Car Purchase Gone Wrong to Hardware Lifecycles and Building AI Products You Want

What if the product you want doesn't yet exist?

We start with the example of a car buying experience gone wrong...

In this episode, Dr. Joan Palmiter Bajorek discusses her experiences with car purchasing, the challenges of hardware development, and the democratization of AI. 

She reflects on her search for a new car that meets her criteria, highlighting the lack of options designed for women. Joan contrasts the lengthy and complex process of hardware development in car manufacturing and medical hardware as contrasting the rapid pace of AI and software innovation, encouraging listeners to explore building their own AI products.

Resources Mentioned for Your Own AI Products

ChatGPT, Cursor, Hugging Face

Sound Bites

"I have a very old car from 2008."
"I would love to get an electric car."
"Hardware is hard, it's difficult."

Takeaways

🚗 Design: Joan highlights the lack of cars designed for women in the market.

💰 Cost: She mentions the high costs associated with developing new cars.

⚙️ Comparison: Joan contrasts hardware development with the speed of AI and software innovation.

🧠 Encouragement: She encourages listeners to build their own AI products, even without coding skills.

🌍 Access: Joan notes the democratization of AI development, making it accessible to many.

🔄 Iteration: She reflects on the iterative process of software development compared to hardware.

Learn More

YouTube! Watch the episode live @YourAIRoadmap
Connect with Joan on LinkedIn

✨📘 Buy the Bestselling Wiley Book: Your AI Roadmap: Actions to Expand Your Career, Money, and Joy. Featured in Forbes!

Who is Joan?

Ranked the #4⁠⁠ in Voice AI Influencer, ⁠⁠Dr. Joan Palmiter Bajorek⁠⁠ is the CEO of ⁠⁠Clarity AI⁠⁠, Founder of ⁠⁠Women in Voice⁠⁠, & Host of ⁠⁠Your AI Roadmap⁠⁠. With a decade in software & AI, she has worked at Nuance, VERSA Agency, & OneReach.ai in data & analysis, product, & digital transformation. She's an investor & technical advisor to startup & enterprise. A CES & VentureBeat speaker & Harvard Business Review published author, she has a PhD & is based in Seattle.

Disclaimer: Our links may have affiliate codes. This is an educational podcast and not intended as legal, career, or financial advice. Seek professional gu...

Chapters

• 0:00 Introduction to Hardware Life Cycles
• 2:21 The Challenges of Car Buying
• 5:29 The Intersection of Hardware and Software
• 8:02 Democratization of AI Product Development
• 10:44 Understanding Target Markets and Iteration

Transcript

0:00

Hey folks, welcome back to another episode of Your AI Roadmap. I'm excited today to talk to you about hardware life cycles and building products you want and in AI how quickly we're building things and also kind of that juxtaposition of hardware and software and AI. So if you've read my book, Your AI Roadmap, you know that I drive a beat up car. I have a very old car from 2008. And it has really good gas mileage which here in United States and probably all around the world is important and This car is paid off. cars are depreciating assets It continues to lose value over the time I had the car and I've driven the car all over the United States. So it gets good gas mileage. It's reliable but today it is beat up and I don't even know if it's worth $1,000. It is a beat up car. So as I've been looking into the next steps, I've wanted to get a different car, probably a used car, and I have different parameters I'm looking for. I'm wanting to get a car that is probably a plug-in EV electric vehicle, because I would love to get an electric car, but just getting a fully electric car, unfortunately the grid where I live for the power grid is not very stable. So if there were an emergency and I didn't have access to electricity and my car didn't work, that would be a very big problem. So getting a plug-in hybrid is what I'm hoping for and I'm hoping for like not ugly colors. I would love to get a car that is blue or silver or white. Like, I don't know about you, but there are a lot of really... ugly cars, I don't mean to be rude, but there's these blacks and dark grays and the look of the car, they're very aggressive. They look like they're mean. The face of the car, it looks like it's gonna growl at you. I'm not in the car space, okay? So if you're in the car space and you're laughing at me, you're welcome to. But I was hoping, okay, was hoping for plug-in hybrid. I was hoping for just reasonably attractive. and i was hoping for reasonably pretty colors and i would really like it to self park um i'm gonna let you in on a little secret here podcast folks i have many skills spatial awareness is not one of them i have trouble with like spatial directions i'm pretty good at sports but when it comes to like, you know, when you walk from your car to a restaurant and then you walk from the restaurant back to the car, I struggle. It is not like I drop a pin to make sure that I can get back because that is just the way my brain works is just not that great. So the idea that the car could self drive, I would love that. We're not there yet. I have friends who are working on that problem. So self driving cars that are truly safe. We're getting there. We're not there yet. So instead of that, I would love it to self park. Okay, I would love it. have a friend and his car self parks to the millimeter. It's a little scary, but it is extremely good. The lidar, the sensors of the car do that for you fantastically, better than human, frankly, with all the fender benders we have around my neighborhood. ah plug-in EV, just decently attractive, certain colors, self parking. these are four criteria. and decently priced. I'm not gonna spend a wild amount of money. Okay, five criteria. And already this car doesn't exist. I was looking at the criteria of cars and um there are two cars that are even close to the criteria I have just listed off. And I was so sad because I was like, wow, I didn't think I would need to compromise this much. really? This is not that many variables. Your researcher friend has very few variables. I thought that was reasonable. And then I realized that these cars have not been designed for me. If car manufacturers 10 and 20 years ago had been like, ooh, the buyer persona is a woman, Gen-Z millennial, and here are her criteria for a not ugly car with these colors and it's self-parks, there would be so many choices, right? On the market, there'd be many, many choices. But instead, um your relatively has enough money person who is a Gen Z millennial female, me, there are no cars that I'm interested in. The products on the market are not designed for me. That's what I truly believe. They weren't thinking of me a decade, two decades ago. And I was like, okay, how much would it cost for a company to start from zero to building a car? I looked this up. We've been watching Rivian, if you've seen that company, grow and compete in the EV space, roughly $1 to $6 billion, according to Perplexity. Thank you, Perplexity. So basically, this is a wild amount of money, and it would be smarter for regular, like already existing car manufacturers to begin realizing that they may need to design cars around a different persona of buyer. Maybe a woman who has a lot of spending power. Anyway, ah I don't know that I can make them see me as a customer, but I am feeling a little sad about it. I will say I have to like really not get almost any of the variables I want for my next car because my current car is about to die. And you're like, hey, I don't care about your car buying experience, Joan. But here's the thing. This is a hardware project. These are products in this space. I am a buyer and I'm telling you about a modern experience. of customers looking at what exists in the market. So hardware is hard, it's difficult. You have to manufacture different pieces, right? There's supply chain, there's lots of different industrial engineers. The paint, I'm telling you about different colors I want. Do they have that in the supply chain of different colors? Here in United States, there seem to only be like six colors. Everyone has to have a six colors different car, like come on, I believe they're different paints. Anyway, there are different pieces of how cars are manufactured domestically, internationally, and that is kind of a set recipe. That comes to hardware. I'd like to say also I've done medical hardware projects, projects in the operating room space, in the emergency department. we've prototyped products that I was like, this doesn't exist yet. if I were severely injured and was getting a surgery or a family member or a friend, I would want these products to exist in the medical space. And I was like, how quickly can this get to market? And they were like, uh five to seven years if we're fast enough. This is a company with money. This is a company with infrastructure already set up. The FDA, the approval was already in the room deciding whether they thought this product was safe from zero, from ideating the product, right? So hardware takes a lot of time. We see similar stuff in software and AI, but you don't see it as much. When we have a car, we see the parts of the car being put together. Okay, I'm not a car person, so I'll stop there. The wheels, the hood, whatever. um When we're talking about AI, we're talking about models. You've heard me on this podcast talk about different models, how we're comparing different things, um anthropic clods. work and the different models underneath that. OpenAI's ChatGPT, there's many different models inside of it, like these different tools, right? But we don't actually see them. We're just, we see in a toggle down, like different words. And so it's, I think, harder to conceptualize what is going on on the backend, what things we're calling, what data sets we're calling through APIs to the cloud, which is just big server farms, like the data, data centers. and warehouses around the world. I really, honestly, the hardware is nice in that way that you can see the different pieces or maybe you see a digital twin, right? A digital version of that hardware. But in software and in AI builds, it's kind of more amorphous. It's a bit abstract. And so from my car buying experience, I was thinking like, what is the takeaway? Is the takeaway that I need to like call a car manufacturer and be like, excuse me, I would like you to think about me and for my next car, many decades from now, you know, here are the criteria I want, like, please start. Or maybe they're already starting and, you know, that car is going to come out in next few years. I don't think so. I can be hopeful. And what I think about, though, for AI products is the speed of development and that people like you and I can build these today for ourselves. So if there is something you want to exist, different productivity tools, you want something to clean your data in a different way, you can begin building this even if you do not know how to code. I'm serious. So I do know how to code. I'm not like a savvy software developer, but I can certainly write code, edit code, debug code, et cetera. And so the ability to read it is helpful. but you can today go to ChatGPT and ask for some code. You can go to a tool called Cursor and get some code. I really think setting up the like, what is the persona? What is the need? How do we solve for the different thing? What kind of features? What aspects do you need for this product? And you can begin prototyping and playing and conceptualizing with AI product builds today. As long as you have wifi, a computer, and time. I truly believe this. So if you're listening to this, I would really hone down and just test out, spend three to six months on one project idea, go for it, get feedback, look at hugging face, ask chat GPT for tutorial. Hey, I want to build this. What are the next steps? What would that look like? You can do your own personal bootcamp in this thing and you probably don't need one to six. billion dollars to build this thing, right? And so that's a really, really big difference from hardware and software and AI builds is just the amount of money needed to build these things. I'm not saying no money. You if you start making lots of API calls, if you use models that are expensive, et cetera, there is money that needs to go into that. But it is not nearly the scale of things. And it really democratizes who can build. We have people in 70, almost 80 countries listening to this podcast. Hey friends all around the world. I can like start reading out all the different places we see you all downloading episodes from. Thank you. We really appreciate your support. um People from Massachusetts and Iowa and countries, India, Hong Kong, New Zealand, Serbia. It's pretty darn cool. Sometimes I have to look at the location. was like, what? Where is that? I gotta check my geography. But literally anywhere in the world, you can begin doing this. It's really democratized. You don't need a whole warehouse manufacturing thing for the car, right? You can build your own AI product today. And I know, I will say, I'm based here in Seattle in the United States, and I do know that there are some models that are only available to Americans or only in a certain capacity and they're rolling out to different places around the world. So that is fair. And VPNs exist also, um not legal advice, but... you have the ability to build things that you want today. I already have some ideas of products. like, does this exist already or do I need to build this for myself? Because there are some things that don't exist that I want. And I believe that there is actually a market beyond just one person, me. N of one, how big of the market is there? And back to, there are a lot of women who wanna buy things who have spending power. I was talking to someone recently in last few weeks. And he was saying, the first version of the product I built for men. And then I realized that the target market, if I build it for women, is actually really a lot bigger. So version two of the product is gonna be targeted very much at women in their 30s and 40s. So was like, okay, that's cool. I can't say that I was very interested in this product, but maybe version two is gonna be so much better, who is to say? But I really think knowing your target market, knowing how they're gonna use it, getting early feedback, from beta customers, ideally being that persona yourself, you can really, really understand it. We use this phrase in software development called dog fooding, which is kind of ugly, but you try the product yourself. If the dog food tastes terrible, maybe you need to go back to the laboratory. You know, that's what I'm saying. So the ability for us to try things and iterate and literally have an MVP, a minimum viable product in like two weeks, right? If you worked on this. I don't know how much free time you have, but maybe that's more like two months or two years, but the ability for us to get things to market, for people to download it, for people to try it is so much faster in software. And I remember leaving that medical hardware project and I was like, wow, I am so grateful I'm in software. Like, woof, that hardware land is hard. It is difficult in a different kind of way. So this was me talking about cars and hardware and software and AI. And I hope that you are inspired to, A, help me get a better car. I'll have to just navigate that soon before my car totally breaks down. But also, and I do use the bus and I use public transportation. Don't get at me. I really only use the car for grocery shopping and hiking. So I hope this has expanded your mind in this way and I look forward to seeing you on another episode of Your AI Roadmap. If you enjoyed this, please download the episode, like, subscribe and send it to a friend. We really are hoping to get it out to more people. If you haven't looked into it, check out the book, Your AI Roadmap, which you could also find at yourairroadmap.com. All right, I'll see you on another episode. Cheers.