AI Rebels

High-Leverage Humans: The Rise of 1-Person Companies Powered by AI

Jacob and Spencer Season 3 Episode 15

What if your next coworker is an AI agent? In this episode of AI Rebels, developer Nim shares his journey building AI Agent Dot App, exploring how solo entrepreneurs can now harness agents to do the work of entire teams. From bootstrapping with terminal-based scripts to creating sticky, memory-driven agents, Nim breaks down the evolution of autonomous AI tools, the psychology of adoption, and the legal grey zones ahead. It's a raw, insightful dive into the future of work, where personalization, iteration, and adaptability define the edge.

https://x.com/sup_nim
https://nime.sh/

hi everybody welcome to another episode of the AI Rebels podcast I'm your co host Jacob and I am your co host Spencer uh today we have Nim with us Nim is a developer based out of Bali he is working on a variety of things but most recently AI Agent Dot App and Studio Dot Gold Nim go ahead and tell us a little about yourself tell us a little bit about how you got into AI Agent dot app and AI in general hello I am guest Nim working on AI Agent that's my main driver right now how I got into it the early days of the AI agent stuff was like in the terminal you could get it to do whatever you wanted but it was like a terminal based like interface which wasn't that good we thought if it was right hard for us to use it it's gonna be hard for the rest of the population to have a go so we kinda just try and took the terminal idea put a nice UI in front of it and just put it out just to see right let other people mess around have a go and just one thing LED to another and now I'm working full time on building AI agents so AI Agent dot app is it was it based on a pre existing agent framework or was this something that that you rolled yourself and and built specifically for your project it stemmed from a baby a G I if you know it yeah like yeah sound dude this name um skips my brain right now but great guy great idea you're OK put it into this you're OK yeah sorry we've spoke a couple of times on Twitter a sound dude great idea and that like sparked this idea we can just get it to cursively talk to itself recursively talk to itself and that thing stemmed a lot of AI agent software it started there then it quickly wasn't that scalable and we found better ways to do certain parts and we started like Hodge podging different ideas into the root baby AI concept now it's um it's a mix of couple of things we use long chain we there's a AI framework called Crew AI if you've heard of it like it's basically simplifies this whole thing down to like three or four variables to get started but you can get very complicated if you want so we use that right now wow and how large is the team it's just me right now I say we just be nice well it's you and your agents it's you and your coworker agents yeah exactly me and my agents yeah I have a call from uh called him in just to give more color to that uh but he he's currently doing some freelance work he's got a mortgage to pay so he's trying to juggle like things and that's just more safer for his mindset yeah yeah yeah no that makes sense and honestly I mean I think that's what we're I was curious because we always like to ask because we're seeing this incredible emergence of small like highly leveraged human capital corporations yeah where it is just a small team it's a one you should copyright that I know I'm gonna write that down right now um but that's this is exactly the trend we're seeing AI agents with 1 to 5 humans really being able to do a lot AI has changed the game on that cause it's so accessible to get information you get stuck on things and if you just throw enough tokens at it and like have some sort of a a like a light direction on what the solution could look like anyone can solve any problem before you had to like trudge through so much crap like other people's Gethub requests issues to find some sort of answer whereas now you can just get AI to do all that which is like days of work saved it's so nice for debugging it's amazing uh bugs that I used to yeah do exactly what you were describing you know dig through Reddit dig through Stack Exchange instead of like a three hour process it's now like 30 minutes you know hey this is happening try this and then that doesn't work try this that didn't work what do you think and then it comes up with something you know that we haven't tried and that works it's great it's beautiful it's it's very freeing cause you don't get attached to the idea cause if it doesn't work you can like exactly you said try this try that then one of the tries is gonna work and you've not lost a a day yeah exactly it really lowers the what's the word the uh iteration the work required for it yeah iteration iteration cost thank you yeah iteration cost especially for writing something like code where it's like previously if I write just a 100 page a 100 line function for something and then it doesn't work out and I need to completely rethink my approach like you know that's now a hundred lines I have to rewrite whereas now it's like well that didn't work I do it instead haha and it takes 30 seconds it's amazing yeah yeah it's it's wild brilliant so Nim with this AI Agents dot app obviously it has evolved quite a bit over time what what have been some of the biggest changes from when you started to what it is now to give you some context I don't fully have product market fit I'm still trying to figure it out there is like inclines of like certain things that work we make some revenue but it's not it's not a Spotify it's not like open AI we're still trying to get to the right thing so to answer your question what have I Learned I think at the very start it was very general it could do anything and everything then but the AI wasn't as good to give it such a broad question and say go figure it out so then there was a time where being niche like targeting say uh SEO content writing or uh customer care or whatever the hell but being quite like god railed was good cause you could just yeah yeah give it a lot of context and I think it's swinging back to the models are getting much better now you could give it a broad request you may have seen like Manas maybe it's like a I think I'm not saying that written name wrong but it was like a an agentic software that came out you give it like a high level task it would open up browsers create folders files yeah run terminals I think it's Manus right something like that so yeah that it's Manus am I saying right yeah yeah that so if that's going back to just general again you give it like a high level task like go get me all of the SVG logos of universities I I did that task like yesterday like I raced it like I had to go on Wikipedia download whatever where I gave it the same task and it just went off and did it in about 15 minutes which is wow really impressive so what I've learnt is I think general was good general was bad to start with then niche became good because the air models wasn't that good but the air models are getting good enough general is back in so I'm curious to hear your thoughts on this um one thing about AI agents I've always been long been interested in them like I ran baby AI when when he released it on Twitter I've messaged him like 3 times trying to get him on the podcast this is another plug yo hey if you haven't heard this come on our podcast um but one thing about AI agents that I've been a little bit pessimistic about and skeptical about is that I think they're gonna require like a lot of um supervisory architecture and and really thorough evals to be truly effective and I'd be curious to hear from your perspective if you think that is true or if you think that that has now shifted um a little bit with with how capable the models are now I think uh I think you're right it depends a real is never bad I would I would look at agents as you look at other humans right you can you hire an intern they probably need much more guidance to like not mess up am I allowed to swear is this swearing podcast or not go for it you're good don't fuck up right you tell them don't do this don't press that button you can stick stick to these like within the confines of these rules but like overtime you maybe hire a senior or a principal whoever and you don't have to trust them that much because you like they've done it 100 times I say all of that to say currently it's all very broad so maybe over time you have these agents that know how you work or how you like your script to be written or CX to be taken care of so by the thousandth time it's doing it is very little god rails needed yeah hmm that's my answer what do you think on that do you think there's merit there or what's your thoughts no I I think so because I I do think that inevitably the end result is a broadly effective agent right because when you think about it's like you know the training for these agents just requires a slightly different type of data than than what they have right now right whereas now it's a lot of unstructured text they need more examples of agentic actions taken um and so I would imagine that you know by by GPT5 probably that is a pretty broadly effective agent on its own you know I'm spitballing here I don't have any inside information I'm not I don't wanna mislead people here but that's that's kind of how I viewed it um and I've done a lot of playing with AI agents recently as well and and a lot of my past pessimism has been replaced with optimism I think it has there's a concept of like memory like I think agents and humans it's just different names I think the mental model is exactly the same as humans if it does if you do something 100 times you'll get better at it to not mess up doing that thing I think that's the same principle with agents yeah no totally and I think this is a powerful principle overarching all of AI I think this idea that it's terrifying but also encouraging that AI is not that different from humans which makes sense it's trained on the output of human minds from the beginning of time right so it would be strange if it were this completely foreign entity and we didn't understand it like once you get into it and you start working with AI it truly is not that different from working with another human you need to give it honestly yeah you need to give it outlines you need to give it clear instructions and then I think you're right particularly agents it'll be interesting to see how this trend emerges if it truly is someone builds an agent that just has done this task 100 times and it's proven and people pay a premium to use this agent because it has done specific tasks so many times and it's built up this memory I think that could be a trend that emerges I don't know it'll be very interesting to see what this looks like that's what I'm going after currently like the give people the tools to build agents then create their own agent give it like the backstory what they can do assign tasks and give it tools but that's quite shallow right it's quite disposable what I'm currently doing is that memory idea where it knows what your last request was and if you were happy with the outcome and if not what did you change what was the next request like from a business point it becomes quite sticky right cause you're training your own agents to be yeah perfectly aligned to what you want and I think that's quite good yeah and it gets better and better as you use it so then the opportunity cost of leaving becomes higher and higher yeah yeah yeah that makes total sense interesting so what as you've told people friends family that you're working on AI agents what have been some have you had any funny experiences with misconceptions people asking funny questions or misunderstandings what have these relation what has it been like as you've delved into the world of AI agents quite interesting cause I think there's a two very polar opposite divide like most of my friends back home or like my family and stuff they don't really care they don't really understand it they hear the news or whatever it is and it's like like doom stuff right you're gonna lose your job next week they're gonna take it yeah so they're like are you trying to replace us and stuff so there's nothing there mad at you yeah you're like yes actually yes dumb questions like that you're easier than you think to be replaced no no on the other side like all my friends here in Bali they're all just like big mega nerds building their own startup so AI is just so like integrated into their everyday life it's common as in we don't even see it as like a it's a tool right we don't talk about our laptops anymore because it just works we do share like good like prompts or good methods of getting the most out of the AI but no not really funny stuff how about you guys like what what's you you guys I feel like got a bit of an inside track but you have like normal friends what do they think of you what you do honestly I think that a lot of people are still just not familiar with it it's anecdotally my my experience kinda kinda matches up with the statistics of you know only about a third of people have heard of AI but I will say out of my friends you know and family who have heard of AI most of them are optimistic and excited about it I mean on in terms of my family like we're all kind of techy people like my both my brothers are working at AI startups I'm doing this but then I I I see a lot of optimism from for it from for example my brother in law he has a creative agency for YouTube creators and things like that right so matches them up with brand deals they also do video editing thumbnail creation things like that right and he said that they have almost completely automated their their cold emailing with with wow AI personas they have a little AI persona on LinkedIn but nobody realizes it's not a real person it's just like you know they went to like this person is not real dot com or whatever it was you know and like generated a picture of like a generic professional looking woman and she does all their cold outreach and she does great haha she crushes it so haha it's interesting to see it start sneaking in here and there and also I think that anecdotally I watch a lot of TikTok and things like that to keep up to date with you know what's going on culturally and anecdotally I'm also seeing a lot more accepted the acceptance of it uh there I think both as the tools get better and as people get better at using the tools there will be more and more acceptance because to be honest artistically up until recently most of the people posting a bunch of AI art were were were kind of from the crypto space and you know they weren't they don't necessarily have the best taste um I hate to say it that way but that's the truth and I'm starting to see more yeah like I I start to see more and more people with uh more mainstream and and better taste start experimenting with these tools for example there's this guy on TikTok Sante Luca who does these ridiculous satires of like the hustle you know the hustle grind fest bro propaganda

you see on tiktokers like I wake up at 6:

00am right but it's all ridiculous AI generated art he's like I wake up every morning you know I spar with Mike Tyson and it's like you know AI generated Mike Tyson and a dude beating up on each other right anyways it's just really exciting to start to see real art form around these tools um Jake what were you gonna say I was just I was gonna say I think it's I'm continually amazed in my relationships how when you're in it when you're in AI and you're talking about it every day you assume people are at least aware of AI and its capabilities and I'm continually amazed when I'm talking to a family or a friend and and they're like oh I just used chat GPT for the first time last week you just use chat GPT like for the first time it's so cute yeah and it happens all the time and I think as this changes this is where I mean I think AI agents dot app is a good example where you saw this terminal approach to agents and you said huh this is not usable for most people let me adapt this and make it approachable and usable for people and I think this trend is gonna continually be happening in AI where once it's just like AI agents like it's it's a very sticky thing once you integrate it into your life because all of a sudden you get all this time back because now you don't have to go do the research yourself you can just verify the research AI does for you and then you can do the more complicated work of analyzing that research or that's just an example but all these things in your life once you integrate it there's no going back why would you go back and lose time from your life when you know you can so I just think one I'm continually amazed in my relationships how early it still is because it seems in my world it's everywhere but it's not that way for most people and I think it's important for us who are in AI to remember that and continue to try to do as you're doing and make it accessible and approachable and very sticky so that people when they use it they want to keep using it very true but one thing I I just wonder like wonder why like why hasn't there's a clear like making these numbers up like a 50 50 people that are so behind the eye and the other yeah just is not phased by it it's not like they're not aware of it cause it's everywhere like it's in the news like I think it's nearly a household name chat GPT but what's the lack of motivation to like play around with it I think I don't know I think there's a yeah I think there's a couple things I think one is there's still a shocking number of people who kind of assume that what AI is doing right now computers have been able to do for years like a really common thing that I see on TikTok and Twitter is like you know someone post like a a video that's you know CGI from like 2014 and it was like oh it's AI it's like you guys have a weird vision like I like I even saw someone arguing like that like you know oh no like AI was AI could do this in 2014 like yeah AI just AI couldn't do this six months ago and so I think that people are not really and and so that leads to my kind of my second point is I think that people are not really familiar with technologies that that develop this fast I think there's another cohort of people who you know they tried out chat GPT three and a half turbo when it when it came out and you know November of 22 2022 or whatever it was right and they haven't tried anything since then and so they don't know how fast it's improved and how much better it is like you still I mean people only recently got over like the whole like I can't draw hands thing you know like like it takes a while for people to update their their understandings of the world at a at a social scale maybe they're not seeing the benefits like yeah and that too like where's you guys and I are seeing actual value getting out from it like on a daily basis maybe this don't it didn't serve a purpose for them today interesting I yeah and I think that's spot on Nim because I think this is the same like the colloquial early adopters whatever those are early adopters and then the main group adopts it and then the late adopters this whole trend I think until until people tangibly see in their lives the value they're not gonna do it I also think it's just human nature to change when you I think a great example is coming in new to a company right you come in someone shows you a a process that they perform and often times when you start asking you know why do we do that why do we do it this way often times the answer is because this is what this is what we do this is what we this is this is the way it's always been done so unsatisfying I yeah yeah and I think that mindset is also in our daily lives like we wake up you have breakfast you do these things and it just doesn't even cross your mind until you've changed to use AI let me let me go to chat GPT for this like I think so many people particularly older generations where they're not exposed to it constantly on TikTok and and these things I think it's gonna take a while for it to even enter your head sorry to cut you off Jacob it's to add to the psychology part of it we did some user interviews at the very early days of Air Agent our app and majority was like business users and there was a like a recurring pattern which was in some ways they were motivated by fear it's like okay they've got competitors left and right how can I have the advantage so if I can leverage AI to get 10% better or do this faster and I to pull this back to the earlier question was maybe this is there's more to lose here right if the if they get shit at business or if someone does it better than them faster than them they lose out money like livelihoods whereas maybe right that's a good enough incentive to just get behind AI and learn maybe we all are in some degree influenced by that right how can we be the best version with the tools we have rather than just let it go by yeah I think that's I think that's accurate cause I mean you look at the the industries that it's being adopted most quickly in and it's uh you know the industries that are are most prone to what you're describing right coding world where you know the getting your app out faster really matters or the uh auditing world that that Jake came from um uh once again it's a case where if if your team is a lot faster you're more likely to be hired I I'm I'm I'm assuming so anyways yeah yeah definitely for sure how expensive fields like lawyers yeah yeah yeah highly competitive spaces yeah that's it that's an interesting insight I haven't heard anyone mention that before yeah yeah no I think that's I think that's spot on I'm curious name have you had any experiences with agents being built with AI agents dot app that have performed an action that they should not have whatever that is and people try to hold you responsible as the platform I can happily say no there there hasn't been do you think that will happen like just in the space I think that's an interesting question I think that's an interesting question it's like if you would abstract it away does the shopkeeper who sold you the gun go to jail for you shooting someone yeah right yeah it's a good analogy yeah so I I I don't know I don't think use the you know use an internet analogy you know most most countries protect platforms from being prosecuted for for what users say on there right at least to some level depending on the jurisdiction I'm sure but yeah I I it'll be interesting to see you know kind of where the legal consensus lands there to me it seems like you know if I went on to AI agent dot app and I made like an AI agent that was I don't know dedicated to making like you know something heinous like like this is a revenge porn bot I'm sure that you probably don't enable that but like you know just to use a a really egregious example haha um it does seem like the I mean obviously the the uh yeah no it's an interesting legal question like where does the responsibility lie cause obviously the user right is primarily at fault but you know technically you're providing the platform I mean coming kind of coming back to what you said name does the person who sell the gun is are they liable for unlawful actions taken with the weapon that they sold and I think this is a good analogy because I think similar regulations will be put in place where assuming everything is lawful assuming that the gun salesman you know checked for ID and did all of these things and and followed whatever those steps are then I don't think they should be held responsible because they did their part and then it it's no longer there but in your example Spencer it's like well should certain guard rails be put up and if the platform side steps those and maybe they should be held responsible but it's gonna be so messy and yeah I have no idea how it will be enforced it's gonna be crazy to just add to that really quickly I think one of the benefits is like I'm standing on shoulders of giants like open AI anthropics Google like they're gonna do their they've got much more to lose if they mess up that's true right they've picked up a lot of exposure insane amount of exposure so if Spencer wants to go make a porn bot a porn bot and if somehow it gets through it's on them yeah yeah you be like was it yeah go hey don't ask me go talk to Sam Altman exactly exactly but interesting question here's here's one for you guys yeah pretend someone is able to deploy a totally autonomous AI agent to you know a server somewhere and after that hands off they don't have anything to do with the agent the only thing they did is they pressed the button to deploy the agent now the agent lives on its own um you know it has some method for maintaining its server cost right whatever that is right we'll just you know pretend that exists pretend this agent then excuse me goes and you know perform some illegal action who do you hold responsible there do you still hold the original developer responsible do you like do you hold whoever owns the server responsible like do you throw the AI into you know AI prison trap it in one of those cloud flare AI labyrinths that they've been talking about throw it on a hard drive and send it to space yeah something I don't know Jacob what do you think yeah uh one first I think it's insane that this is a conversation we're having it's like what if someone creates a sentient being like if we were to take it extreme if someone were to create a cyborg totally sentient right that and then they just set that cyborg loose and that cyborg slowly devolved into all it does is rob banks and hold people hostage is that inventor liable because there was a bug in the cyborg's code it was supposed to be a bank teller but instead it's robbing banks and there's this bug I know this is getting insane but now that you said it like that actually to jump in I do think it is yeah the developer that is at fault OK autonomous agent it's not fully autonomous the root the root started somewhere right and it's the ideas that you would have put in or the developer would have put in so I think that person would be liable yeah I wanna make another point here real quick which is Jake you were saying like it's insane that we were talk that we're talking about this at all and it is but also at the same time if you think about it we've been talking about this concept for a long time yeah go read Frankenstein yeah right like this is exactly we are we are literally discussing like the exact moral quandary at the center of that book right and so this is part of what gives me optimism for the future when it comes to AI and humanity's ability to adapt to it is like we've been telling ourselves stories about this for years already we have already some some cognitive and social tools for discussing these topics we just need to kind of kind of re leverage them into a new area yeah totally agree I kind of wanna go back to something you mentioned Nim about your family who's not really in the AI space and seeing things in the news cause I think I have a similar experience with a lot of my family and friends um but what do you tell people when they voice this fear that AI particularly I think it's become even more pronounced with AI agents this idea that AI is going to take jobs cause AI agents are so tangible you can watch it perform a task on your computer and it's very easy to see okay this is gonna replace these jobs what do you tell people to help help their fears I say skillshare.com no no I like it no I honestly don't know like I think I it I honestly don't have a clean answer here but essentially you wanna skill up in some way that mm hmm you can't be automated out tomorrow but that's not really comforting that's what I think in my head right the answer the thing I would say is probably don't worry it'll be fine yeah but what would you say I actually honestly don't have an answer to that what what would you say I know it's hard I like your like it'll be fine cause I think it will be I truly think it will work itself out we have adapted to so many fundamental shifts in technology and society over the millennia of time that I think we're gonna be fine we are extremely adaptable and we can do it I think also another important thing to do when talking to people about AI is acknowledging their concerns right cause there are obviously legitimate concerns with AI like it it wouldn't be exciting if if it you know if there weren't right it wouldn't be transformative um and so I think that it goes acknowledging their concerns goes a long way cause then it's like yeah you know like that is a problem here's what people are doing like for example like you know the the energy usage issue right yes it's a problem to be solved but the good news is is that like the people developing these models are incredibly heavily incentivized to solve it right if if any of them can unlock like a significant energy performance gain like that's that's you know billions of dollars unlocked for them trillions of dollars unlocked for them overnight and then the second point I like to make as well going to what Jake was saying is like yeah we've we've we've endured a lot of fundamental shifts in society before like you know yeah it's it's gonna be bumpy in parts but in the end this is progress this is what progress looks like progress is rarely if ever you know clean it's always messy it's always hard to do but you have to do it yeah'cause stagnation means death exactly that was so well put I wish I said that so you can say it now we'll shift that over to you yeah we'll change his voice have you said it yeah run it through 11 laps here let's quickly mouth it out yeah we'll we'll overlay the video perfect yeah yeah um oh man I just feel like this is the Wild West there's so many things that are gonna shift and change can I ask you guys a question yeah what's what's been like the most interesting you talk about AI for 60 episodes and you live it outside what's been the most interesting thing you've came across it could be like that's such an obvious answer obvious solution or that I would have never thought of that or something that's like you think about Julius AI um I am I'm super impressed with what they're doing over there um if you're not familiar with it it's it's an AI data analysis platform really really well done uh and they're they're an incredible example of what Jake and I what Jake was talking about earlier of like the you know they're like a what was it like a five person team Jake I can't remember what row said but something like that tiny team and already doing you know like 500,000 in monthly revenue something like that like just uh I maybe it wasn't monthly maybe it was annual I don't remember exactly point being 500,000 users okay yeah that's what it was 500,000 users thank you very good um yeah and and that that really stuck out to me as like kind of a bellwether of of where all startups are headed right like like this is this is this is how startups are gonna run in the future hahaha it's gonna be a team of you know a small team of under 10 people probably for the first year or two of its growth and they're only gonna start bringing people on when they need when they need more people for the purpose of interfacing with the public basically yeah um and it's gonna yeah I'm just I'm just fascinated to see who else can replicate their success mm hmm totally yeah they're a great example of what is possible for sure I think something I know I mentioned this earlier but we had a professor on his name is Professor David Wood and we asked him we've been talking a lot about what does how is AI changing the education landscape are are teachers basically just grading chat GPT now cause all the papers are authored by chat GPT at this point what does this look like how does this work how are professors particularly at the collegiate level really grappling with this I think when we asked him his answer summarized a lot of this approach that the strategies we use to live a successful life are the same strategies we use to work with and integrate successfully with AI cause he basically just said yeah I tell my students to use AI and then I just the bar for my assignments now has gone up cause I assume if you don't run it through AI what are you doing why would you not and he and then he said also some of the nature of my assignments have changed some of the things that I know they need to know as a person maybe it's more of like a live vocal exam right where you have to give a presentation on these things that you need to know yourself but then if you're writing a paper yeah you better run it through Chat GPT if you're writing code for an assignment he said my assignments have all kind of escalated where before we just didn't have the time to get certain things to a certain level he said but now I can take them through the full life cycle of things because they're using AI I don't know so I think just speaking with him has been a theme we've seen where using AI is not doesn't need to be and it shouldn't be this these completely new and earth shattering strategies and I think that grounds us when we approach AI yeah and uh kind of riffing off of that I think that we're going to see a shift in education back towards handwriting and this is kind of off the wall but reason being so first there's all the scientific evidence about like you know handwriting like handwriting out something by hand you know secures it in your memory better than typing it on a computer etcetera and also it's an easy hack around people using AI right uh they could still use chat GPT but like either they're still they're gonna have to copy it down by hand and so either way it's like you know you're forcing the student to engage in the act of learning and so I think that we'll see kind of you know the equivalent of that play out across a lot of educational domains of of reinvesting into by necessity reinvesting into like very very very fundamental skills because that's what that's what unlocks you know more capable use of the higher higher level tools it's the same idea when everything's mass produced the the craft craft produced item is of much higher value I think there's probably some sort of link there to the handwriting and I think to quickly say it to the small teams I think that's really good that means it'll take you maybe two people like you and you and your best mate to spin up a business and just have another like have multiple goals at figuring something out I think that's brilliant the education I would have never thought of it like that he's just made that professor has just made it harder for them to cheat but by also raising the bar is gave the kids that want to learn it's like okay now go build an app the kids yeah he might have put like the first level 1 and level 2 but they go still have to go figure out okay I need a database and it gives them a ceiling a headroom to learn new things on their own and ultimately that's what you're trying to teach in school right the skill of learning exactly motivate people to learn more it's brilliant yeah and I think that's the trend we need to see across the board I think he's a great example but I think every single industry the answer is not a the answer is not to just run away from AI the answer is how do we adapt it's not going anywhere what do we need to change so that we use it in the proper way to further my yourself and humankind cause I think it is a real fear that AI will a lot of people will start to use AI as a massive crutch because it's so easy to do and stop thinking another one of our guests that really stood out to me and this is slightly tangential but I'll connect it in don't worry um was his name is Bryce Judy and we brought him on the first time we brought him on he had he was developing a product for developing language sets for under resource languages particularly I don't remember the exact right now it was a indigenous language in the Yucatan Peninsula in Mexico right um and so he gets it all ready to launch he's he's trying to sell it to people and then like you know as soon as he's trying to sell it to people Google announces that they have added like 10,000 new languages or something to Google Translate and you know the the all of the languages that he was trying to target were were included he's like oh well that sucks and so within a matter of you know I don't know probably six months maybe less yeah he completely pivoted and started a new product targeting you know it's called Tariff Low I think it was using AI to calculate tariff duties and and this is another just I bring this up because it's another great example of what AI is able to do for people which is like you know his whole idea based around AI got blown up by AI and he turned around came up with a new idea executed it with AI and now he has funding and employees right like it's it's amazing what leverage these tools give people who are highly motivated and willing to try difficult weird things ha ha ha especially just quickly jump on that I think the key thing was the highly motivated part because if it if it took him six months and there was no way for him to like get back to that spot and it would take him three years that's gonna take a much harder hit whereas he knew the tools and he could he could get back to something viable yeah exactly in a month yup too yeah I love that if you up your upskill yourself in AI it's applicable where like yeah no matter what happens you can still use those to shift yeah and and to that point people constantly like I I see this all the time like oh no no no point learning coding now no point learning how to write now no point learning how to draw now it's like that's all a lie learning all of these things is now even higher leverage than before if you are a really good artist you're gonna be able to use the you know use image generation a lot more effectively than somebody who is not a good artist because you know what looks good you know what to prompt for you know what to look for etcetera yeah the same thing goes for you know co generation tools right like it's it's applicable everywhere like reinvest if I could give advice to to anybody listening to this episode like reinvest yourself into fundamental skills don't think that AI removes the value of them it's an amplifier agreed yeah I say that all the time it's not it's not an amplifier it's it doesn't replace it amplifies what you have if you don't if you're not an artist it's not gonna make you an artist unless you are motivated and you have it coach you through it Spencer mentioned Bryce who had to pivot and obviously AI Agents is a huge space but there are some big players in the space are you nervous of AI agents dot app being being hit by chat GPT coming out with an agent platform these big players is that a a worry a constant worry I stay I lose sleep over this no not to that extent but it of course it's a worry everyone wants it's the it's the it's the natural path forward so anyone with any sense is looking in this direction which is not that good if you're already here but I think I also think it's back off but you can't right how am I gonna say back off to open AI you know the I'm a small I was here first but I also do think it's the idea when the tides rise all boats rise with it so if they're doing such good work and they're sharing their learnings that's like $1 billion worth of research I'm getting for free yeah so right how do you distinguish yourself like do you have any ideas of how to remain relevant yeah I think I have to just go talk to businesses there's a bunch of businesses that like already in the database that have used it my strategy for the next like 2 3 months is go in and I mentioned earlier about like the memory or building like building up knowledge to your agent I think there's something there mm hmm and it goes after that Spotify model become sticky do you have any ideas actually I'm what are your thoughts uh no I think this is I think what you said is important is the cause open AI can provide a platform but they don't have what you said that personalization so I think that bridge of that expertise that you could bring seems like that would remain relevant you know as if you provide that also as these larger providers start launching their own tools it grows the size of the pie where you know once people are aware that these capabilities are possible some of them are gonna be looking for alternative option options you know whether it's incapabilities pricing options etcetera right and so I think that what I'm hopeful for for for you know platforms like yours is that that growing the size of the pie maybe you don't necessarily need to grow the size of your slice as much just because there's there's more to go around yeah it'll be it'll be really interesting to see how the economics of the space work out I'm excited maybe I'll come back to do this podcast like in a year's time yes where we would love that who's open AI you know ha ha ha it's AI agents are up ha ha AI agents are up it's named after AI ha ha ha ha ha ha ha I love that haha so as we're as we're wrapping up here um if you if you had to give advice to someone who's looking to get started in this space building a business or whether they're just you know looking to start using these tools what advice would you give them you just have to just do it like pick something that you care about and just go at it to some it has to have a finish like yeah set some sort of goal like every business I've started the objective was to make some money and okay I made some money I had a clothing brand made some money I was like this is not a viable business I don't like brick and mortar been before that was a photographer videographer this musician even right and I was like I aim for one goal like I set myself a goal of like make some money to see if this is a viable idea if nothing I Learned via via the path to that to summarize I waffled on a little bit the advice would be to just do something to a point where you're happy with like alright I did this yeah there's no point in you starting a project and bidding it two weeks later and you've got nothing to show for it it has to have a finished in some way enough to so you can put it on your LinkedIn profile that's the match that I'm looking for I I like that no I love I love this idea of putting a finish cause we often hear the advice of just start do something pick something get going but I think that idea of having a specific goal is crucial one because it's not easy still it's not easy to build a business with AI people are like now you can just do it no no it's still very hard and I think this idea of having a specific goal is very motivating and allows you to measure your success so I echo what Nita said I have a little bit of color there that's good advice cause I used to start a lot of things but if you start 100 things but you finish nothing you've done nothing yeah like it's not very satisfying as for yourself you've done nothing for the last two years where is you might have finished 10 unsuccessful projects that's still quite nice right right that's it what advice do you have for me guy guys Spencer keep going honestly I think that is big advice in the AI space those that are motivated to keep going will naturally evolve and survive and I think find success like it's a war of attrition yeah like yeah I agree with that yeah if people are wanting to follow you follow what you're up to your progress what's the best place to do that obviously we'll put AI Agent Dot Apple put that in the the description of the video what about socials LinkedIn Twitter at sup name s u P underscore n I m that's where I'll be perfect excellent we'll drop that as well thank you so much we definitely want to stay in contact so we'll see where you're at you know over the next year I'm very excited to see what happens for you thank you great chat boys look forward to talking again thanks for coming on