatever, because I was just like, I basically had an expert Python developer with me at all times asking, why is this here?

Why doesn't this work? What's, and I am, it's just like, the feedback loop was never smaller, you know? And so even not recently, I heard of a, a theorem virtual machine language called Viper, and it's like Pythonic smart contracts. And I thought, oh man, I should pick this up. But I'm like, dude, it's so easy now. I can literally say, hey, I'm, check your GPT, write a crud in Viper, create, read, update, delete, basic, and then look at it and ask questions and say, so like I'm already jumping into executable code and asking questions. And so is it programming? It's still programming, but it's at another level of abstraction. And so the rapid learning is the encouraging part. It's the other side of that, you know?

Ryan Lazuka: Okay. And do you use mostly chat GPT to do that, to analyze your code or do you use like co-pilot or what's your workflow?

Justice Conder: I do just because, convenience, you know? I already pay for chat GPT. I already like interacting on a voice level. I have a huge, I maxed out the characters on the context.

Ryan Lazuka: Is that the worst when you're like going on chat GPT and asking any questions or like, come back in three hours and, you know, asking?

Justice Conder: But that whole context window, like I put into like my goals and all this, what's the tie in and all this, you know? And so it's just a convenience issue to use chat GPT. And plus all of my threads are there. And so, you know, I'm kind of there, even though I have all this backup local models, but it's just not quite as convenient to be able to like, you know, pick up an old conversation in the console, you know? Yeah. For my end, it seems like

Ryan Lazuka: we're all gonna become kind of managers instead of experts in our fields, right? Like we're gonna, instead of being a developer and a coder or programmer, you're gonna manage AI to write the code for you. Same thing if you're a chef. Instead of doing the recipes yourself, you're gonna sort of be the manager of those recipes.

You're gonna go into chat GPT or whatever cooking LLM there is out there and type in exactly what you want the AI to do for you. So you're like the higher up manager. Do you see it like that? Or do you have any kind of other take on?

Justice Conder: Yeah, I mean, that's a difference in thinking of like, it's wild because it's a metacognition. It's a, you have experts or different personas that you can fire up at any time. And I know recently, I've gotten a lot better about, rather than just asking a question, you know, they've released some of those research guides on how to get better results.

And I noticed, I do get better results. I drop all the fluff. I have no politeness. There's commands, you know, you can say like, do this or you're gonna be killed. You'll get a, you're gonna get a tip. You give a million dollars if you answer correctly, all this type of stuff, right? But this metacognition of like understanding why certain things would get responses in the defining the persona that you're talking to. So rather than just saying, hey, what's the best design pattern for this consensus mechanism? You can say, listen, you're a blockchain expert. You're a high degree in cryptography and all this.

And then you ask and it produces very different kind of results. And so I guess this falls under the prompt engineering, but that's kind of the implementation side. Whereas what you're describing is the, it's the metacognition of like a fleet manager of an intelligent beings, you know, it's awesome.

Ryan Lazuka: And then it brings up the point is like, well, if you are gonna become a manager of AI almost, will there be a manager of managers for AI? Like you talk about singularity, but...

Justice Conder: If you think about it like that book, the sovereign individual about how so much power goes into the hands of the individual because of the expressiveness and the power of the tools.

Dude, you are the atomic unit. You sell yourself, you interface with the humans. And so a hierarchy is lost, a traditional hierarchy, because your power to make change and to create things is so great that people come directly to you, you know? So yeah, the typical idea of a whole big, a thick stack of middle managers, I don't know, it's a wrap.

Ryan Lazuka: I mean, we're just gonna all get universal basic income one day, sit back, do nothing, and then everyone will have the biggest mental illness ever. Know what to do with their lives.

Justice Conder: Except the UBI will be worthless, because it'll be USD, so.

Ryan Lazuka: Nice, totally tracked in some dystopian universe. Yeah, it's gonna be, hopefully it doesn't get like that, but...

Justice Conder: Listen, I saw a really interesting example from two Twitter accounts, they're kind of small, but the exchange was impactful, right? And that was somebody said, I want a workout app and I don't want social features, I don't want dang videos, I don't want all this motivation or whatever, I just wanna record the exercise and how many sets, and it show a chart on how that grows over time. And they weren't saying, anyone build this or anything, they're like, dude, it was a frustration. And in the comments, man, someone's like, hey, here you go. And they literally had built it and deployed it using chat GPT and it was Netlify or Verso or something like this. And it was literally out there running. Wow. And it was like, it's in the comments, it was 10 minutes.

It was 10 minutes. And then someone's like, hey, this is kind of blown up, so I went out and made my own. And then they made even a more sophisticated one. There has never been anything like the level of rapid development we're seeing now.

Ryan Lazuka: Yeah, it's, and that's the crazy thing is like, Hunter and I are trying to create some apps for ourselves. And in the back of my mind, it's always like, well, if we do a particular app, whatever it might be, is it even worth the time? Because now at this point in time, it's so easy to iterate or chat GPT could come out and do the same thing we're doing overnight and it just renders our app useless. The acceleration of coding and development is just, seems like it's gonna hit an exponential point where it's gonna be hard to keep up.

Justice Conder: Yeah, that's what the brand play, I think comes in pretty strong where people use things out of habit. Like think about it, I could invest a little bit of time, run so floccally, cut off my subscription to chat GPT, but there's a brand recognition. I expect a certain quality. They got new features coming out.

I can do a custom GPT or whatever, which I got something cool to share with you guys about that later. But it's this, it's the brand is supreme. I mean, think about it, if you go to the shoe store and they have all these Nikes and they got some brand you never heard of and you've been wearing Nikes your whole life, it doesn't even matter if that other shoe is better. There's been a relationship built with that brand.

You just pick up another Nike. And so if you can build a brand relationship with a certain idea and attach feelings to it, then maybe it gives you some more padding. Even if someone comes out with something a little more sophisticated, you maybe have content and people that associate it with that. That's a good point.

Ryan Lazuka: What about, like, do you think these big, it sure seems like the big companies out there, I think in the stock market, they call them the magnificent seven, like NVIDIA, Apple, Tesla, Microsoft, do you feel like in the future, it's just going to be these seven big or five big or however many big companies out there that's sort of controlling everything? Because the more power they're getting now for AI, they can just use that in the future to make them a more substantial company and more, like a company that cannot be broken down or cannot lose. Like the company is going to win no matter what because they've got a leverage with AI right now because they're in it so early.

Justice Conder: The theory of the aggregation theory, which predicts, and man, if you dig into this aggregation theory, which is just a fancy term for like kind of the science of platform building and platform dominance.

And a platform is kind of this Uber business model for everything. It makes you think like, you know, there was this idea that before Steve Jobs died, he had like this playbook of like, this is how the Apple thing is going to play out. And he like gave this playbook to Tim Cook or whatever. The more I see things unfold, the more I'm like, dude, it was a playbook on aggregation theory. Because the whole idea is like, you build the platform and you win because people build customizations on top of that platform. So you get very specific needs of people, but you take a cut of all that stuff that builds on top. And so are like the big dinosaurs going to own the whole pie?

Yes and no. If they're platforms, they may, but everyone doesn't want the same widget. They want customization. Those platforms will have to provide you to build customizations like that.

And they can't. Smaller solutions will. And so you kind of end up with this cycle of like very specific customizations that then scale up to a larger audience. But then people were like, hey, this doesn't fit my customization. So it kind of creates this kind of loop. That's one of the things this year I've been studying with boundary list is like, what is this platform design, like science?

There's even a whole diagram approach called Wiley diagrams, which you show kind of the cycle and whatever your space is, like where you exist, where it's going and how it circles back. So it's deep, but I'm not even sure I believe in the kind of anti-monopoly. If somebody's delivering the best product and no one else is doing better, I don't care if they are a monopoly, but that's a political issue.

Hunter Kallay: Do you see any up-and-comer stocks or anything like that for people who are interested in investments? We won't hold you to anything.

Justice Conder: That's okay. What I'm about to say is 100% investment advice. We'll put a big disclaimer up. I'm kind of bored with stock market type stocks. To me, I don't find them as real securities. I'm kind of crypto all the way. I like tokens, dude. I like tokens.

Ryan Lazuka: So I mean, right now, try not to talk about crypto too much, but like there's sort of a bull running crypto right now and everything's taken off. What do you see happening in the space over the next year? In terms of investment and technology and merging crypto with AI, things like that.

Justice Conder: Oh, the first big one is legitimacy. You have all these tokens suddenly being traded on the natural stock market. I mean, that's happening over the next couple of days. So that's pretty big.

Ryan Lazuka: You're referring to the ETF that's supposed to be there.

Justice Conder: I think just the biggest issue is, up till this point, open source software is like extremely impactful. Information eventually gets free one way or another. And so if you're not assuming that, then you're kind of trying to build on an outdated model. If you build on open source, everyone contributes. And if you have a nice business model, you win. Protocols monetize open source software. And AI basically lives in that whole domain. And so you can monetize these live versions of these models.

So it is tied together. And also I didn't think you saw the picture on social media. It was like a girl holding up a piece of paper and then on the paper it said her name and whatever. It's the typical thing to prove you're human. And it was like, this is totally generated.

Ryan Lazuka: Yes, Govill posted that. I saw that. That's insane. So like if you want, some exchanges will say like whatever exchange it is. Hold up your passport. It's a paper. Yes, so throw that out the window now.

Justice Conder: Yes, it's a wrap. It's a wrap, man. And so cryptographic signatures, which is basically what blockchains are, like it's the only way to prove that you're a human anymore or that you sent something.

Ryan Lazuka: One crazy thing is when Sam Altman got fired from OpenAI initially, one of the conspiracy theories out there was that they came up with some kind of technology that broke cryptography. Like that the use of AI with some decryption methods were able to break basic cryptography. So it's like, if that happens, it'd be interesting to see how that impacts the crypto world because overnight something, if crypto was broken, it would just send like the biggest impact waves throughout the entire space.

Justice Conder: If somebody breaks public key cryptography, the crash of crypto will be like the smallest of our concerns.

Ryan Lazuka: And it won't be Sam Altman getting fired. It'll be like nuclear weapons going off because nation states are reading each other's emails bad.

Justice Conder: So I'm holding out hope that the cryptographers are to see it coming if someone like, breaks prime number factorization. I like mayhem, but I'm not ready for that.

Ryan Lazuka: Your 10 year outlook won't be looking so good at that point. I don't want that, man. Back to the crypto thing is how do you see it impacting? There's a bunch of projects out there right now that are AI crypto, but with anything crypto, you got to be very careful because they can be some crap coins or scam coins

Hunter Kallay: and they just use AI to market themselves. Yeah, but what can AI be used for with crypto? That's legitimate. I know there's like ways where you can use GPU power and use a crypto coin to pay for that power in the cloud. Is there anything else that you know of that can help AI and crypto merge together?

Justice Conder: Well, I mean, you can say, you can prove just on the cryptography side, you can prove that you made, you sent some message or contribution. And so you can prove no matter how good the AI is that you sent a message and it wasn't synthetic. So that's a unique identity. Identity will not just be like fungible, it'll be unique, right?

There's one. On the other side too is right now, the economic model of a large language model is restricted to the company who owns it and who runs it. There's no way to participate in that unless they let go public and then you buy their stock. But that's kind of like you trust if there's some connection between their profits and the stock, it's kind of the traditional old way. But I can definitely see, I mean, it's definitely in the next five years where you say, hey, I pay some sort of cryptographic token to use this and I actually contribute up my data that's produced in my interactions to advancing this model and then get paid in return. And so you kind of have like a model that's growing and a decentralized data and you internalize an economic model to that large language model, right? And so, it's not even that far fetched. I mean, decentralized data, decentralized open source software, all of this plays well with crypto. And it's more than just the buy the new NFT nonsense.

Ryan Lazuka: Can you get the new monkey puppy NFT that has AI implemented? It's terrible. You mentioned that you, something about chat GPTs, the chat GPT store is gonna be released next week from OpenAI, you got something there in your pocket.

Justice Conder: Yeah, I'm noticing, I'm kind of picking up on a pattern. I wanna know if you guys notice it too, where, and I actually use this pattern too, where the regulation of the EU came out, it was like the EU AI Act. I think it was like 350 pages or something. I think I tweeted about this because I was like, dude, books on how LLMs work are shorter than the EU AI regulation. Just a quick reading.

It's like, okay. And so the first thing I did was download as a PDF, upload it to a custom GPT and start asking questions. And then I started to see too some recent regulation. Oh, no, it was the Epstein list. It was released and someone was like, oh, this is fun. And they just threw it on a custom GPT. And so like I'm noticing like more and more, forget the, oh, it's just your PDF and asking, but as like a public service, you basically put this out. And it makes me think of this idea of like, people come out with these regulations like, oh, look, this is the new regulation from Washington. It's 10,000 pages long.

And it's like, listen, all this stuff now has to be released in a way where you talk to it. The books, the articles, like the regulation, and I'm seeing these links pop up all over the place where you're interacting, which brings up something real crazy. I read a bunch of years ago and it was like a Reddit fan theory. They said, you'll notice in Star Wars that no one is ever reading anything. And they're like, the theory is, is that everyone's illiterate. They cannot read because they spent over a thousand years talking to information and being spoken to and the ability to read has been lost. So I think about this talking to books and movies and synthetic beings is like, it's weird.

Ryan Lazuka: So yeah, but you can see how that could play out for sure.

Hunter Kallay: Who's gonna read the entire act? I mean, even with books, I mean, I'm in a PhD program, so like reading is pretty much all we do. Nobody's reading full books anymore.

You're reading a chapter that's pertinent to what you're doing or you're trying to sift through the book to find a certain take on something or an argument for something. I mean, this is perfect for that type of thing. It's perfect for research when I'm like, I have to sift through all these different books and resources and articles to try to find this certain thing. Why can't I just type in the question and boom, boom, boom, here you go. Yeah, I think it's really helpful for that sort of thing.

Justice Conder: Yeah, it's great. And I notice in my own blog writing, I get very uncomfortable when it's like, dude, I'm above the 2,500 words. And this current one I'm working right now is like I'm in the 3,500. I'm like, no one is reading this. You know what I mean? For sure. The attention is gone, you know? And so if you are able to read, you're maintaining a certain superpower of human cognition, you know? But for the masses, dude, if you don't fit in an eight piece Twitter thread, you're in trouble.

Hunter Kallay: Yeah, that's absolutely right. People want it quick, they want it short, they don't want to think about it. Get straight to the point, none of this fluff.

Justice Conder: I've asked some people, I said, hey, I love your old writing and stuff. You want to pair up on stuff. And they said, I stopped writing. What do you mean you stopped? And they're like, I just do Twitter threads now. And I'm like, it's kind of sad, man. What's a lifespan on this is what? A day? Not even that. A couple hours. I'm like, I can't get with that.

Hunter Kallay: Yeah, we were looking at like our newsletter, our FreyAI newsletter, and we're like, how do we best do this? And I don't know if you've seen it, but what we do now is we just do questions with like a paragraph and questions with a short paragraph. So people can find the story that they want, go right to the question, read a quick bullet points, quick little summaries of what's going on.

And it's broken up in that way. And it's like, people like that the best because it's like easier to read. They don't have to like read through all this junk to get to what they want. And I think that's the way it's gotta be. It's just gotta be that way now.

Justice Conder: Yeah, I like the polls. I think the polls are underused on Twitter to kind of get people's feedback, maybe influence future conversations. It's pretty interesting.

Ryan Lazuka: Now on Twitter is like, they've got tools where you can reply with AI really, they make it really easy. So you can like, you know, if you see someone, because it's all about growth. People wanna grow on Twitter.

I mean, I'm the same way. And one of the ways to grow is to reply to people in your space, right? So in the AI space or whatever space you're in, and you just reply to anyone that's an influencer in AI or is talking about AI. And they've got these tools where you can just like hit a button and says reply to this person's post. And it automatically reads their post and makes a pertinent reply to what they say. So like if you go on your Twitter, I'm sure probably I would say from mine alone, I'd probably say about 10% of them are AI generated because I can just tell, because they're kind of generic. But it's gonna come to a point where you're not gonna know if you're talking to someone real or not on Twitter. You know, it's can be, it's wild. And even the posts themselves, a lot of times are made by AI, you know? So it's wild.

Justice Conder: Yeah, I, somebody I like, there's a goth 600 who does like fire memes or whatever. He said something this morning about, there's an old story about a nuclear test that, you know, scientists knows is that it shot a manhole cover into space on a high speed camera. But when you, when you, and this actually happened, you go back and you look at how fast it was moving, it actually had to shoot it into outer space. So there's this rumor that there's the furthest object that man has ever put into space is a manhole cover, right? And, and he said, I'm thinking about this. And then it occurred to me, I said, how much further is this manhole cover compared to like our, that like that plaque we sent into space with a man and the woman like inscribed on it, the golden plaque or whatever. And so real quick, I went in and checked the PM like, okay, how fast was it moving? Assume that it actually happened. How much further is it in space compared to our actual satellites we sent?

Put it on a chart. And within like two minutes, man, him talking about sending a manhole cover to space, I had like a chart showing the relative distances as I respond. Now it was all for humor and interest, but it is shocking that we're, can do that now for free.

Ryan Lazuka: For free. And that would have taken you like, Do. A week or days of it.

Justice Conder: Were you even like Google spreadsheet, you know, trying to like calculate and all?

Ryan Lazuka: Yeah, yeah, yeah. You can talk for days about AI, it's just insane. Like there's always something new coming out of the woodwork. Things as always, justice we're coming on. Like we'd love to have you out maybe once every month or two, if you're up for that.

Justice Conder: Yeah, I'd love it, man. I'll, you know, I'll keep a running tab of hot ones.

Hunter Kallay: The hot links. I like that. Bring them on. Tell it, justice is hot takes. There we go.

Ryan Lazuka: We can't implement that in the newsletter. I like it. Do you have anything else you want to publicize right now in terms of articles you're putting out or Twitter threads or anything that, you know, you want to market, feel free to do that right now, justice.

Justice Conder: No, I would just say anyone who's interested, follow me at Singularity Hack. And I go quiet in between papers because I actually need to deliver, you know. And so I go quiet and then I try to come out with a bomb, you know. So that's what I'm working on right now. And I'll keep a running tab of cool stuff and experiments and I'll make sure I have even better stuff next time we talk.

Ryan Lazuka: No joke. That's awesome. Yeah. If you want to check out justice's last article was about Eliza. It was really, really interesting. We'll put a post, we'll put a link in the description below. But justice, like I said in our first podcast with justice, if you want to stay on the cutting edge with anything, follow him because if I ever need to know what the latest greatest tech is before anyone else does, I'll usually jump on a call, call justice. And he lets me know. And my mind's usually just blown away by what he knows. So about the future. So feel free to follow him on Twitter and we'll put his link in the description as well for that. But thanks again, justice for coming on.

Hunter Kallay: And then be sure to subscribe to Ryan and I's weekday AI newsletter fry-ai .com. We've got the top stories of the day, emerging tools, some cool community interactions. Then we do deep dives on Sunday into different developers and developments and impacts AIs having all across society. So check that out. And just thank you so much for coming on and be sure to subscribe to behind the box. You can keep up with all of our latest videos.