Episode 11: All That AI BS: An Infodump of Facts, Fears and Futuristic Predictions

Lee Zuvanich: Welcome to Founder Problems Podcast, where today we’ll be discussing AI and all of the BS surrounding it. So more specifically, everything from the harms and evils of AI to why we’re using it anyway and what we think the future of AI might be. I’m here with my co-host, Sarah and Zach, who will be helping me unpack this very contentious and exciting topic.

Sarah: I am Sarah Schumacher and I help entrepreneurs and solopreneurs understand how to use and apply technology in their businesses primarily with websites.

Zach: And my name is Zach Oshinbanjo. I’m a former founder who has lately been spending his time in pre-development for housing additions and building large commercial garages.

Lee Zuvanich: And I’m Lee Zuvanich, a startup founder and agency owner that has been building AI systems for my platform and others for the last two years, [00:01:00] you could say I’ve been drinking the Kool-Aid watching for signs that I’m in Jonestown. 

Sarah: Ominous.

Lee Zuvanich: very literally though ’cause I’m guiding us into it. I’m like, let’s go guys.

I don’t know where we’re headed. let’s start with the impact of ai, our hopes and fears. Specifically let’s start with fears. What do you all think of when you think of ai, what’s first thing that comes to mind?

Zach: The first thing that I think of is moving towards a new technology stage. You start with something like the steam engine, you move into. Whatever’s after the steam engine, and then you move into this phase.

Now we’re moving into electrical vehicles and things like that, but I feel like technology, as far as AI is the next frontier as far as how the entire world advances into the next stage. We’ve already gone through industrial, we’ve improved manufacturing to a degree.

We’ve gone through this process. So I think AI is just the next step that’s gonna accelerate commercial development globalization in, in, in ways that could [00:02:00] be good and bad. But I feel like that’s where AI kind of land for me specifically.

Lee Zuvanich: That’s quite a picture.

Zach: Good.

Sarah: yeah. The main thing in my head at this moment, just ’cause I finished a newsletter about it, is the way it’s going to negatively affect things in ways that are currently unseen. And that’s just the idea of releasing something that is untested, that is not really controllable. It’s almost like a virus, in some ways. You have this thing that you’ve created that once it gets out into the world has an impact that you cannot plan for and see. That’s being jumped into full steam ahead without people understanding the implications of it.

It’s gonna take time for that to happen. And then when it happens, it’s gonna be so baked into stuff like it’s not, oh no, AI is taking away jobs. That’s the obvious thing people are talking about. And it’s that deeper layer underneath that is the thing I’ve been thinking about [00:03:00] lately.

Lee Zuvanich: Yeah. So I’ll share mine. First thing I think of is literally the robots. Maybe go do a little research, look up AI robot. I think it may be it was at Stanford. but they’re already able to program. Robots that can walk and talk and do basic household chores and respond to the user, like ingest being told to do certain tasks and then them just the way that a human would.

So I think of robots I think of armies of robots, and it sounds very dystopian, but we are quite literally already there. So that’s what I think of.

 Beyond the most dystopian fears, the first thing that pops into a lot of people’s minds do you have any more grounded concerns that are impacting your day-to-day usage? Like for me, I would say it does give me pause when I know that [00:04:00] saying please and thank you to AI apparently costs a lot of energy and water from the cooling systems that are being used to keep these mega processors cooled down while they handle these, amounts of information requests that everyone is making to them.

The water doesn’t stay cycled in the system like I thought when I did the research recently, it turns out it just evaporates into the air. that water and the energy that are being used, those are typically being taken from disenfranchised communities. So if there’s a data center being built near you.

Just know that’s the debate. It’s actually going to impact resource usage. If you can’t run your AC in the middle of the day in July, it might be because there’s a data center near you soon. So that’s maybe a more serious concern that I think about as I’m using AI anyway.

Sarah: That would be a good example of it affecting the system in a way that people don’t understand . If your AC doesn’t work and you’re like, oh, something must be up with a no, it’s ’cause [00:05:00] there’s no water cooling system available or the power grid is stretched thin or your power bill goes up locally.

That kind of thing is the stuff where I don’t think the average person’s gonna think about that. And that’s just one thing of many.

Zach: I think there’s always room for improvement, obviously. But I think you have an opportunity with companies, I think it’s called Cerebra, that’s improving tenfold, 20 fold, something like that. The energy efficiency of people who use artificial intelligence. So you even have virtualization that makes a server access a lot easier.

I think the first year that the combustion engine was more mainstream, people died or something like that. Or you could even go back even further and say something like, with horses, how many people have been kicked and killed by a horse? Millions. So I think it’s an unfortunate reality that over time we’ll get much, we’ll get more efficient.

Lee Zuvanich: few eggs to make an omelet. Huh? Zach?

Zach: I hate to be that guy, but I think [00:06:00] over time people will get way more efficient, we’ll utilize way less data way less energy less water evaporated. We have

Sarah: Oh man. See,

Zach: like that I think will get better, but it just takes a little bit of time.

Sarah: I gotta push back on that though, because corporations have to be incentivized to care about that.

Zach: Fair.

Sarah: Positive changes never happen unless they are legally regulatory requirements. You have to be more efficient. That’s the thing that concerns me about a lot of the stuff that’s happening is everyone’s oh, ai, it’s so cool.

Let’s build it into everything. Let’s give these people bajillions of dollars and all of this stuff. And no one’s really discussing that.

It’s the same as all of these corporate dudes using plastic for every single, water bottle and soda and whatever else, and then being like you consumers need to recycle more and you need to stop using plastic straws.

The problem is not us as individuals using a plastic [00:07:00] straw every now and again, it is corporations choosing to source all of the material in plastic at an insane rate. And so it’s a marketing campaign to push the responsibility of the problem onto the consumer to distract from the fact that they’re making these choices at the top.

So my concern is, yeah the water resource, like why is it not a closed system? Like why wouldn’t you, if that’s the case, and maybe some of ’em are Right.

Lee Zuvanich: to build. Yeah. Some of them are closed

Sarah: Exactly. And so they’re not going to do anything that affects the bottom line unless the local government is I’m sorry, you can only build a data center here if you insist on making sure that our communities don’t suffer lack of water.

Lee Zuvanich: to

I’m from Kansas City, as are you guys, and I think our listeners primarily are a data center is coming to Kansas City and it was lauded as we’re bringing jobs and if you do the research, I feel like our city got sold out. I didn’t see anything about negotiations to make sure that power and water usage would be fair and manageable and that, I think it’s a Google data center that they would be giving back to the community in different ways.

It’s

Zach: There’s two [00:08:00] coming.

Lee Zuvanich: minimal. Oh really? Yeah.

Zach: Getting Google and meta locations here.

Sarah: Oh, that’s true. Meta they’ve already built.

Lee Zuvanich: It’s who wants to see technology flourish in the Midwest, this is not that. is not Google and meta building offices where technologists get to go work together. This is just them seeing cheap land with people who aren’t guarding their resources. And I’m talking about our mayor like other legislators who should have been better at asking what does our city get? What are you gonna give

Sarah: That’s the problem is that the average political person has not the slightest idea how any of this crap works. These people are all dinosaurs anyway, they’re not understanding the implications of any of this and being like, wait a second, maybe we should have some safeguards in place.

No one’s asking those questions. Everyone just sees money. 

Lee Zuvanich: I feel like if you’re the mayor of a city and you can’t Google, why is a data center bad before agreeing to it, you should not be the mayor of that city. But we’re getting off topic. So

Zach: Yeah. It goes down.

Lee Zuvanich: balance us out? [00:09:00] Yeah.

Sarah: Doom.

Zach: Yeah. I think that’s doom and gloom. There’s always gonna be a perverse incentivization for adoption of new technology. I think if you even look at some of the things that have happened within data management, patient data, there was some legislation that put forth subsidies ’cause everybody was like, oh, it’s so expensive.

If I make everyone’s data interoperable and I give patients complete access to their information, it’s gonna be so expensive, I can’t do it. So the government did their favorite thing to do that helps to incentivize corporations to do quote unquote, the right thing subsidies. So they introduced subsidies that allowed businesses to offload the entire cost of transitioning to an interoperable system so that they could all switch all the data platforms over to a singular system so that whether you’re in Kaiser Permanente in California, or you’re going to St. Luke’s or something in the Midwest, your information could be exchanged. Whether it’s easy or not is an internal process, but the information should be in the same [00:10:00] format.

There’s always top goal of any city Yeah. The welfare of its people. Yeah. But the other part of that is economic prosperity and development. E-D-C-K-C is the body within Kansas City that negotiates and understands how do you bring these large organizations here so that we can our job output, improve our involvement in the technical technology ecosystem, and don’t get left behind.

But those are some of the things that happen unfortunately, or fortunately.

Sarah: Because everything is so positive around AI, the average person listening to us probably is a little more in the know about the AI space than the average dude on the street that’s Chad, what now?

Whenever people are overly positive about things, I feel the need to push back, this contrarian impulse to be like, hold up let’s look at the other side.

So maybe I’m leaning negative in a lot of ways because I’m not seeing people talk about that enough. It’s just all this, oh my gosh, I vibe coded a whole [00:11:00] app. Okay. Not really, but All right. So yeah, so we can get into the positives of it, but I think opening with, we need to be realistic about the dangers as well.

I think that’s important.

Lee Zuvanich: And a caveat that we’ll get into more soon is we’re all using it. We’re all

Sarah: Yes.

Lee Zuvanich: We are using it. I’m using it daily. I’m building systems that

Sarah: Yep. Yep.

Lee Zuvanich: But before we get into that, does anyone have anything positive to say about ai?

Zach: I think it’s gonna allow certain routine ordinary tasks to become way easier to execute and do at a larger scale. An excellent example, government efficiency. So the government is the largest employer in the world.

They have troves of information, money going out, data going in and out that used to, require 50,000 people to just at CSV files over and over to look for inefficiencies and things that just don’t make sense is the perfect [00:12:00] implementation for ai.

Read this 5,000. of transaction information and tell me where it’s messed up. You don’t necessarily need to spend taxpayer dollars on something like that. You don’t need an entire department dedicated to something like that.

Sarah: I think the biggest positive that I see in it is that it levels the playing field for small teams and even solopreneurs to have access to and build the same type of technology as larger businesses. So you don’t need a ginormous corporate team with a group of people that are specialized in X, y, z, to churn out whatever you can build things that enable you to produce at a higher level.

The idea of building a custom GPT to do obnoxious tasks very quickly while you’re working on using your brain power for things that really need your brain power, that to me is a huge positive. You as an individual solopreneur can have automations running in the background.

You can have these tools that you build and [00:13:00] use that are unique to your business, that allow you to be more productive. Your output can skyrocket without you having to bring on a bigger team of people. I work primarily with like solopreneurs, consultants, coaches, people that have like deep expertise.

They’ve worked a job. They’ve had a career, they really know their field and they’re like, yeah, this isn’t stable anymore. Or I just got Doge, or I just got laid off, or I want to be my own boss. ’cause my work environment’s toxic, whatever. So a lot more of these people are choosing to go out on their own.

And when they’re able to do that and then have access to some of these AI tools, it makes it a lot easier for them to compete. I see more and more fractional positions for knowledge workers that are at a high level. I think it’s gonna be more difficult for junior people you’re at a senior level with your expertise and you know how to take advantage of some of this stuff, you can really turn out some quality work without having to have access to an entire marketing department or whatever. I.

Lee Zuvanich: Yeah, I would agree with that. I was able to create graphics for a walkthrough for a [00:14:00] client recently. They wanted to see some visuals and wire frames for a future app that they wanna build. And I was able to use chat GPT to create some really beautiful graphics that would’ve cost me a week and hundreds or maybe even thousands of dollars working with a designer. Now, on the other side of that, those designers that I would’ve used could probably really use the work. But if they’re not also using AI to do that more low level stuff and using their expertise like UX research and UI design knowledge to go deeper ’cause I can’t use chat GT for those things. Then they are in a dying field.

This is why this is such a contentious topic. For me, that’s a positive as a business owner for them, it’s probably a negative. I don’t know.

Sarah: It’s an issue with junior level people in any design, writing, whatever, because it’s gonna separate the mediocre players from the people that really have the ability to think big picture and be strategic because you still need the strategy. It doesn’t matter if you have, chat GBT to write for you.

If you don’t [00:15:00] understand how to write your website copy in a way that actually sells to people you’re not gonna end up with anything that works. And that’s high level strategic stuff. So if all you do is push pixels around all day and do generic design and you’re not thinking at a high level, yeah.

It’s not gonna be great because people can get to that point on their own. honestly, our logo would be a good example for the podcast. We did a bunch of, I don’t know, whatever you used, but AI generated things be like, oh, I like this style, I like this color palette. Here’s kinda what I was thinking. And that was really helpful because I’ve worked with clients for years on this kind of thing, and most people cannot tell you in words.

Can you send me examples? What are the logos do you like? I need to see something because how you say it in your words may not be how I interpret in those same words.

Lee Zuvanich: Of revisions. If you can imagine us doing it together as we’ve done in the past,

No, not that. No, not that. I don’t know how.

Sarah: Yeah, exactly. And we completely bypassed that entirely. We didn’t use AI design the logo. I worked off of the layout concept of okay, we want this angle and this thing to work, but the [00:16:00] amount of work to still create our logo after that, I spent more time on the logo than I did on the entire website.

I can say that because it takes all the refining and the little elements and like redrawing lines and whatever, and then typography style stuff with, the founder problems, typography. That’s a lot of work to do. You could get part of the way there potentially, but you still need someone with a design eye to refine it.

It’s not gonna look finished. We’re not at that level yet. Will we be at that level? I don’t know. But you still need higher level people.

Zach: The bigger threat isn’t necessarily creativity. Because you can still utilize AI to create a lot of things. The bigger piece where people were gonna be in peril or have some issue is if your expertise isn’t your service isn’t what you do. If what you’re

Sarah: Yeah.

Zach: execute is completely just ground floor out proposal, light work, just pitching things out, yeah, it might consume your work, but you have the infinite opportunity now to augment [00:17:00] everything that you do a way that makes you more competitive in that space.

If you have exceptional experience in a very niche thing, like you’ve done 7,000, logos, what you can put together is gonna be able to articulate different angles. What about this sentiment? What about this feeling? ’cause AI isn’t gonna be able to deliver to that degree.

Sarah: Yeah. I think the takeaway for this, you need to uplevel your skills. If what you do is in the knowledge workspace and you feel like you’re at threat of being replaced, you need to uplevel your skills.

You need to become the expert. You say 7,000 logos, I would say 7,000 logos specifically for food packaging brands.

Zach: Yeah.

Sarah: The more you can kind of niche in on something, oh, I wanna use this color palette. Eh, you’re a food company.

You don’t wanna use that color palette for food. Like that kind of stuff. You need to lean into that. People really need to uplevel what they know and become the expert at this point. And if you’re not doing that, it’s, yeah.

Zach: Look at it like, what is it? Cybernetic [00:18:00] enhancement is a little Star Wars poking here is there’s one of those planets. There’s a people going around and some people take their hands off or whatever, robotic hand, somebody their eyeball out so they can like scope in. You gotta augment, cybernetic enhancement. That’s where we’re at.

Lee Zuvanich: I feel like we’re talking to

Sarah: Pretty much

Lee Zuvanich: who are like, oh, I just don’t understand computers. And that other office said that we have to get ’em or we’re gonna get left behind. But it’s so scary and they’re like at their clipboards and their fax machines and their printers and then the people who understand where everything is

Sarah: pagers.

Lee Zuvanich: Yeah. Yeah. They’re ordering their desktop computers and people are walking into the office going. Oh, we gotta learn this now. Great. This is gonna take my job.

Sarah: I think that’s a good parallel though, because there was the printing press, there was radio, there was tv, there was computers on the internet, and we’re at the next phase of that. That is even more disruptive than any of the stuff that’s come before it. That’s absolutely where we’re at is another shift.

Lee Zuvanich: Yeah. And not to disparage or [00:19:00] undermine the very real implications of what that kind of shift causes. Like people are losing their jobs, they will

Sarah: Yeah.

Lee Zuvanich: And I think that we’re in a climate in this country right now, and also globally probably where like Zach was saying, legislators are like, gotta crack a few eggs to make these

Zach: Okay.

Lee Zuvanich: We gotta innovate, we gotta stay ahead. We can’t, be left in the dust, so let’s let that data center. And yeah, people are gonna have to stay scrappy because not only are people going to lose their jobs, but I think that overall whole industries are being changed right now. And it’s gonna be forever. It’s irreversible in my opinion.

And we could go one of two ways. We could end up in this utopia where we say we already have the resources available in this world to cure world hunger and better mitigate war and disease and famine. And with AI helping us with the data analysis and a million other factors, we can get more efficient.

We can replace a lot of menial labor. We [00:20:00] can free people up to be creative. We could do a global living wage or a national living wage or whatever. There’s so many things we could be doing with technology as it frees us up and we get more sophisticated in scaling operations so that it’s also efficient. But the flip side of that coin, if you’re in a capitalist society, is those profits are gonna go to a select few people. As they get more efficient, the wealth will just be more and more consolidated. And that’s what we’re seeing happen in the us which is why there’s a lot of unrest. I don’t want this to become a political discussion, but I think that legislation, as Sarah was saying, is maybe how we pick our path there. It’s not about do we use AI or not. It’s about how is our city and our state and our country going to respond to the fact that we are absolutely getting more efficient and the wealth that this creates is going somewhere? Where’s it going?

Sarah: Yeah, I think the average person needs to be informed of that. That’s kinda what we’re doing here is know what’s going on so that you can say something [00:21:00] about it.

If you see something, say something.

Lee Zuvanich: And I’m

Sarah: But AI.

Lee Zuvanich: and I wanna be part of the solution. I love ai, I want to use it, and I don’t wanna be contributing to the downfall of society as I do.

Talking more about usage do you all wanna share how are you using AI day to day in your personal and professional lives right now?

Zach: For personal, I would almost just categorize it as lazy. Like lazy things that I don’t want to devote brain power to do. If I look in the fridge and I see celery, zucchini hamburger meat and potatoes, what I make with this, I don’t want to think, I don’t want to search.

This seems like something that’s super easy. Actually in hindsight, given what you said about water earlier, sounds terrible to use it for that. But I use it for a lot of things like that. Or even vacation planning. Best places to go in Chicago or something like that, wife, daughter, give the age going between this time, this place, what are the top 10 things to do? We don’t like doing X, Y, Z this is what we want to [00:22:00] spend. Gimme a list of 10 things that would be great to do during this time. And you know what? While you’re at it, build me an itinerary that allows me to travel between each of these things using public transportation so you can spit out a complete I itinerary and agenda. Go to the children’s museum, go to the Maggie daily, go to Rainbow Cone, and it all sequence all that out rather than me spending six days looking through Almanac and all this other archaic stuff. I can just use AI for something like that.

As far as a business case, I use Claude for a lot of almost consulting type work, where if like I have an idea or a pitch that I’m working on, I go in there and I feed it context.

Pretend you’re a 50 year veteran working in government and you do this and I give it enough context and so I can have it advise me. So somebody who should probably be concerned it’s like these. Mega consulting firms like McKenzie that may have some issue there.

And then perplexity is this Google that’s boosted.

It allows you to [00:23:00] search things, get some sources, and then I use some of the other main ones. But the two main ones that I’ll mention here for brevity are Otter, allows me to record and take transcription notes that I wouldn’t be capable of doing otherwise. at the end of the day, get a summary and then reclaim, rather than doing this back and forth with everybody, Hey, can you meet Tuesday? Can you meet Wednesday? Can you meet Friday? I say, this is the availability and it finds the time and gives me transition time in between.

So those are the main ways that I use it.

Sarah: So I use Fathom instead of Otter for work. Which I think is similar, but it auto records meetings and then it creates a transcript and then it will also spit out like checklists of here’s what you said you would do at this meeting. That’s really nice. So that’s one I actually did not even have on my list. It auto joins my meetings and that’s where I think AI is useful. It just runs there.

It’s like your internet connection. [00:24:00] You don’t think about it as long as it’s working. What else? I use Chat, GPT and Claude. Both. The open AI API integration I’m already using for some things. And it’s just it’s the biggest one, it’s the most common, it’s the most accessible. So it makes sense to build like a custom GPT, which is like a custom chat GPT bot essentially. There’s some stuff I wanna build that I’m trying to decide if I would rather go with the Claude Project version.

So I’m in the middle of researching that. ’cause my concern with a lot of this stuff is around privacy and the ethics of the company and everything else. And yeah, OpenAI stuff is, nice and convenient, but part of me is being a little bit more of a stickler, I think about which one I wanna use.

But anyway, so I’m using both right now . The number one thing that I use AI for is coding. And I don’t do a lot of coding typically. I don’t generally need to, and I should be doing even less of that.

But it’s gonna be something like I need a custom PHP snippet or something like that for a specific function or a CSS thing. And it just is so much faster because I know [00:25:00] exactly what I need and how to prompt it

Lee Zuvanich: Wow.

Sarah: Hey, write me a snippet that does this, and then I just drop it in and move on with my day.

So an example of this that I had as a client in the last month wanted to add a last updated by date to the bottom of all of their website pages, which is not a typical feature.

It took me 26 minutes to do that. That’s it. And that was writing a snippet. Installing it, testing it, refining it. ’cause then I had to make sure that it showed up and styling it too. So it’s not just adding the functionality, it’s also styling the functionality and making sure that it shows up above certain elements at certain times.

Blah, blah, blah. 26 minutes start to finish is what I build the client for that. That’s it. And that was purely because of using AI. AI makes it easier for people who have a wide breadth of knowledge to do more stuff themselves. So I can do a little bit of everything. I don’t really code, I’m not really a coder, but I know enough to be dangerous and that gives me the ability to do a lot of stuff myself if I want to.

So that’s how I’m using it a lot. [00:26:00] Coding. I use some image editing things like the built-in ones in Photoshop. The generating image stuff as an artist really bothers me ’cause it’s all been trained on our work and I’m not a fan of that.

What I do think is handy is using it for things like, I had a client who had a photo where she’s holding a water bottle. I used AI to take out the water bottle. So that’s just her hand now. And that would’ve been, an hour of manually photoshopping this stupid thing.

And so I just wouldn’t have done it before. ‘ cause I’m not gonna charge for that. That was perfect use case in my opinion. For a code snippet, I needed a really long list of zip codes formatted a certain way.

So I just said, Hey, format these done. I would’ve had to type like 167 quotations and commas and it would’ve been stupid, a waste of 10, 15 minutes of my day. That is again, perfect. Just fix this for me. Will you , so it’s small things like that. I do not use it for writing.

I do not use it for writing. I

Lee Zuvanich: I would never, yeah.

Sarah: Outlines, yeah, I’ll do an outline or I’m working on one that is a proposal writing thing because the way I write proposals, oh, in a number of pages and [00:27:00] features we discussed, blah, blah, blah. So that’s kinda how I’m using it at work.

I don’t really use it too much for personal. I have done it for trip planning stuff like Zach mentioned. I plan on building an ai meal planner. Based on how I do meal planning, so like a custom thing.

That’s probably it. Honestly, I mostly use it for work.

Lee Zuvanich: Yeah, so I’ve been using primarily chat GPT, although every time something new comes out that people are excited about, I go try it. always

To chat GPT. not sure I wanna do that long term. If Claude could catch up, then I would like to possibly switch.

But yeah, everything from I created a homework helper for one of my kids so that it would just step them through problems without giving them solutions. I’ve created custom chat bots that answer really technical questions for my clients when I don’t have time. Just what do we do to the system? It’s security operations. You’re not gonna understand any of it, but here’s the list of what we did, how long it took, and now here’s a chat bot that will explain exactly why it was so important and what will happen if we [00:28:00] don’t do it. Integrated my Clickup platform all my tasks and brain dumps into a custom chat GBT bot which they provide the tools for. This was before I started getting better at coding recently. But they make it really easy, right? They’re in the platform to do even API integrations, and it was able to. Talk to me and say Hey, here’s your schedule for the day. Here’s what think is the highest priority. You have these things prioritized. I think you forgot this one. Let’s make sure you get these done. Check in again when you get this done. And it was like an accountability coach, project manager, all in one that really I thought was awesome. And I started to try to get into vibe coding, which is where you just like code without understanding the code you’re writing basically, and you let AI write the code and then, just enough to implement it and push it out there and stitch it together. but you don’t know enough to fix it when it breaks typically.

So I started trying that before I got into [00:29:00] this coding bootcamp I’ve wanted to get into for eight years. And I’m excited to be in, I’m in the middle of it right now. I hated that I was coding and I only understood maybe. 30 to 40% of what I was looking at, and if I needed to debug, I was kinda lost. And AI would start looping what if we try this? What if we try this? And it was none of it was working. I knew I could just then hire someone to help clean it up and that person would be like, really, dude? So we’re in this really interesting moment right now where vibe coding is a joke that only produces broken apps.

And if you go to the subreddit, vibe, coding or whatever, see all these people who are like, oh my God, I launched the startup. I built this app. It’s amazing. I didn’t even have to pay anyone. Look, check it out. And then 24 hours, 48 hours later. we got hacked, guys. We got hacked. Oh my

Sarah: Hacked.

Lee Zuvanich: down. They’re such easy targets, they don’t even realize there’s no security foundation, there’s no best practices, there’s all this exploitability. So I stepped away from that, [00:30:00] went to the coding bootcamp, and at first they let us use ai. And then after the first week, they were like, all right, no more ai your code is too clean.

You all need to make mistakes. You need to struggle a little bit to understand this stuff better. And I cut all the AI out and I hated it. ‘Cause it was almost like Zach was saying it was a part of my brain. It was like augmenting A DHD also, it was helping me dive into a project instead of overthinking it. And I had to stop doing that and sit and stare at a blank screen and think okay, I need to make a list. What is that? That’s a loop. That’s a map. That’s an array. Whatever it is. Instead of letting AI just be like, oh, you use dot map and you do this and that, and I wouldn’t have to do any recall and it wasn’t sticking. So I’ve learned through this experience that I’m in the middle of that there are just some things that you really shouldn’t use AI for if you want to be more and invest in yourself. Honestly, just [00:31:00] like Sarah was saying about personal writing, I would never use it to replace my own The writing that I do on LinkedIn, like the reason people hate LinkedIn is because it’s all a bunch of self-serving drl.

Sarah: It’s very obvious

Lee Zuvanich: and it’s just

Sarah: when it’s when AI’s used for it.

Lee Zuvanich: with AI and I don’t know why I just started blowing off. I think I started blowing off steam on LinkedIn ’cause I thought, no one’s here anyway.

No one’s gonna read this. And I talk about business and nobody else wants to read that, so I didn’t feel like it was appropriate for another platform. I had a lot to say about these business things I was dealing with, now I really enjoy doing that and I have a specific and unique voice.

And even if I trained AI to do it, I wouldn’t get the personal benefit of getting to rant and blow off steam and I love to write as a creative process for me, and I would be stealing from myself by doing that, and I would be just contributing to more of that low level crap everyone else is turning out. I think that people can sense that and that it does make a

Sarah: A [00:32:00] hundred percent. It’s so obvious. Writing is how you learn what you think. It’s the process of writing that helps you clarify your thinking. If you can’t write it, you don’t really understand it. So if you are outsourcing that piece of things, you’re not really deeply learning things.

So it’s the actual process of riding that really matters. And that’s the thing I think is really dangerous about people. Oh, I’ll just have AI write this for me.

Something you said made me think of the the analogy of maybe a kid that’s learning to ride a bike for a first time, you’d be like, Hey, here’s a dirt bike.

Here’s a motorcycle. I’m gonna give you an engine. Have fun. Maybe they need to work on balance first. Maybe get to taking off the training wheels before you’re like, here have this Ducati, yeah, no,

Lee Zuvanich: off and scrape their knee and fear

Sarah: Yes.

Lee Zuvanich: ’em on a dirt bike.

Sarah: Exactly. And that’s what AI feels like in a lot of ways to where if you were a brand new developer, and you’re like, I’m learning this for the first time, it’s gonna be a handicap.

You’re not really gonna deeply understand all these things. The example I gave of the code that I wrote for the last updated date. That was really interesting because when I prompted that initially, [00:33:00] chat, GPT took it in completely the wrong direction like way off in left field.

And I had to say no, start over. Forget everything just start over and do this. And the only reason I was able to say that was because I knew what I needed to do because I understood what needed to happen and what parameters need to be used and all this stuff. So if you don’t understand that, and it does, like you said, it gets stuck in loops a lot, it’s not gonna go well for you.

You need to understand at a core level how things work before you start doing

Lee Zuvanich: Totally.

Sarah: ai. I do think it’s interesting for using it as like an educational tool though, because you could be like, Hey, this isn’t working. Can you explain why? And so in some ways there’s almost like some hands-on instruction.

I think it just really depends on your mindset and how you use it. Is this a I’m gonna have this do my job for me, or am I gonna use this as a lot of people, I seem a lot of people say thinking partner. Use it as a thinking partner. That’s an interesting idea if you’re learning something.

’cause it can help you break down concepts and learn the better. But that’s only if you slow down and do it. So it’s interesting. It’s such a double-edged sword in so many ways.

Zach: The writing piece is important. [00:34:00] I went through a, I don’t know, like a six month phase where I tried to be a super posty thought leader. and I think I tried to do that organically, maybe four times.

And I was like, this isn’t my natural field of practice. So AI button. And I just started doing that. And I think the first thing you notice is there’s certain syntax and the AI voice that it likes to do. And so I probably did. Maybe like 10 of those, but that’s simply because I just, I don’t like to write. And so if you’re trying to become a poster child or like the person who is the expert in a particular thing, then using your organic voice probably has the most value. But I think there’s other cases, like what I was trying to do, my expectation and thing that I’d been told is. Nobody knows your building if they can’t see it. Like it’s the expectation that everybody gives you. So I’m like, yeah, post every day. I have an opinion on this. I didn’t care. This is what people want to [00:35:00] hear about and this is what they’ll engage with me if they see that I talk about it.

That’s a dangerous thing. ’cause at a certain point I put into Che Bt or Claude, I was like, gimme 52 posts. I just want to be good for the rest of the year. I don’t want to think about this ever again. The post got more and more horrendous with each post. It was just terrible.

I was like, this sucks. And I think I just stopped doing it all together.

Sarah: Okay. So I would say like the way to use that is saying, here is who I’m talking to. Here’s what I wanna talk about. Here are some sub subjects I’ve covered in the past. And you could have it spit out ideas or outlines. You could create a content calendar of that, but then you write the content. I could have it do that right now for things and be like, oh yeah, that reminds me of this story of this one time.

And oh yeah, it’s so frustrating when I see people do that. So you can use it as a tool, but you don’t outsource it completely. So I think there’s this line, you have to walk where you need to understand the importance of developing your own voice. As an artist, it takes a lot of trial and error, painting, drawing, [00:36:00] whatever you’re doing to develop your own style.

And you have to work really hard on that. And it takes time. That’s why it’s so frustrating to see the Miyazaki stuff get ripped off the studio, Ghibli stuff, get ripped off. That is so recognizable because he’s been refining his. Style for longer than I’ve been alive. That’s why you recognize it.

It takes time to do that. It takes time to develop your writing style and your writing voice. I would say the same is true for music as a musician. If you wanna be able to play like jazz music, you have to understand, first of all, like the technique and be able to position your hands and do all the physical things.

And then also understand things like chord progressions and how everything fits together in music theory. And only after you learn those fundamentals could you jam with a jazz band. But the effort that it takes to get to that level of being able to do that supposedly effortlessly is years, right?

So what’s happening with a lot of people is they’re bypassing the years of developing an artistic style, of developing a writing style. And then so far there’s no way for people to play an instrument, like a robot hand or something [00:37:00] that doesn’t for you,

Zach: I guess they have

Sarah: like a cybernetic hand maybe.

Zach: airports

Sarah: Oh yeah. Yeah. You just pretend that you’re, yeah, maybe one of those. But yeah, it’s just really important. You can’t skip that. And so that’s, to me, a little frightening to think about younger people now, completely skipping those things. ’cause there’s a lot of fulfillment you get from that too, of being like, I’ve worked really hard on developing my voice or my writing style or my art or whatever.

And

If you skip that, it’s just really hollow.

Zach: yeah,

Lee Zuvanich: Yeah. Yeah. I would wanna hear

Sarah: Yeah.

Lee Zuvanich: if you decided to go back to posting, first of all, I would say like practice of using AI for things and then them not. Being high quality is unfulfilling for the person using it as well. Like you stopped posting and the discipline of just bravely putting yourself out there and like your Zach, your personal brand of humility mixed with if you get on a topic you know a lot about, then you are confident. But you [00:38:00] are a really humble person in general. And I think as a founder who understands all the things you understand, you’re. and breadth of knowledge is pretty impressive. I think that’s really unique and I would want to hear that voice. I don’t wanna hear AI tell me what Zach is building.

I want to hear you be like, I did this and I fucked it up and I really don’t know if I should keep doing it, but actually I do wanna keep doing it because I really care about these people I’m helping. Like you have a journey that you’re in the middle of. And I love hearing you personally talk about it. you ever decided you did wanna write, if you wanna develop the enjoyment of writing, that’s what I would be gravitating towards. If I knew that AI was telling me what you were working on I would not be as interested. I might be like, let’s see what the AI has to say about sacs work today. But like a, I want a bullet point summary I’m your investor. I don’t wanna hear like pontificating ’cause that’s not you. And I think that’s where people go wrong with ai. Like use AI for what it’s good for. Give me a summary. Don’t give me like ai. [00:39:00] Thanks. I wanna hear what the human person that I can relate to, ’cause they’re also going through this struggle.

I wanna hear what they think.

To lecture you, I’m Sarah and I both love to write, so like a nerve for us. 

Sarah: It’s a different, angle too. I like to educate people. So my stuff is typically don’t make this mistake or watch out for this, or whatever. And that’s just because I love teaching people and helping people uplevel their skills.

So that’s my angle. But yeah, it’s, everyone’s got a different version of it. Some people have really wonky senses of humor and it’s really entertaining. One of my favorite writers on the internet is just, his turns of phrase are just beautiful and he lambasts the tech industry and it’s so unique and I love it but that’s very much him. So everyone just needs to find like what their thing is that, just showcase your personality and that takes work to get there. 

Lee Zuvanich: Agreed. Just to wrap things up let’s finish this out with, do you think this is headed? Like 10 years, five years, maybe even next year? Where do you think we’re gonna be with ai?

Zach: In five years be more ubiquitous than it is [00:40:00] now. It’s already immersed itself in almost every aspect of our lives, but you’ll see it starting to contribute to those Jetsonian type things where, like you mentioned figure, they make a robot that’s going through sorting dishes, washing some of those repetitive household tasks.

I think we’ll get closer to getting a lot of that within the next five years. would almost imagine let’s just say of jobs will be augmented in some capacity, shape or form by the implementation of artificial intelligence. And I think we’ll also start bending the curve, I would hope, in the next five years towards sorting out a lot of the negative effects, like using less water, using something else, or if we’re using AI to build supercomputers or things like that, we address the challenge of limited resources that we’re up against, or, landfills or filling up to the brink. Uncured diseases and things like that. So you can, I expect within five years we’ll [00:41:00] start bending the curve. It’s gonna be everywhere and will reduce the damage that it has on society at large.

Sarah: Okay. I agree with the 75% and it being part of everything, like I think it’s just gonna be ubiquitous. It’ll just be baked into everything. On the one hand I don’t agree on it being better. I think maybe people start to realize and draw attention to it. I don’t actually think it’ll get better in the next five years.

I think it’ll get a lot worse where it gets better if it ever does. But the other angle to that, that I see is there will be a backlash and a reaction to AI where we’ll see more of a return to a lot of analog stuff. So people will start to realize that they can tell the difference between a brand created with AI and a brand created with a professional brand designer.

And so it will become like this it’s almost like a luxury good. Oh, this wine was. Hands stumped by children in the, I don’t know, whatever, like some ridiculous, like really expensive, exquisite hand bottled by peasants in this remote village and shipped to you on [00:42:00] mules.

I don’t know, right? It’s gonna be like that. But for like creative industry type things that have been overtaken by ai. And so it’ll almost be like, oh yes I use this company that refuses to use any AI in its work. Or just as honest about you need to work with a real human being. Or we don’t use any chat bots.

If you call us, you only get real human beings on the line. We are a customer service company of human beings like that. I think we’re gonna see some reaction to that, where people will lean into that as a positioning angle that is, we are maybe not anti ai, but we need a term for this.

AI enhanced, not AI first, AI second. Maybe it’s AI second, because Duolingo just was like, oh, we’re AI first now, and there’s all this backlash to that. So maybe there will be this term that is like AI second, humans first enhanced with ai, but it’s a different

Lee Zuvanich: I think the

Sarah: thought process.

Lee Zuvanich: because that would be like saying businesses will be reminding us that they use computers.

Other people don’t use computers, but we always use computers or like

ubiquitous. And to your point, there is ethical and [00:43:00] un ethical callouts for things in other industries and maybe we’ll learn to bake those in. I think it’ll all just become one big seamless monster like we just saw this week. OpenAI is confirming that they’re going to start building solutions for healthcare with all that implies, right? We don’t need to get into that. That’d be probably another 20 minutes of discussion about how we feel about that.

But it is what it is.

Sarah: Next week we talk for an hour about just security. Sorry.

Lee Zuvanich: I have a lot of personal feelings about that as someone who’s been harmed by the medical field, but also, there’s terrifying implications, right? As Zach was saying, industry marches on, this is happening. It is here. It’s only becoming more ubiquitous. So I think the future looks like it being a part of everything everywhere.

Most people didn’t have a language for understanding that the systems that we used already had algorithms built in. Otter is a good example. You guys were referring to that [00:44:00] as an AI platform. It wasn’t always an AI platform.

A lot of what it didn’t need large language model sophistication. It only built that stuff in over the last two years. Everything we were using still had algorithms that were sophisticated that didn’t have large language models. So similarly, it will just be built into everything and you won’t really know. What’s what. It’ll just feel like it’s a part of every system, but it’ll be really obvious in the things that we see in the news where people are having human and animal companions that are really just robots. That is obviously, we’re right on the cusp of that because they already have robots that can do tasks and they already have AI that can hold conversations. So we’re gonna have that kind of augment our society in ways that we grew up watching in dystopian movies. it’ll impact the way that we do combat the way that we’ve been using drones. It’ll be like just more ubiquitous types of things. I can’t think of anything that it won’t touch.

Honestly. I think it’ll be everywhere.[00:45:00] 

Sarah: I think my takeaway on this or my perspective with the ai, secondary ai, first way wanna call it is just being intentional about the use of it. Because I agree, I think it is gonna be everywhere, but the big thing I want people to think about is. Just, intention is honestly what I say for literally everything.

Anything and everything. ’cause it’s true. Don’t just add AI just because everybody else is doing it, or you think it’s cool. It, everything should be done with intention. So if you have a system that’s not working, you don’t add AI to fix the system, you inspect why the system’s not working, and then you create a plan that might include ai.

But that’s the biggest thing that I would like to see more of is just instead of slapping AI into every new startup so that you get more funding, because it sounds sexy, you actually have a reason that it makes sense to use it

Using it in a way that has been well thought out.

Use AI responsibly. That’s my takeaway.

Lee Zuvanich: technology responsibly.

Sarah: Yes, that’s right. It should be, it’s more broad.

Lee Zuvanich: I think that we’ll see [00:46:00] more of like you guys were saying, people one of two roads. AI will be part of computer usage everywhere and everything. And so then it will become, do I subscribe to technology being in everything I do in every moment of my day? Or do I want to go live on a farm in the middle of nowhere, off grid? And I think that there will be a movement people creating those supposed utopias only for them to realize how much help and harm was being caused. Or like retreat spaces where people disconnect for a while and come back and try to reframe and then get addicted to their phones again.

That’s where we’re at right now. It’s just gonna get more intense as it invades every aspect of our lives. I say as a founder building AI systems. 

Zach: Tough.

Sarah: Is the first step. Yeah.

Lee Zuvanich: yeah, 

Sarah: for all of us.

Lee Zuvanich: is what I build today and then my retired like life is running [00:47:00] a retreat space where I just teach yoga again and no one can get through to me ’cause my phone’s always off. So we’ll see. so just to wrap up let’s talk about most valuable insights you feel maybe you’ve uncovered during this discussion or what your takeaways are. you have anything that just like crystallizes where you’ve landed here?

Zach: I would say for me, the main thing is you have to be prepared for this new wave that we’re coming into. It’s not something that necessarily needs to be feared, it just needs to be better understood and see how you can implement it into your life. Some of the jobs and things that are gonna be changed. like I said, it’s not based on expertise. It’s just out of function and ability to perform and make output.

Sarah: Basically what I said just be intentional. Uplevel your skills, understand how it’s gonna affect you, and yeah, I.

Lee Zuvanich: Yeah I would say keep an eye on what your legislators are doing and are they protecting us as we [00:48:00] watch this technology infuse every part of society and just be intentional and aware. Stay aware. All right good luck with that and enjoy this edited by AI episode and have a nice

Sarah: Partially edited.

Lee Zuvanich: Yeah, thanks guys.