Seattle, Washington October 16, 2028
I decided to get new boots today. We haven’t had snow in Seattle since 2021, but winter’s coming and the rains have gotten increasingly torrential these past few years, so it was time.
I feel like I need to share my experience buying these shoes today; partially because I like digging behind the scenes with various businesses I work with, but also because I’m not sure everyone really understands the quiet revolution that’s happened in shoe retailing over the last six years. We are in the midst of a shoemaker renaissance, thanks to a revolution in 3D printing and artificial intelligence that is utterly changing the economics of shoe-buying.
It’s a story worth understanding.
Mr. Lee, Modern-Day Cobbler
Roosevelt Shoes is my local shoe store. It’s part of the community-owned Roosevelt Mall, which already includes some 70% of the commercial storefront in my neighborhood. The store is owned and run by Lee Ji-hoon, a first-generation Korean-American whose father repaired shoes at the University Village Mall here in Seattle (and who actually repaired a pair of boots for me back in early 2010’s).
The best part of buying my boots at Roosevelt Shoes today was Mr. Lee and his absolutely passion for making shoes. It’s not just a job for him; it’s a calling. He wasn’t just trying to sell me some boots. It’s like Mr. Lee believes there’s some ideal pair of boots floating out there in the noosphere and that it was his job to bring it to life though this partnership between me and the artificial intelligence he uses. I’m pretty sure Mr. Lee’s surname is the Chinese symbol for “plum tree,” but it would be so much more poetic if it were the Chinese symbol for “li” – the embedded, transcendent patterns of nature.
At any rate, Mr. Lee is a brilliant guy whose enthusiasm and genuine passion to serve make me effortlessly want to send everyone I know his way. Funny how that works.
I’d known about Roosevelt Shoes because I walk by it all the time on the way to the gym. I didn’t actively pay attention to the shop until it was time to get a new pair of boots though, and Rover, my Virtual Personal Assistant, flagged it as the best match for my needs. Part of the reason Rover zeroed in on this store is that I actually have a financial stake in its success, since it’s part of the Roosevelt Mall, which is owned by my neighborhood, and therefore by me. The other reason though is because Rover knows I value human connection when I shop.
Rover set up an appointment for me and over the course of twenty minutes or so Mr. Lee and I worked together with an artificial intelligence program called “cobbler.ai” to design my new boots. I then headed back over later that afternoon once the print job was done.
I love these boots (see Mr. Lee holding them in the picture above). They’re so comfortable that, if it weren’t for my wife, I’d probably wear them to bed tonight.
The Artificial Intelligence: “cobbler.ai”
One of the more interesting parts of my boots story is the artificial intelligence Mr. Lee used to print my shoes. It’s a distributed package called “cobbler.ai,” which is only licensed to authorized local shoemakers.
Cobbler.ai was developed by serial entrepreneur Tony Hsieh back in 2022, sort of as an antidote to some of what happened after selling his high-flying online shoe retail business, Zappos, to Amazon back in 2009. After the sale, Tony became increasingly interested in the impact that businesses have on local communities. Just a few years after the acquisition, he started pouring much of his energies into revitalizing the old downtown of Las Vegas through something called the “Downtown Project.” Tony is a tech guy, so I suppose it was only a matter of time before he connected the dots linking shoes, cutting edge technology and local communities.
Here’s what I’ve gleaned about cobbler.ai from written reports and from talking to Mr. Lee and a few other shoemakers here in Seattle. Like many artificial intelligence systems these days, cobbler.ai is designed as “intelligence augmentation” – in other words, it’s there to augment the capabilities of the local shoemaker. Its intelligent cobbler agent has a customizable persona, who works with the human shoemaker. The agent handles all the data integration and modeling work, while someone like Mr. Lee is there to pick up on all the nuances of working with real humans. Watching Mr. Lee interact with “Frank” (his name for the agent), it’s clear that Frank works for Mr. Lee, and not the other way around – which is more than I can say for the centralized online commerce sites these days.
The data integration module of cobbler.ai allows a shoemaker like Mr. Lee to pull data, with my permission, from various relevant data sources that will help the two of us design the right shoe for me. If I had been buying running shoes, that might have included data captured from the iMove elliptical and treadmill machines at my gym in Greenlake, but in this case, it’s data pulled from my GroupHealth doctors along with my chiropractor and Tui na practitioner. GroupHealth provided Mr. Lee with detailed models of my skeletal and muscular structure, along with some background information on a knee injury I suffered back in 2011. My chiropractor and Tui na team provided more nuance on how to tweak the thicker sole of a boot design for someone like me who’s gotten used to flatter, barefoot designs.
The design module is one of the more interesting things about cobbler.ai. In this case, the underlying 3D model came from Merrell, which has a long history of making boots. The model gives cobbler.ai a sound platform that easily accommodates my personalization modifications. Function is only part of the design, of course. Aesthetics are also key, and here the underlying Merrell design is easily modified using a variety of techniques. The first relies on my own underlying aesthetic, which is constantly tracked and updated by Rover, my Virtual Personal Assistant. Rover knows pretty much everything I’ve purchased over the last eight years or so, and it pays attention to my attention, noting when my eyes linger over certain types of artistic styles and designs. Cobbler.ai knows how to incorporate that information into its designs, at my request.
The other thing that’s interesting about the cobbler.ai design module is that it doesn’t necessarily lock me into who I am and what I like today. It has very sophisticated methods for drawing design inspiration from a growing network of “fashion forwarders”– people with a much keener design aesthetic than mine who are constantly pushing the edges of creative expression. Cobbler.ai goes out of its way to play nicely with this crowd. As you can see from the picture, I went with a gold treatment on my boots, and that’s definitely beyond my typical look, even though I tweaked it a bit based on my own preference for dark outlines and contrast.
The Economics of Commercial 3D Printing
Cobbler.ai beautifully illustrates the decentralized, economic reality of current additive manufacturing, or 3D printing. The vision of a household, general-purpose 3D printer hasn’t quite panned out the way many of us thought it might twenty years ago. We may well get to a Star Trek “replicator,” capable of seemingly synthesizing matter out of thin air. But we’re not there yet.
Printing commercial-grade products is a challenging feat, especially those made of “multi-materials” (think of the composite of leather, rubber and nylon that comes together in a shoe). The cobbler.ai printer costs around $35,000, according to Mr. Lee, and that doesn’t include the artificial intelligence software, which averages an additional $23,000 a year.
The point is that even though we have simpler household devices like today’s carbohydrate printers for pastas, crackers, cereals and the like, we’re still quite a ways away from reasonably-priced household printers for complex consumer products. As a result, multi-material 3D printing has led to a renaissance of small, local businesses, like Roosevelt Shoes and the rest of the cobbler.ai network.
Regenerative Ownership and Stakeholders
One of the more interesting aspects of the cobbler.ai network is the way the money flows. When Tony Hsieh developed the business model earlier this decade, he used a deliberately “regenerative” framework to ensure the long-term sustainability of its stakeholders; you know, the people most essential to its value-creation processes over time.
Shoemakers like Lee Ji-hoon actually own the franchise network behind cobbler.ai. This makes enormous sense to me because it’s people like Mr. Lee who actually train the artificial intelligence, making it smarter and more capable over time. In fact, after talking with Mr. Lee in a bit more depth, I learned that while the legal structure of cobbler.ai makes use of franchise law, in practice, the owners see it as a kind of a modern-day guild.
Membership in the guild is quite selective. Existing owners want new owners who are smart and hardworking, people they can trust with the business, but also people they can trust when it comes to training the artificial intelligence over time. The AI is the core intellectual property behind the whole venture, and the key to the network’s long-term viability even after multi-material 3D printers make it into the home.
I asked Mr. Lee about this selectivity and he sees it as a natural development of an operator-owned network. There is a rather long waitlist for cobbler.ai franchises here in Seattle, although there are still plenty of franchise opportunities in smaller markets around the country. The network is already in Canada and parts of the European Federation and is preparing for further expansion around the globe.
Looking at the cobbler.ai network directory here in the U.S., I see that its members are fairly diverse in terms of ethnicity, but that women seem to be underrepresented. When I asked Mr. Lee about it, he only shrugged somewhat awkwardly and mumbled something about the fact that they were working on it. I guess even the most progressive of organizations still have their blindspots.
Where there do seem to be a large number of women within the cobbler.ai network, however, are amongst the fashion forwarders. Mr. Lee introduced me to his 18-year-old daughter, Sook, who does an amazing job of translating local fashion trends here in Seattle into design elements for shoes in the cobbler.ai network. She gets a cut of sales when her design elements are used by customers, and while she’s not in the rock star ranks of the professional fashion forwarders yet, it’s enough money that she’s been able to save up to put herself through design school at Cornish here in Seattle.
Fashion forwarders aren’t the only cobbler.ai stakeholders who share in the upside of the network. When my GroupHealth skeletal, musculature and other models are used in my shoe design, money flows back to the health cooperative. The same is true with my local gym, American Athlete, with the biometric data from my workouts there. It’s not a lot of money, mind you, but distributed ledger technology has lowered transaction costs to the point where these kinds of small-scale financial transactions are now feasible. Over time, and with enough volume, they might well add up to something.
Cobbler.ai is also committed to supporting a broader base of stakeholders as well. Our neighborhood development corporation, the Roosevelt Mall, owns a ten percent stake in Roosevelt Shoes, which means that my wife and I own something like .0004% of the company. That’s not a lot, but along with the rest of the other retail stores here, it’s starting to generate a nice annual dividend, especially now as more and more shopping behavior shifts back to local again.
What’s also quite inspiring is how committed the cobbler.ai network is to sustainability and to turning the market for shoes into a circular economy by pushing the cutting edge of biomaterials in 3D printing and by providing incentives and solutions for customers to help them separate the technical and biological nutrients from their old shoes.
Bootstrapping a New Economy
It’s fair to say that the market for shoes is a reflection of much broader forces now affecting many sectors of the economy. The shift back to local, though still in its early days, is well underway and the positive effects on local communities are already quite striking.
Just as important though, one of the things I enjoy most about the cobbler.ai story is the partnership it highlights between humans and machines. This kind of partnership is not a foregone conclusion, of course. The commerce giants like Amazon aren’t resting easy; their centralized 3D printing and drone delivery model is still a real challenge for local 3D printing-based retail stores. But that model is missing the passion of people like Mr. Lee, and for me and the other rapidly-growing base of customers of Roosevelt Shoes, happens to be a difference that makes a difference.