Latest / The Koerner Office - Business Ideas and Deep Dives with Chris Koerner / Learn AI Automations Before You’re Replaced by Them ⏐ Ep. #192
Transcript
- 0:00Are there any low hanging fruit ways that people could make
- 0:02money with these tools? We're not a technical person.
- 0:04You're not allowed to say that anymore.
- 0:06You can literally go on YouTube, ChatGPT and learn whatever.
- 0:10We live in a beautiful time. Platforms like NEA, make.com or
- 0:13gumbo. It's a perfect place where a non
- 0:16coder can experience when it feels like to build something
- 0:18that's actually useful. I have no doubt you'll be able
- 0:20to build cool stuff. I see a lot of people, Chris,
- 0:23like get frustrated. Oh, it's not working always like
- 0:25this. It's like give it a few tries to
- 0:27read up, ask ChatGPT. SMS are often right.
- 0:30Optimist make money. I challenge everyone watching
- 0:33build at least one after this episode, at least one
- 0:35automation. Use the heck out of all the free
- 0:37trials and most importantly, have fun learning.
- 0:40When you're having fun, you're not working.
- 0:42Why? Don't you start by telling us
- 0:46like what type of like side projects that you're interested
- 0:50in right now? I'm very much interested in a
- 0:53lot of like AI agentic sort of stuff where being able to
- 0:59construct agents to do certain tasks right, Like something as
- 1:03simple as read my emails and then perform X task.
- 1:07I think that type of stuff interests me more from a
- 1:11learning perspective than from actually building and shipping
- 1:15product. I think if you're going to build
- 1:17something like that, you run the chance to compete with the
- 1:21giants like GVT Chat, GVT Claude and Gemini.
- 1:26And I think the sweet spot in this, in this space we're in is
- 1:29either to build a niche product like Corniche, however you
- 1:32pronounce it, or is to somewhat build something that as the
- 1:37models get better, your product gets better as well, right?
- 1:40I think, you know, early on we saw this a lot and even now
- 1:43we're seeing it happen where people will build features that
- 1:48the models get better and then your feature gets swallowed,
- 1:51right? Now, you can continue to carry
- 1:53on with branding and marketing, right?
- 1:55But I think if you're wanting to build a business, I'm, I'm
- 1:57hoping you want to last a little long, it helps to either build
- 2:01your product, aligning yourself with the models or building it
- 2:04so niche. So niche that you know, whatever
- 2:08space you're in, they love you and they love the solution you
- 2:11have. So they could care less if the
- 2:12models could do it. So that sort of space interests
- 2:15me. So like to your question, I, I,
- 2:17I build a lot of projects, side project, and that's for your
- 2:21just so I can know what it takes, at what speed I can
- 2:24build. I'm very much at a point where I
- 2:26don't really have any ideas that I'm willing to like Shepherd and
- 2:30say this is the one. So in the meantime I just
- 2:33sharpen the axe till the idea comes.
- 2:35Yeah. What agents are you seeing that
- 2:38are actually like super useful, super hands off?
- 2:43Because like with my experience with agents, they get like 90%
- 2:47of the way there, but that 10% is so onerous that they they
- 2:50can't just run without my input. What are you seeing today?
- 2:54That's like a process, a gum loop, integration, or anything
- 2:58that you could classify as an agent that really makes your
- 3:01life better, or it could make anyone's life better.
- 3:04Yeah. So I'll be a pessimist and
- 3:06optimist. The pessimist in me will say
- 3:09that we really haven't seen true agentic workflows yet.
- 3:13Everything so much so far has been things being glued
- 3:19together, right? When you look at NEA or like
- 3:21Zapier or make.com, it's really like a bunch of these tools
- 3:25being glued together, giving them access to these models as
- 3:29tools, which is impressive, don't get me wrong.
- 3:31I will say the one thing that shocked me in the last couple of
- 3:35days is Claude Code. I know Cloud Code is on the
- 3:37pricier side, but really like, you know, and for the people
- 3:41watching, I, you know, for one month, maybe cancel every other
- 3:45subscription and just get one month of Cloud Code.
- 3:47Because it's one of those experiences where I've been
- 3:50using it the last couple of days and it will go on for like 20-30
- 3:54minutes, building features of debugging itself, right, writing
- 3:59documentation. And it felt like everything
- 4:02we've seen and experienced so far was, in my opinion, pretty
- 4:06mediocre. And Cloud Code is sort of like
- 4:08I, I tweeted upon recently, it's sort of opened my eyes where we
- 4:11really haven't experienced agentic, right?
- 4:14Because a lot of people have this notion that if it's
- 4:17agentic, it has to be fast, which is not the case, right?
- 4:21If something is agentic, I actually want it to take it's
- 4:23time and get it right. Versus I have this automation, I
- 4:28have that automation, I'm patching these things together,
- 4:31ensure they're firing. But the reason why there are
- 4:33mistakes and like you said, there's that 10% where it just
- 4:38can't jump over is because there's no feedback loop for it
- 4:41to reiterate itself. You have to jump in and either
- 4:44tell it, do it again or change some things up.
- 4:47But with Cloud Code, it's actually the first time I've
- 4:49seen, oh, it'll keep going till it gets it right and it will
- 4:53research and it will do all these things that in short, we
- 4:56haven't reached agentic yet, but we're starting to get there.
- 5:01Practical advice, What does it mean for the end user?
- 5:03What it means is it's actually important for people to fill in
- 5:07that 10% gap. And here's the reason why I
- 5:09think there is value in being the person who knows, like who
- 5:14knows how to get the best out of tools.
- 5:16You know, the better driver, like we all can drive, right?
- 5:18Some people are better than others, right?
- 5:20We all can build tools and automations.
- 5:23Maybe some could do better than others, right?
- 5:25So I would say that yes, we're not there yet, but I would take
- 5:28this time to further build knowledge, further build
- 5:32strategies, further get used to the tools, not because the tools
- 5:35won't change in the next three to four years, but the knowledge
- 5:38that you acquire now will have some benefit as things get
- 5:41better. Do you think that Claude code is
- 5:44impressive and just as helpful to a non-technical person or a
- 5:47non coder as it is for you? There's two sides to this.
- 5:51I think it's impressive to anyone when you see the level of
- 5:54output, but I think the level of admiration will be different.
- 5:58Right There is this love hate relationship I have with people
- 6:02who vibe code and vibe coding in general.
- 6:05And the reason being is I see a lot of people get frustrated
- 6:09after maybe 3 hours of vibe coding an app and it doesn't
- 6:14work out and something breaks and they flip out on the tool
- 6:17and they get pissed. And it's like, welcome to
- 6:19software development. They don't have the perspective
- 6:23that you do you. Know what I mean it takes time
- 6:26right? So This is why even when like
- 6:29you know the models were crapping the bed and weren't
- 6:32that good and are just starting to get decently good.
- 6:35The simple fact that it would build a simple UI or it would do
- 6:39simple tasks was, was was amazing to me because again,
- 6:43the, the value of, of software and how much time it takes for
- 6:47someone to architect software. We're entering a time where, you
- 6:50know, code is starting to become cheap, cheap in a sense where
- 6:53it's for most businesses. It won't be your Moat.
- 6:57But I say this to say that a lot of people don't have that
- 7:02understanding of how hard it is to architect a good software.
- 7:05And what I mean by good software is software, if not architected
- 7:09correctly for the most part is like milk.
- 7:12It will go bad really quickly. And This is why like companies
- 7:16will always have software engineers because even though
- 7:19the software is built well now maybe packages go out or there's
- 7:22a new way of doing things or the there are new features you have
- 7:26to build. So software is not like wine.
- 7:28It does not age well. It ages terribly right.
- 7:31So if someone has that appreciation and that
- 7:34understanding and they start to vibe code, I think you'll have a
- 7:37better experience versus, oh, I prompted 7 times and it doesn't
- 7:41work. And I work at tempo.
- 7:42So I, I, I have the privilege of seeing people's prompts.
- 7:45People will be like, build me a database.
- 7:49I'm going in circles to say that whether it's cloud code or it's
- 7:54tempo or it's all these other tools.
- 7:57I do think the only way people can be successful at vibe coding
- 8:00is if there is at least a fundamental understanding of
- 8:04technical concepts and those concepts can help you drive, you
- 8:08know, prompts, decisions, etcetera, etcetera.
- 8:10Yeah, So I think Cloudcode is impressive to anyone, but if
- 8:14you're a developer, it's it's pretty, it's pretty, it's pretty
- 8:17rad. It's pretty awesome.
- 8:18Would you say you have to be a developer to get the full use
- 8:21out of it? Otherwise just try raplet or
- 8:23something. Here's what I say.
- 8:24I think we're entering a time and a space where I personally
- 8:28at least, maybe I haven't lived long enough.
- 8:29I don't think there is a time in history that we can use to sort
- 8:35of help guide us into what steps are the best.
- 8:38So there's a part of me that's like, try all the tools, all of
- 8:40them. There isn't a single one that
- 8:43you should try to fact. If I were most people, I would
- 8:46be cycling through all the tools, one month, this tool, the
- 8:49next month this tool. Even if it's the hard and
- 8:52technical ones, right? I think there is value in
- 8:56learning how these tools work. And the same thing in software
- 8:59development. The learning is usually in the
- 9:02pain where something breaks and you have to figure it out,
- 9:06right? And before AI developer will go
- 9:08to Stack Overflow or go try to find this one obscure blog
- 9:12written by some guy seven years ago.
- 9:14Like that pain of trying to come up with a solution or understand
- 9:17why something isn't working, although very, very painful, is
- 9:21very, very valuable. So I say that to say, as these
- 9:26beautiful vibe coding tools come out, do not rob yourself of the
- 9:29painful learning that you can benefit from exploring every
- 9:32single tool that exists. So yes, Cloud Code is definitely
- 9:35more of a power tool for developers.
- 9:37I still suggest you should check it out because what's the worst
- 9:39that can happen? OK, tell me about NAN and why
- 9:42you think it is special. The idea of being able to chain
- 9:48certain tasks together, certain actions together and build
- 9:52workflows is pretty impressive without having to write code
- 9:55like platforms like NEA, make.com, like you know,
- 9:57whichever 1 floats your boat. I think are a great place for
- 10:01someone who is an idea guy or gal should be experimenting,
- 10:05right? Or gum loop like there's a lot,
- 10:06there's a lot of great tools out there, right?
- 10:08I think it's a perfect place where a non coder can start to
- 10:12experience what it feels like to build something that's actually
- 10:15useful. So if I was a non developer, and
- 10:18I've thought about this for a long time, a lot of the stuff
- 10:21that I would do is I would play with a lot of the tools like I
- 10:23mentioned, but I would also start to automate even the small
- 10:27things in my life, right? It doesn't have to be perfect.
- 10:30I think you rob yourself of having to like, not everything
- 10:34has to be assessed. Maybe there's some things you
- 10:35build for yourself, right? And in doing this, you prepare
- 10:38yourself to be that person who builds something.
- 10:40That's pretty awesome, right? So I challenge everyone watching
- 10:44build at least one after this episode, at least one automation
- 10:48that does something that you do on a regular basis, right?
- 10:51Let's say you get a lot of spam emails, right?
- 10:54Maybe something that reads those and you use GBT to respond in a
- 10:58hilarious troll away. I don't know what it is,
- 11:00whatever it is. Just something to learn.
- 11:02Yeah, just something to learn. I think there is value in
- 11:05learning and often times, especially with like Twitter,
- 11:07every one is perfect, everything is perfect.
- 11:10People don't make mistakes. At least that's how it looks
- 11:12like when you're looking at other people's stuff.
- 11:14But I really challenge you to to just build, have fun, enjoy the
- 11:18process, break things, learn, enjoy the pain.
- 11:21But that's what it's like being a developer, right?
- 11:23So I think there's certain things that even as the models
- 11:26get perfect, they're not they're not going to read your brain
- 11:29unless you put the chip in. And I don't think it'll do that.
- 11:32But I say that all to say, embrace it, right?
- 11:35Embrace it and just have fun. And if and if it's fun that
- 11:37you're having, it doesn't become stressful because I see a lot of
- 11:40people, Chris, like get frustrated.
- 11:42Oh, it's not working. Oh, it's like this.
- 11:44It's like tools don't get built in a day, at least something
- 11:47valuable, right? If you're building a lab, which
- 11:49I understand, but you know, most people don't even know the
- 11:52different types of databases and are frustrated when after one
- 11:55prompt, the front end's connection or the back end's
- 11:59connection to the database doesn't work.
- 12:01Yeah, give it a few tries. Try again.
- 12:03Try tomorrow. Maybe read up ask Chad GBT One
- 12:06thing that I do is for every place I get stuck at just
- 12:11because I want to learn from it, I will literally either
- 12:13screenshot whatever area I've experienced or where I've got
- 12:15stuck at. Last ChatGPT what happened and
- 12:18how do I learn from this? I think just investing in these
- 12:22little small actions prepares you for tomorrow's world.
- 12:25Because I'm I'm one of those people that doesn't think like
- 12:29I'm not AAI is going to replace everyone type of person.
- 12:32I am a person who says AI is going to change things and the
- 12:35people who change with the times and update themselves with the
- 12:39times will be the ones who reap the rewards.
- 12:41There's a same pessimists are often right.
- 12:44Optimists make money. Yeah, I love that.
- 12:48I love that thing. What is the coolest automation
- 12:51you've built, and are you willing to show us how it works?
- 12:54Yeah, so my favorite one, it was an AI pizza ordering agent.
- 13:02Oh, that's cool. Basically, it went to my
- 13:05favorite spot, which is dominos.com, Domino's dot CI
- 13:10live in Canada, Domino's the best pizza place in the world.
- 13:13So what it essentially did was I, the user had a chatbot
- 13:19interface. So I also built a web app for
- 13:21this. So it wasn't just an automation.
- 13:23And I'll kind of explain how you can build both.
- 13:27There is a chatbot interface, and this chatbot interface I'll
- 13:32use Circle, had an AI agent I built.
- 13:36I use agents loosely, but you get what I'm trying to say.
- 13:39Yeah. And this agent interfaced with
- 13:45Domino's API. Now I will say the reason why I
- 13:48didn't publish this or I don't maintain it is because the API
- 13:54that it used to interact with Domino's was like not an
- 13:58official API. It was like some guy who built
- 14:00it six years ago and like there's no really secure way to
- 14:03handle payments. So it's a little SUS, which is
- 14:05why I only built it for myself. But basically the way it worked
- 14:10is user would come on the site, the AI agent would notice that
- 14:14they're on the site, would say, oh, would you like to order
- 14:16pizza? And then it would give you 4
- 14:18options because that's how I configured it.
- 14:20And then it would go back and forth.
- 14:22So it would have this conversation with you where it
- 14:25asks you a question, you give it an answer, that answer becomes
- 14:29gets stored in ADB. And then it keeps going back and
- 14:32forth. And then once everything that
- 14:34was needed to make the order was set, once the AI agent had
- 14:39everything it needed, would then call the Domino's API, make a
- 14:42payment and all that stuff. Here's where things get
- 14:45interesting. And maybe we'll have this as
- 14:48homework for people is I think, NAA web hooks.
- 14:54Yeah, so NEA and I'm pretty sure gum, that's what's hot in the
- 14:58streets now, I believe. But with NEA, they have what you
- 15:02call web hooks. And basically what's awesome
- 15:05about this is you can connect whatever automation you've
- 15:10built. Like let's say I have an
- 15:12automation that every time I get an e-mail, I'm a YouTube, I get
- 15:15people who e-mail me if I need someone who make my thumbnail.
- 15:19So maybe I have an automation that, you know, read all my
- 15:22emails and if someone mentions thumbnails, ask how much, right?
- 15:32Something simple, right? I don't want to go back and
- 15:33forth. I just want to know how much,
- 15:35right? If they're emailing me, they're
- 15:37probably sending me their portfolio, what they have ask
- 15:40for how much. And then you can get fancy where
- 15:42you can get it to send you like a message to WhatsApp, Telegram
- 15:46and all this stuff, right? But what's interesting with
- 15:49webhooks is I can connect this automation with a web app and
- 15:56that's when it becomes interesting.
- 15:57So let's say you have an automation where a user just
- 16:02needs to give you their e-mail. And then what you will do is you
- 16:06will scrape the Internet for a specific shoe and if it goes
- 16:11below a certain price, then it sends them a notification.
- 16:15You can build that automation in NEA, right?
- 16:18But how do you package that? How do you sell that?
- 16:21Or how do you build an interface that makes it very, very simple
- 16:24for someone to use that automation, right?
- 16:27Webhooks is where it comes in handy.
- 16:29So again, I'll have it highlighted here.
- 16:31You can literally just search and and and then webhooks and
- 16:38the first doc will take you there again.
- 16:41Will you get this right the first time?
- 16:43Is this painful? Will you have to vibe code a few
- 16:46things if you don't know how to code 100%?
- 16:49But if you do this, most people don't know how to do this, I
- 16:52have no doubt you'll be able to build cool stuff, right?
- 16:54So I'm very much interested in how do I turn these cool
- 16:59automations or really agents into something that a user can
- 17:05use, right? So This is why you can maybe
- 17:08call this vibe coding plus vibe automating.
- 17:13So vibe coding plus vibe automating, you take something
- 17:17you built in NEA. I will admit, as a developer, I
- 17:21like building these myself, right?
- 17:24But I still learn how to use the tools because again, I want to
- 17:27be up to date with the times. But if you're someone doesn't
- 17:30know how to code, you take some of these automations and then
- 17:34you take a vibe tool I'm BIOS, I'll say tempo or any other vibe
- 17:39code tool of your choice. I would really learn in building
- 17:43something combining these two right?
- 17:45I think you start to build applications that are more
- 17:49useful than your standard landing page, but you also start
- 17:52to see how these different components program how a website
- 17:56works, right? Like you don't have to be
- 17:58developer to know that, OK, every like functional app has a
- 18:02front end and a back end, right? The front end is the user facing
- 18:04side. The back end handles all API
- 18:07calls. You know, if a website has
- 18:09username, password and all that stuff, that's authentication.
- 18:13If I'm doing something on a website, I log out and I come
- 18:17back in and information is there, that means the website
- 18:20has a database, right? These things are not shouldn't
- 18:24be guarded as ohh only technical people should know this.
- 18:27I think it's very easy excuse. I think you can know base level
- 18:31how these things work and I guarantee anyone who knows even
- 18:35like, and This is why product managers, designers, they cook
- 18:39with these type of tools because they have technical knowledge.
- 18:42They're not technical people, but they have technical
- 18:44knowledge. So when you combine that
- 18:46knowledge with these AI tools, whether it be automation, vibe
- 18:49coding, cloud code, whatever it is, almost always those are the
- 18:52people who end up, you know, having the most fun and building
- 18:56the coolest things. The non-technical people, the
- 18:58product managers, Well, people with a kind of a base level
- 19:02technical knowledge, but not coders.
- 19:04I say not technical people in technical organizations or
- 19:08environments, yeah. OK, when it comes to building
- 19:14your average automation like this, what tools do you like to
- 19:17use for what types of tasks? And you know what, let's let's
- 19:21expand that out. What LLMS?
- 19:22What tools are your go to tools for certain tasks?
- 19:27Yeah. So I will draw a line.
- 19:30I have let's say a spectrum, right?
- 19:33Let's say from user facing cool Vivi to genius code difficult,
- 19:46right. Let's say that this is a
- 19:47spectrum, right. If I'm building anything where
- 19:52code is being generated or I feel as if the task is very like
- 19:57required some sort of calculation or whatever,
- 20:00anything difficult? I'm always going with Claude,
- 20:05almost always, right. And then on the opposite
- 20:10spectrum, let's say I want to build like a nice image
- 20:13generator, right, for people to generate cartoons, or I want to
- 20:17build a poem generator, or I want to build an agent that
- 20:21talks in slang, right? Almost always open AI.
- 20:24And if you use open AI enough, like I don't want to open my
- 20:28personal ChatGPT because it show my history and stuff, but my
- 20:32ChatGPT starts to talk like me. Like it will say bruh, but
- 20:37BURUH, because I talk like that, right.
- 20:41So like the model is sort of designed to be in a very
- 20:46psychopathic and weird way, like almost personable, right?
- 20:49Yeah. So yeah, it almost becomes my go
- 20:52to for like, I don't even want to be to that guy, be that guy.
- 20:55If I'm building anything B to C and I wanted to have a
- 20:59personality, I'll always go for something like open AI and for
- 21:03anything again, agentic, writing code, complex stuff.
- 21:09Binary Binary outputs. I'm going for something clawed
- 21:14and then Gemini honestly is like an All Star where Gemini can
- 21:19sort of be both, right? And there is a use case to
- 21:23Gemini where like Gemini's also the most cost effective where
- 21:28and maybe we can have another graph here just to paint this
- 21:33picture. One of the things that
- 21:35especially if you plan on building something that's going
- 21:37to be used a lot or scaled is something you want to think
- 21:40about early on. If we were to compare price,
- 21:44Quad is up here. Gemini is, you know, very cheap,
- 21:50and then open AI, depending on the model, is very right there,
- 21:55right. So there are a lot of things I
- 21:57would take into consideration in the initial building stage.
- 22:00I'm not really worried about cost, and neither are you.
- 22:03So I wouldn't really think too much about that.
- 22:05I think you should only worry about cost.
- 22:07When you know you've made more than $1,000,000, then it makes
- 22:10sense. Yeah.
- 22:11But until then it's just a matter of what type of app are
- 22:14you like building. Anytime I'm building anything
- 22:17user facing, building a basic chat bot, I'm always going for
- 22:21open AI. But if I'm building any coding
- 22:23assistance, anything that I feel like requires more agent,
- 22:26especially with cloud four 4.0, I'm going with cloud right?
- 22:31And then Gemini's like an All Star where like I you can pretty
- 22:35much use it anywhere. They have an image generator.
- 22:37They also have 2.5 pro, which is a great coding assistant.
- 22:40It's not as good as cloud 40, but it's pretty up there.
- 22:43So this is how like I sort of play it myself.
- 22:46Now for someone who is a beginner and you know, let's say
- 22:51you're building some sort of automation, you want to pick an
- 22:53agent, just go with open AI. Just yeah, just go with open AI.
- 22:57Almost always open AI should be enough for you.
- 23:01And I would go open AI and Gemini.
- 23:04Claude is on the more expensive side.
- 23:06I don't want people to lose money.
- 23:08So I would almost go always go with open AI.
- 23:11And if you're really like cost is a big thing, then Gemini.
- 23:15So that's kind of how I look at these models.
- 23:18This is a spectrum I think of in my mind.
- 23:20Where do perplexity and grok come into play for you, if at
- 23:24all? I'll be honest, like these
- 23:27models, like these three just do it well for me.
- 23:31Like I have not used grok in terms of the like the API.
- 23:35Now I've used grok like to like answer tweets to describe tweets
- 23:39only. That feature is great, but it's
- 23:40never been like compelling enough for like I wanted to use
- 23:43it like I wanted to use the model Perplexity.
- 23:47If I'm going to be honest, Open AI is like deeps research and
- 23:51search is just such a great product that I used to use
- 23:54Perplexity, but not anymore. So like, I'm satisfied with
- 23:58these tools, right? There isn't something those
- 24:01tools do or those models do that is not here.
- 24:03Now, if one of them launches a new model and it's exceptional
- 24:07and it's well and it does something better than these
- 24:09guys, guess what? I'll move over, right?
- 24:11Part of the game is being able to know OK, what's out what's
- 24:15poppin, which one's doing great great and then being able to
- 24:18move over right. Just because this is my set
- 24:20today doesn't mean literally tomorrow won't change because I
- 24:24will tell you before cloud four point OI was literally using
- 24:27just open AI and Gemini cloud 3.7 was not good in my opinion
- 24:32and it just felt like a waste of money.
- 24:33So I literally stopped using Claude and it was just Gemini
- 24:36open AI 4 point O dropped and I'm completely on the Claude
- 24:40train now right. So part of the game is being
- 24:44flexible and malleable to like all the new coming out.
- 24:48Are there any like low hanging fruit ways that people could
- 24:51make money with these tools? Like whether that's building a
- 24:54wrapper or starting as like an agency or services business, AI
- 24:58implementation, AI automation implementation business.
- 25:02What's top of mind for you nowadays?
- 25:04Yeah, like, actually, I had an interesting conversation at the
- 25:07office today. And someone was like, we were
- 25:10talking about, like, some of the new apps that have come out, and
- 25:13myself and a fellow engineer were like, yo, like, some of
- 25:15these things are so easy to build.
- 25:17Again, talking from an engineering mind.
- 25:19And we were shocked that one of the companies we were looking at
- 25:21raised so much money, Like, I think upwards to like, $10
- 25:23million for something that honestly would have probably
- 25:27taken us maybe two to three days to build.
- 25:29Right. And someone who was there said,
- 25:33you know, often times because we're all in this tech bubble,
- 25:36we feel like some of the stuff that's normal to us isn't
- 25:40impressive to somebody else, isn't valuable to somebody else.
- 25:43And sometimes it just takes Poppy out of that bubble.
- 25:46So I say that to say the best opportunities, at least that I
- 25:50think right now are those where you look outside of this tech
- 25:54bubble, right? Because in this tech bubble,
- 25:57look who you're competing with, right?
- 25:59There are so many bright minds and so many engineers and
- 26:03non-technical people and technical people.
- 26:05This tech bubble we're in is not the best place to start,
- 26:09especially if you're a beginner, right?
- 26:11So if you are not a teenager and you're an adult, you've had some
- 26:16experience or you work a job, you've been part of some
- 26:18industry, I would literally start there.
- 26:21What are some things in your industry, at your job, at your
- 26:24office or your company? What are some workflows?
- 26:27What are some tasks? What are some things that you
- 26:29dread? You would be surprised how
- 26:31receptive people would be if you solved problems that you
- 26:35genuinely understood, right? Let's say, for example, Chris,
- 26:39I'll throw out a free idea. I really wanted to build an AI
- 26:42bookkeeping app. Tax season is hell.
- 26:45Keeping up with receipts is annoying.
- 26:47I download all these apps, QuickBooks 0, whatever.
- 26:50They mess everything up. Everything is, you know, not
- 26:53organized. Well, my rent transactions are
- 26:56categorized as business like. It's a mess, right?
- 26:59So there's a lot of manual and tedious work.
- 27:02And there was a part that's meant I wish this app existed
- 27:05and maybe I should build it, but I didn't build it.
- 27:08The reason why I didn't build it because I'm not a, I'm not a
- 27:10bookkeeper, I'm not an accountant.
- 27:12So I would build something surface level and sure it would
- 27:15work for me. But imagine an actual bookkeeper
- 27:19and actually imagine an actual accountant, someone who's in the
- 27:22trenches, who knows the pain points, who knows the long
- 27:26tasks. I think if you are that person,
- 27:29and I'm not saying you have to be professional, I'm saying you
- 27:32just have to have a job, literally.
- 27:35And I'll talk to the folks who don't have a jobs to the
- 27:37youngest. For those people, that is the
- 27:40best place you should look. It's terrible, especially as a
- 27:43beginner to be like, what's the next?
- 27:46Like can I maybe build a model or try to do like something
- 27:49egregious and insane that's in this tech bubble that there are
- 27:53thousands of people who are in this space daily building it.
- 27:57But if you're someone who's non-technical, who is having fun
- 28:00using these apps, who's learned how to build, you know,
- 28:02something cool with these tools, start from industry in right #1
- 28:07the best problems to solve are your own #2 the best problems
- 28:10are to solve are the ones that you're in that industry, you're
- 28:13in that business, you're in that.
- 28:15Like that to me is the gold mine because you know the customer,
- 28:19you have the customer, you work there, you are there, you know
- 28:22the problems and you even know how to communicate, right?
- 28:24Because building it is Part 1, being able to market it as a
- 28:28whole nother behemoth on its own.
- 28:29And you can do that well if you're in that industry, right?
- 28:33So I would start there. My place, my job, my people,
- 28:38whatever that situation is that you're familiar with solving
- 28:41those problem solving. For example, I have a a friend
- 28:45who works with, works at a pharmacy and he's having fun
- 28:50building things for pharmacies. And it's not because he's a
- 28:54super technical genius, but it's because he's worked in
- 28:57pharmacies. He knows how annoying billing
- 28:59is. He knows how annoying like the
- 29:01software is, that where they register medication and all that
- 29:05stuff. He also knows all the compliance
- 29:06that's required, right? And these are all boring
- 29:09problems for someone who's not in that space to solve, but the
- 29:13for someone who's in the tech space to solve.
- 29:14If you're stuck in that tech bubble, what does solving
- 29:17pharmacies problems sound like? It sounds boring, but that's
- 29:20where the money is. If you can solve the problems
- 29:22for specific industry especially you're part of, you're golden.
- 29:26I love it. I mean, even if you're you're
- 29:28everyone's in an industry, but you're having a hard time
- 29:30thinking of problems in your industry, ask Chaochi PT like,
- 29:33hey, I'm a pharmacist. I want to build an app for other
- 29:35pharmacists. What are some common pain
- 29:37points? I would use deep research for
- 29:39this. I would say I want you to go
- 29:41deep. Like go to Reddit, go to message
- 29:43boards, go to social media, find all the gripes, all the the
- 29:47negative tweets, all the complaints about the
- 29:49Pharmaceutical industry and list them all out.
- 29:51For me. It lists out 20.
- 29:52OK, take the seven that you think are most relevant, paste
- 29:55those down and say, OK, now I want 20 more that are a lot more
- 29:58like these seven, so you can just keep refining it.
- 30:01Pretty soon you'll have 30 to 50 problems and all of them will
- 30:04ring true to you as someone in the industry.
- 30:07And then you can say, OK, I'm going to use.
- 30:09I'm going to take the top three and build wrappers from these
- 30:11three and distribute them to my friends and coworkers just to
- 30:15test and play around with and see what feedback I get and then
- 30:18iterate from there. Right, Yeah, 100%.
- 30:21And I think like you mentioned, like going back to actually
- 30:24using AI to do the research, right.
- 30:26I've actually done something similar where I go to church
- 30:29every Sunday. I serve at a church, I teach at
- 30:31a church. And you know, there's certain
- 30:33things I've realized that I'm like, oh, like I live in Canada
- 30:36as well. So for example, payments for a
- 30:39lot of charities at the softwares are either terrible.
- 30:42The percentage they charge is very, very high.
- 30:44And like you can't just use like Stripe, you have to build a
- 30:47website and stuff like that. So I was like, I'm going to
- 30:50build a payment solution for my church and I did that.
- 30:54If I wanted to turn it to a business, I could.
- 30:56I don't have time, but if I wanted to, I could.
- 30:58But you see, and I almost guarantee if I wanted to build
- 31:01that business, it would do so well because it's a problem I
- 31:06experienced and I know the industry experience.
- 31:08I built the solution and I'll price it much fairer than the
- 31:12competition. It's just a perfect formula.
- 31:15So I think starting where you are and speaking to your people
- 31:19and when I see your people, I'm not talking about race, creed or
- 31:21anything like that, but I'm talking about your industry, in
- 31:24your business, in your line of work.
- 31:27I am telling you the things that if you think in a tech bubble
- 31:30mindset, you always think this is simple, this is boring.
- 31:34But at the end of the day, first principles, people solve, people
- 31:39pay to get their problems solved.
- 31:40As long as you're entertaining people or solving their
- 31:43problems, you should be getting paid, right?
- 31:45So that's how I think of things. Don't be too tech inflated.
- 31:50I think sticking it to your industry and the problems in
- 31:53your space and solving that is is a great way to start.
- 31:56I like the back to your bookkeeping, your AI app for
- 31:59bookkeeping. It reminds me of Cal AI.
- 32:02And for those listening, Cal AI is an app that basically said My
- 32:06Fitness Pal has hundreds of millions of users.
- 32:08It's a nine figure company or it has $100 million plus of
- 32:11revenue. We're just going to do that.
- 32:13Except AI first, right, And that one sentence, that one liner,
- 32:18that model is like that, but AI first, right?
- 32:22Instagram was like, oh, Facebook, that's cool.
- 32:24How about mobile first, Facebook pictures forward, mobile first.
- 32:29We're not even going to have a desktop site at all.
- 32:31We don't even worry about that. And all these others, other apps
- 32:34in like the late 2000s, early twenty 10s, their whole model
- 32:38was just, oh, we're going to take that website and do mobile
- 32:41first. An actual desktop version of
- 32:43this may never happen and that's fine.
- 32:45So now it's just like, oh, we're just going to do that except AI
- 32:49first, AI first and foremost, AI foundational.
- 32:52So that could be QuickBooks, except AI first.
- 32:55That could be anything except AI 1st and QuickBooks, MyFitnessPal
- 32:59and all of those. They're going to integrate AI,
- 33:02but it will they will never be AI first.
- 33:04They'll never be built with AI first and foremost, right?
- 33:08And one thing I would add to that too is like a lot of the
- 33:10like the new apps like Cal AI and stuff that you mentioned
- 33:13that they have one big differentiator.
- 33:15And this is like I plan on making a video on this and this
- 33:18is going to be a Moat for a lot of people.
- 33:20And that's user experience, how your app looks and feels and
- 33:23notorious for two things. If you know me, Percy, my like
- 33:27IDE like the where my where I write my code, I change the
- 33:31theme every like 2-3 days. Browsers I change every 5 or 6.
- 33:36And the reason being is I love the place that I'm working or
- 33:42the apps that I'm using or the sites that I'm using.
- 33:43Whatever I'm using, I want it to look and feel nice.
- 33:47I say that to say that most people, most people, most
- 33:50people, most people, most people.
- 33:52If you focus on crafting a beautiful and elegant user
- 33:57experience that most of the time is almost enough to do well.
- 34:03You have like, for example, there are ticketing softwares
- 34:07like where you can like use like, you know, you can like,
- 34:10let's say you're running an engineering team and you need to
- 34:13track like tickets and who's building what and you assign
- 34:16tasks to stuff. There's stuff like Jira, for
- 34:18example. Jira has been around for like as
- 34:21maybe as long as I have and it's an ugly app.
- 34:25Yeah, yeah. It's it's ugly and but then you
- 34:29have something like linear which basically is JIRA, but look at
- 34:36that which one? JIRA accept pretty.
- 34:40That's their their business power pretty.
- 34:42Fast right like they're very thoughtful with every single
- 34:45thing that they do and I think in a world where everyone is
- 34:48saying oh, AI is coming for the devs, AI is coming for this the
- 34:52people who care about these little things right how the app
- 34:56is crafted how the user experiences the the features,
- 34:59what makes things pop what makes someone don't want to just use
- 35:02their app even when they don't need to I think those people
- 35:04went big yeah. UX becomes a Moat.
- 35:07I call it the sauce. I think UX becomes a Moat where
- 35:10if you're someone and homework for anyone who wants to do this
- 35:14is I would collect sites and apps that you wish you built,
- 35:19not because they made money. We'll worry about the money
- 35:21stuff later. No, yeah, they made money, but
- 35:23because it made you feel a special type of way for me, like
- 35:26I use linear and every time I use it, I'm like, man, I wish
- 35:30I'd built this. Like I wish I could put my face
- 35:33on this and say I built this and that's how software is good.
- 35:36So I think being able to craft and develop a taste for how user
- 35:42experiences should be, how apps should be, how sites should be,
- 35:46even as a non-technical person can be very fruitful.
- 35:49I actually think, Chris, there's going to be a time where even
- 35:52though the models are going to get better that and this has
- 35:55happened for years, but I think it's going to happen even more
- 35:57that a lot of non-technical people are going to hire, are
- 36:00going to partner up with technical people, right?
- 36:02I think we're still going to enter.
- 36:04We're still going to be in a world where you like devs are
- 36:06needed and we'll help supercharge things.
- 36:09But I think it comes to a point where the non-technical person
- 36:13will start to be valuable too if they take these things into
- 36:17consideration, right? But if you just you go to some
- 36:20app and you're like, build me no plan, no idea, no understanding,
- 36:27right? And again, this is not an
- 36:28excuse. The excuse isn't, oh, I'm not a
- 36:30technical person. You're not allowed to say that
- 36:32anymore. You can literally go on YouTube
- 36:35and learn whatever. You can literally go chat GVT
- 36:38and learn whatever, right? So say that to say we live in a
- 36:42beautiful time. Well, like they say, if you
- 36:44build it, they will come. And I think you would say if you
- 36:47build it beautifully, they will come.
- 36:50Like beautiful apps? They do it.
- 36:52I mean, look at it this way. Let's say you're looking for a
- 36:55Co working space in Canada. You go tour two of them.
- 36:58You go into one and it's like a little dirty.
- 37:00It's kind of gross and there's no artwork on the walls and it's
- 37:03just bare bones. You could tell that they cheaped
- 37:05out. They didn't put much thought
- 37:06into it. They just wanted to get the Max
- 37:08value per square foot as possible.
- 37:10And they did some things to kind of look nice, but you could just
- 37:13tell this is nice. And in a vacuum, if this were my
- 37:16only option, I'd probably use this because I need a Co working
- 37:19space. And then you go to the next base
- 37:20next door and it's gorgeous and it's there's indoor plants and
- 37:25there's furniture from Restoration Hardware.
- 37:28You're like, that's a $12,000 couch.
- 37:30I feel good in here. This is good energy.
- 37:33Like you're going to not just choose that one, but you'll
- 37:35probably pay twice as much to use that one.
- 37:37So everything I'm outlining is obvious.
- 37:40Like everyone would rather choose that one.
- 37:42But we don't treat our software like that.
- 37:45We just throw out something our MVP.
- 37:47And then if it works, then we leave it like that forever
- 37:50because if it's not broke, don't fix it.
- 37:53But like, look at Airbnb. They were designed first from
- 37:56day one and it's $100 billion company.
- 37:59Like that was their whole strategy design, right?
- 38:02We're going to have a beautiful design.
- 38:04And I think, I think it matters. I think it matters more than
- 38:06ever. And like, unless you are
- 38:10building something very innovative and game changing,
- 38:16maybe you can get away with having ugly UI.
- 38:19But we're entering a time where the expectations high.
- 38:21Like if I see a site and it looks like it was made in 2005,
- 38:26unless it's a government site or whatever, like, or like some
- 38:30like, you know, farm like whatever, like a hospital site
- 38:33or whatever, there is no excuse because they they have like
- 38:36compliance and all that stuff and or banking, whatever, that's
- 38:39fine. But any business, any SAS, any
- 38:43startup, like design is super important.
- 38:45So make sure that, you know, you develop that taste in that craft
- 38:49and it comes from, you know, looking at sites that you like.
- 38:53For example, I can go on the gum loop side and maybe find, you
- 38:57know, I like the little diagram they have here.
- 38:59And another thing you can do is actually go to popular like
- 39:03frameworks, like for example, there's magic UI, like this is
- 39:07like a component library. Like you can see these different
- 39:10designs. Yeah, sure, they're meant for
- 39:12developers, but again, just being able to develop that taste
- 39:17of like this is a nice animation, right?
- 39:19And then you have obviously the Creme de la Creme Apple side.
- 39:23Like what's beautiful about Apple?
- 39:24Why, like, what's their style? You see they have a specific
- 39:27font. You see that they're spacing and
- 39:30the letters and mental grid like I actually spend a considerable
- 39:35amount of time just doing this, admiring these beautiful design,
- 39:38elegant user experiences. There's another one.
- 39:41This is a crypto startup, but nonetheless, they're again, like
- 39:46a lot of people hate. I bet if you went to the site
- 39:49and then you just used it, you'd feel a certain type of way,
- 39:51right? So I think these things become a
- 39:54Moat and you don't have to be a technical person to develop is
- 39:57taste. And if you're betting on the
- 40:00fact that the models are going to get so good?
- 40:03That it'll be able to build stuff for you.
- 40:06I think being able to develop the part of you that can help
- 40:11guide the model to build stuff like this is important, right?
- 40:14And that comes through time. That comes through building
- 40:17stuff and breaking it. That comes through like, you
- 40:20know, testing out all these tools.
- 40:22And I don't hear anyone speaking this, but I highly suggest make
- 40:26some friends that are developers because I know everyone's
- 40:28talking about all developers are going to get replaced.
- 40:30But I'm telling you, you have you a couple of developers
- 40:33friends that you keep around you buy them pizza everyone in a
- 40:36while. It once in a while I am telling
- 40:38it will yield results not only from knowledge, but also from
- 40:41like, you know, maybe you partner up together and build
- 40:43stuff together or maybe they know of tools that you don't
- 40:46know or maybe they even answer or help things like the amount
- 40:49of like people that I know who are like cousin of mine was vibe
- 40:53coding and the tool he was using.
- 40:55I'm just couldn't get the API stuff down and he like spent 2-3
- 40:59days. So guess what I did, Chris, I
- 41:01hopped and not fixed the firm. It's a 32nd fix for me.
- 41:04It's easy, right, I think. And then he continued to build
- 41:07his app, right? So make some developer friends.
- 41:09I think it'd be very helpful. That's a good idea when it comes
- 41:12to beautiful design. Do you know of any AI tools out
- 41:14there where you can say hey copy and paste this website or copy
- 41:18and paste this app? Unfortunately, many of them will
- 41:23not give you a one to one copy. You will have to.
- 41:28This is a workflow, I would say if I were to do that.
- 41:30So step one is screenshot, right?
- 41:35Step 2 is obviously feed screenshot to AI, which AI build
- 41:41again, whatever vibe code tool you want to use it it honestly
- 41:45at this point in terms of design, does it like it doesn't
- 41:48really matter. You start to see that certain
- 41:51models have like certain design tastes or like leads certain
- 41:55ways. That's kind of up to you to
- 41:58decide, right? All of them.
- 42:00But you're going to get a result, and the result is maybe
- 42:0440% of what it actually looks like.
- 42:08The way you get the 40% to 100 is by going through every
- 42:14section of the site that doesn't look pixel perfect and telling
- 42:18it so. Say, for example, I take a
- 42:22screenshot here, right? Let's say I take a screenshot
- 42:26here and let me just pull the screenshot here and I drop this
- 42:32in the model. I tell it, make it look like
- 42:34this. Uh huh.
- 42:35I almost can guarantee you it will not.
- 42:38It will get the layout down. It will get product resources,
- 42:41pricing, customer blog, content. Contact now and get the buttons
- 42:45down. We'll get the text down, but it
- 42:47won't get it. It'll mess up some place.
- 42:49So what my job is now is to go section by section and start to
- 42:54tell it to fix it. So for example, something that
- 42:57you might not have noticed here, look how rounded this button is.
- 43:01This is not a square button, right?
- 43:02It has a little bit of roundness to it.
- 43:04So what you're going to do is you're going to, you can
- 43:06screenshot this again and then upload it, or you could just
- 43:09tell it the buttons aren't the same.
- 43:12Please round it more. So that's one part you can fix.
- 43:15And then you can maybe see that. Notice how there's a grade, like
- 43:20it gets darker here, right? It's not just white, like it's
- 43:23white then turns to grade and starts to get darker and darker,
- 43:26right? You start, you'll have to
- 43:28pinpoint Yep section where it does not look perfect and then
- 43:33make it perfect by telling it. Now here's advice that might
- 43:37sound counterintuitive, but hear me out.
- 43:41If you're building something, and I used to say this and I'll
- 43:46explain why I used to say this, but I'm going to say the
- 43:47opposite. If you're building something, I
- 43:50think the landing page should be the last thing you should focus
- 43:52on. I think design should have its
- 43:56own time. If you're thinking about
- 43:59designing features at the same time, especially as a
- 44:01non-technical person, I wouldn't do it.
- 44:03I think design deserves its own time.
- 44:05I used to tell people build a landing page first.
- 44:08And I think for very, very beginner people, I will say
- 44:11that. And the reason why I say that is
- 44:13the dopamine, right? It's.
- 44:14Momentum. Yeah, it's momentum, right?
- 44:16It's like you get the landing page done and it just feels
- 44:19great. It's exciting.
- 44:20Like you, you have this dopamine rush.
- 44:23It helps you, you know, to keep going.
- 44:26But if you're on the discipline side, you've done this before,
- 44:29this isn't your first rodeo. I would start to build, you
- 44:32know, the stuff that you're actually trying to build.
- 44:34And then you could put the landing page on top and focus on
- 44:36design, right? But if you're a complete
- 44:38beginner, you know, start with your landing page, get the dope,
- 44:42immediate, get excited. But I would not start with the
- 44:45design first unless I have no idea what I'm building and I
- 44:49just want to have a nice landing page.
- 44:51Yeah, beautiful. 2 questions #1 Can you explain in two sentences
- 44:57what MCP is and why people should care about it?
- 45:00The simple phrase I've been hearing is the USBC for LLMS OK?
- 45:07Meaning like my the new iPhones have USBC, my MacBook has USBC,
- 45:11that person's Android has USBC, right?
- 45:14We all can same use the same charger, we all can share the
- 45:18same SSD drive, right? It makes it easy for all these
- 45:22different devices to use that external functionality in the
- 45:25same way. What MCP allows you is it allows
- 45:29for LLMS to easily access external services, right?
- 45:35So for example, Linear, right? Linear is a ticketing tool,
- 45:39right? Let's say I have an LLM here, it
- 45:41could be Clod, it could be whatever.
- 45:43And I have Linear, I have Linear over here.
- 45:47Now, if I want my LLM to have access to Linear without MCP,
- 45:55let's just say MCP doesn't exist, I would have to go to
- 45:58Linear's API documentation, right?
- 46:01I would have to learn how their documentation works and then I
- 46:04would have to basically write some code and I have to, I'm
- 46:10assuming I have some back end. I'd have to write some code to
- 46:15connect to Linear's API. Not difficult, but what happens
- 46:20when you have 20 different things you want to connect with?
- 46:22There's many different APIs documentation, not everything is
- 46:26the same. It gets very annoying.
- 46:29MCP is a standard where third party services like Linear will
- 46:34follow a specific rule set on how to build their MCP server
- 46:39and the LLMS AKA the MCP client knows exactly how to connect to
- 46:44these servers, right? It's like the USB port.
- 46:47So the LLMS have the USB port Linear and all these third party
- 46:51tools are building their product with a USBC in it.
- 46:56So it's easy to connect and plug in, right?
- 46:59And for most people, this might not sound like a big deal
- 47:03because again, there are still LLMS that connect to APIs.
- 47:06But from a production standpoint, from building out
- 47:09these robust, like if we ever wanted a Jarvis, right?
- 47:12Like an Iron Man where you tell the assistant, oh, fetch my
- 47:15e-mail, contact this person, make sure to close this ticket
- 47:19and all that stuff to build it without something like MCP, it's
- 47:23a very robust job. Now, MCP isn't perfect, there's
- 47:26still a lot of caveats and stuff like the Anthropic team and
- 47:30stuff for figuring out. But it's to say that it's a
- 47:33standard for LLMS and it means that we're going to be living in
- 47:37a world where the LLMS will essentially have access to
- 47:41anything that has an MCP server, which at this point is almost
- 47:44every 3rd party SAS service out there.
- 47:46OK. OK.
- 47:48Last question, is it worth someone trying to learn how to
- 47:51use MCPS and NAN or 8020 of it is learn how to use Lindy and
- 47:58gumloop and zapier and you're good to go or do you think it's
- 48:02worth taking that extra time? OK, so my advice is different
- 48:06for different people. If you don't have a lot of time
- 48:09on your hands and you maybe have one or two hours a day, I would
- 48:16just be an avid user of these LLMS, right?
- 48:19Use ChatGPT before you Google it.
- 48:22Maybe ChatGPT at first, right? Get into that habit of using
- 48:25those tools. But if you are someone like
- 48:27maybe you don't have 4 kids and you don't have this big family
- 48:29that you have to take care of, I understand not everyone has the
- 48:32time, so I want to be understanding that.
- 48:34But if you have all the time, you work 9:00 to 5:00 and after
- 48:36five you're free. Use everything, learn
- 48:39everything. Maybe you'll like something,
- 48:41maybe you won't, right? Just listening to what I think
- 48:43you should learn is pointless because literally as soon as
- 48:47this episode airs tomorrow, one of the tools by launch a crazy
- 48:51update and maybe everything I've said changes.
- 48:53The one thing that I want people to take from this is we're
- 48:57entering a time, we're entering a new, I don't want to sound
- 49:01cliche and say industrial revolution, but we're entering a
- 49:04time where things are changing and being part of the change
- 49:08means learning everything. Some of it will fail, most of it
- 49:11will fail. Some of us will succeed, but the
- 49:13knowledge, the experience, they can never take that away from
- 49:17you. So use all the tools, Learn all
- 49:20the tools. You know, use the heck out of
- 49:23all the free trials. A lot of these tools have free
- 49:25trials. Take advantage of all of that.
- 49:27And most importantly, have fun learning, right?
- 49:30I think when you're having fun, you're not working.
- 49:35When you're having fun, it's not work.
- 49:37And when it's not work, you can do it continuously, right?
- 49:39So that's the biggest advice I would give to people.
- 49:42Have fun building using these tools.
- 49:45Man, that was beautiful. Thank you, Mike.
- 49:48I appreciate that. It was amazing.
- 49:50Where can we find you? Yeah, so I'm on X at Ross
- 49:54Mickey. I can maybe pop it up here.
- 49:56You can find me here. I troll every now and then.
- 49:59I also have a YouTube channel where I talk about dev, non dev
- 50:03stuff. Obviously.
- 50:04Chris, I really appreciate you bringing me on here.
- 50:06It was a pleasure and hope people learn something.
- 50:09Of course. Thank you.
- 50:10All right, what do you think? Please share it with a friend
- 50:12and we'll see you next time on the Kerner office.