Eli Feldman, Co-Founder & Partner, Shy Bird

(upbeat music) - What is actually happening in restaurant tech? I'm Kristen Hawley, Founder and Editor of ExpoDite. I can't miss newsletter about restaurant technology and the future of hospitality. - And I'm Brandon Barton, the CEO of Bite. We are the leader in chaos solutions for the fast casual and quick service industry. And I've been in restaurant tech for, oh my gosh, 17 years. This is the simmer. - We are moving past the headlines and into the real talk, interviewing the smartest people in the business to discover what's powering restaurants now and most importantly, why you should care. (upbeat music) - Hi, Kristen. - Hi, Brandon. - Welcome back. - Yeah, welcome back. - It's nice that we're using the word welcome. We'll get to that in a second, but we took a little hiatus and it was nice to take a hiatus. I thought we'd get right back to it after Labor Day and it turned out we didn't, and that's okay. It's conference season, it's busy. - It is busy, you've been busier than I have, but we did get to hang out on Monday, we're recording this Friday. So Monday, the 15th in New York at the Welcome Conference. - At the Welcome Conference. The Welcome Conference is turning into one of my favorite days of the year. Like, it is really something to look forward to. There's a whole bunch of people from the world of fine dining that I don't exactly see on a daily basis because I'm not working in that world anymore. And I know that I will see them at Welcome Conference every single year. And it's like, I said that to a bunch of you, "Oh, here's our yearly hug. Like, how are you? Good to see you." - Yeah. I've described it as something that I find very inspirational and like fills my cup in a very unexpected way. - So wait, why don't you give the little brief or on what the Welcome Conference is because maybe something doesn't know. - The Welcome Conference is put on by Will Goudara, formerly of 11 Madison Park and author of unreasonable hospitality, which if you haven't read it at this point, just go get it, it's been years. And it is a day of speakers that has evolved. I think it was more restaurant focused in the beginning, hospitality focus, Welcome Conference. And Will and team have done a really good job of broadening sort of the definition and the scope of welcome and hospitality to include speakers like author Adam Grant this year was great, the, wow, I'm gonna forget his name, the Savannah Banana's guy, you were like racing across cold, you were racing across town to like make it off your flight from your other conference. And I was texting you and I was like, he's on stage, man, you gotta get here. And it is meant to be a day of inspiration and also like actual tips for improving the hospitality of your organization through perhaps unexpected or interesting means. - And by the way, if you're listening to this, you would enjoy the Welcome Conference. I'm fairly, I'm fairly sure, right? Like it is something that you'll pick up one or two things and see a bunch of people that you know, you probably know, but you'll pick up one or two things. I love, I love just all the frameworks that people share on stage because I-- - It's a little Ted talkie. - And exactly in a good way. It's like, oh, I think that too, but holy smokes, I've never thought of it as a system. I just thought of it as the way that I like to teach people how to do stuff or the way that we do things internally or whatever. And it's like, if I just systematized it and wrote it down, it would actually be like a playbook. And so I kind of like call it the unreasonable hospitality team. I feel like that's like the banner now that puts on the Welcome Conference. So I told the team, including Brian Cannellus and stuff. I was so psyched this year on the speaker. - It was really good. - It was really good this year. And that's not to take anything away from previous years, but something was special and inspirational this year. Jesse Cole opening up that, I don't know, I've never, I want to go to a Savannah Banana's game now. - Did you just miss them at Yagi Stadium? - I know, I know, but I'll just, I'm in so many other cities at all times, so it'll be fine. - Well, we did, and we had a great time. And then the Welcome Conference also famously has the dueling after parties from Open Table and Rezzy. It is the manifestation of Reservation Wars 3.0. - It has become a thing. It's actually so funny. And we were like, everyone's like, where is the party's gonna be this year? And this year, Open Tables was again, once again, as in previous years, not last year, but the year before just like the restaurant at Lincoln Center. - Yes. - And where the conference was. So it was like right across. And then Rezzy vented out Postice downtown, which was like, I love, by the time we made it downtown, I was just walking around that restaurant, I was saying, I love this restaurant, I love this restaurant. I do, I go there every time I'm in New York, I take my children and the, I first Postice was very formative for me because it was like really hot when I first lived in New York City as a young 20-something, working and publishing. - Right. - It was a destination. You had to call to get in. It was like the reservation phone line. And so I really, I enjoyed it. And I took a little snooze on the bus from the shuttle bus, the charter bus from the OpenTmo party. - Oh, you did, that's right. You actually took a little, you had a adult break while all of us were yelling and screaming about what. - Injever break, yeah. It was like Will was taking videos. You were chatting some, like a chef. I don't know what was happening. And I was so surprised. - No, somebody's opening a restaurant in the West Village and we were all trying to figure out what space he was, he was choosing. And it was like me and Colin. - Is that why you were screaming Kelly and Dean Kav company at the end? - Yes, yes. We were all picking like random places that were like, 'cause he's opening, I can't say who or anything, but they're opening like a unreason, not unreasonably, but like a uncharacteristically large space in the West Village. And they don't exist. So it was like, just by the process of elimination between Colin, Kamak and myself and like two others, we could figure it out. If I was like, what are big restaurants in the West Village? Anyway, I thought the after parties were great. It's really great to see everybody. I do like love the natural tension of who's going to a party when the Rezzy bus sits outside of Lincoln and picks up people that get their first drink at Lincoln and bring them down to the past East Party. - Yeah. - And the tension is kind of palpable. - Yeah, thanks for holding the bus for us, Rezzy. Sorry about that, it was my fault. - Sorry, they were tired of your fault. - I was last one. I know, but I was talking to a CEO. - Oh, it was fine. But it's, I think next year. Okay, here's just early prediction for 2026. There will be a DoorDash 7 rooms party somewhere. - Oh no, I can't do that on Monday night, man. - It's not Monday night. It's welcome night. It's not Monday night. - Okay. - It is your professional duty to pit all three. - We'll see you next year. - So that's my prediction. They have to get involved. It's fun to be able to do. - We know one of those person makes like the NPR reporters have. We can do some like podcast reclive. And thank you to everyone who said hello to Brandon and I, because we don't usually appear to get together. - That is true. - But you're a tall person and very recognizable. And there were several times, and I'm serious. Several times where people knew who you were because you're more outgoing than I am. And then I would talk and they were like, "Hey, you're Christian. It's true of you. It's live." It was really fun. - Yeah, maybe that's why also it was the best. The best welcome ever. And welcome celebrity. Going from a high to a low, I think it's reasonable for us to mention, pretty big playoff this week from Olo. - I actually don't know the size of it. - Yeah, I don't, all the sizing is rumor. So like I don't really feel like playing in that range, but it sounded like a three digit number of people. Toma Bravo closed last week on the acquisition and this week a good number of people let go. I, several folks in my world had predicted that. And like if you were to chat GPT, what does Toma Bravo do when they buy a company? Like, oh, their playbook is like get in and cut costs and whatever, and like is this a new, like anybody who's ever worked with private equity, this is not a new thing. But really sucky day for a bunch of people. You know, I hope those people are seeing things like this like an Olo friends group on LinkedIn, people posting jobs and stuff like that. So I hope people are gonna land on their feet. But I think it, I think there's like a broader like up leveled kind of thing there that I had not expected where, you know, they're cutting a whole bunch of people. And I think that, you know, in going to some of these actual tech conferences, not the welcome conference, I wonder if people are scared about that for the service that they're going to be getting from Olo in the future. You know, I think that there's a little bit of nervous energy around it. So I'm very curious to see how the next couple of months play out if there's any like large brands at Leave or anything like that amongst the transition. Not that they should. I mean, the service is probably the same, but like, I don't know, that's kind of what happens. So it's something that I'm keeping an eye out on. And I hope those people do land on their feet because that's shitty news to get. - Yeah. - Well, that might be the whole story. We'll continue following it. Well, through our adventures, we sent a note over to somebody who you recently wrote about Eli Feldman. Eli Feldman is the owner and operator of Shybird amongst other things. And Eli's joining us on the pod. Eli said, "When do we jump on just now? "Long time listener," which again, warms our heart. And honestly, Eli, we know each other a little bit, but we really know each other through some other people that we both really respect. And Chris, you've spent a bunch of time with Kristen. And honestly, you're one of the smartest minds in the world of full service restaurants and technology. I'm so excited for this. Honestly, to learn things from you. I mean, you recently hosted an AI summit for restaurants in Boston and Kristen wrote about it. So you guys should talk about that. - Hackathon. - Hackathon, sorry, summit, hackathon, same thing. - Give them things. - Yeah, I hope you've built some cool stuff. And to be honest, I think you're running full service restaurants in such a correct and modern way. So again, excited about it, Eli, welcome. - Yeah, super excited to be here. Thank you guys for having me on. When I look at the list of people who've been on here, I'm not sure I necessarily belong, but it's cool to be here. - Sir, please. - Very cool to be here. - We like to interview the doers, not just the talkers. So-- - Yeah, that's fair. - Yeah, and I'm always inspired by your focus on technology and practical applications and uses of technology in your restaurants. But in the recent interview that we did, and the reason that I wanted to really focus on your efforts in AI is because you think about advancing the entire industry, not just making your restaurants better. You always have an eye, I think, and I think like this too, as to like the bigger picture of what's going on. But I do want to start by talking, can you just like, I wrote about this in detail in expedites, so we don't have to go through the whole day. But like what made you want to get a bunch of restaurant people in a room and talk about AI and then builds, you know, some MVPs, maybe we'll call them to help wish operations. - Yeah, I mean, what made me want to do it, you know, I think at the core of everything that we're doing, whether it's Shybird or newsletter or anything, is really just sort of rooted in the idea that like, you just think restaurants are really important. Like, full stop, like, you know, you employ one in 10 people in the country, one in four to five spaces, like, which is crazy. Like the grounds where retail spaces is a food use. And I think all those numbers are just going to keep going up. So it's just like a wildly, I think, important, but also undervalued and misunderstood industry. But it's also like legit broken. The economic model just doesn't really work. And I just think when I've seen AI, it was like in, when I first saw it, it was the first time that I saw something from a technology that I was like, okay, this actually could make a meaningful difference. And in the way that nothing, I don't, I would argue nothing really before it has. And so then just kind of dove in and then started to see some of the benefits of it. And it was like, we've got to talk about this stuff. We've got to, as an industry, get our hands on it, not wait for the big tech players to kind of deliver us the their version of what all this stuff is. We've got to be able to influence what their version is. For some of us, we should recognize that we don't need their version at all. Maybe not all of us, but some of us. So that's kind of what got me into it. I was frankly surprised by the level of interest. - You spent it up pretty fast. - Yeah, I mean, I've generally found when you're doing something with restaurant people, giving them more than seven to 10 days to think about something as a recipe for them vowing out. - It's true, yeah. - So unless you're doing something like the Welcome Conference, give people a max of two weeks to commit to something. So yeah, we put it together pretty fast. Once we got the great night, I worked with agent AI, which is sort of an offshoot of HubSpot as a partner in that. And they were amazing. I mean, it was just incredible to have some people in the room who could really facilitate. - So why don't you just go back real quick and like, if you were to give a couple of things, like what's broken? I mean, like I think everybody listening understands that things are broken in the industry, but like what is most broken to you? And maybe the places that you think can most be addressed. I mean, there's a pricing issue that we've known this forever. I mean, burgers are the same price that they are. They were 20 years ago and yet the cost of goods has gone up, you know, 300%, right? But like from your perspective, what are some of those things? - Yeah, I mean, to start like just profitability. And I don't mean I'm like, yes, I'm somewhat of a capitalist, but I'm not like, I live in Boston. Like, and I belong here for whatever that means. But we can't sustain an industry that one in 10 people work in on four cents profit. Like, it just doesn't work. And, you know, in a lot of ways we got here through what you're talking about brand and like rising cost of food, inability to sort of charge in response to that, what is kind of necessary. But I also think in addressing some things that we didn't address for a long time. Like this industry oftentimes when it was making 12 to 20 cents on the dollar was doing it based on exploitation. And I think we're doing less of that now, which is good. But now we got to solve the other side of it and somehow still make restaurants accessible places for people to go of all classes. And so like that's, I think at the core of it is a profitability problem. You know, I think, you know, not to get too sort of esoteric about it, I think a lot of that is just predicated on like, it is a 19th century business model trying to survive and thrive in the 21st century. And there's a reason that nobody builds stuff in the middle of cities anymore. Like you don't make things in the middle of cities anymore 'cause it doesn't make sense economically. But we make a lot of food and we gotta figure out how to make it actually work for everybody who is a stakeholder in it. In particular, those who work in it and those who take the risk to open it. I would argue. And maybe on the optimistic side, like something clicked for you when you saw the ability of AI to do things. And clearly obviously put together the Agen AI plus you guys, you know, for the way I love that. You know, I followed Darmesh on LinkedIn. Like it fascinated by everything that's happening there. Like it is so cool. What for you are you optimistic about, and especially as it relates to full service? Because I see a lot of the AI applications in the QSR and the fast casual side. But like full service, it's harder. You still, I still think you need a bartender. You need a server. You need a, you know, you need great cooks that are executing on like, you know, sub, you know, big Michelin level food, you know what I mean? It's so what excites you? A whole bunch of things. I mean, I don't think you necessarily need to think about AI as replacing people as the way to make it exciting for full service restaurant operations. I think that's a bit of a misnomer. I think AI gives us an opportunity to do something that none of the tech players are really aligned and incentivized to do, which is actually create a coordination layer across disparate systems that don't talk to each other. And so like actually having meaningful understanding and 360 understanding of your restaurant, by just sort of pushing everything out of the walled gardens into one place. And then being able to point the smartest machines ever built for $25 a month at that one place. I mean, how could that not be exciting? Right. Just take all the data I have and point cloud at it or point chat GPT at it and say, how can I make this better? And one of the things I think about is a lot of people think of restaurants as sort of backwards technology wise. I don't agree with that. I think technology just hasn't caught up to where restaurants need it to be, which is like real time. That's super sweet. That's super sweet. Real time, actionable, insightful, not about data, not about information, but trustable insight and then action. And without AI, we never got there. So we just, these companies have just been dumping data on us for 25 years. And there's no bandwidth for that. There's no time for it. Oftentimes there's not the right know-how, even if there's the right instinct. I'm sure somebody will be like, you should have said this. So I might as well say, can I make like the steel man case for you on that? Totally. Yeah, I'll first compliment you here. Eli, you're one of the smartest restaurants I've ever met in my life. That is like full, hands down, full stop. And I mean that only knowing you a little bit, okay, right? Do you think, I think the steel man argument back, or not do you think, but I think the steel man argument back is simply your average restaurant owner is not going to react with the same kind of speed and agility and preciseness that you might with said data. I just, I mean, I can think about how many restaurants and this could be a total, we could sidebar conversation. I don't want to go on too much of a tangent on it. But to me, unequivocally, putting your restaurant online to sell food online was a no-brainer pre-pandemic. And how many restaurants had not had that set up? And by the way, third party versus third party, clearly there's a battle channel. One versus the other. The other gets you more volume, the other cost more, whatever. But how many restaurants didn't think that they should sell their food online and now do? And it took a freaking global pandemic for people to realize that that was a good idea. I mean, why? So maybe the argument back would be like, hey, there are so many things, and it's just so hard to get restaurant tours to adopt them in the way that is needed for a number of reasons. Yeah, I don't fully disagree. I think, I'll be honest. I have a lot of empathy for restaurant tech companies. We're a terrible customer. We are. We have no money. We have no time. We want everything customized to us, even though 80% of what we're doing is largely the same. We only sort of think about that last 20%. So I do have empathy for all of these companies. It's what they're doing is brutally difficult. You have to internalize the reality of the operator if you're one of these companies. And I think there's a reason that people didn't think about putting their restaurants online. And one part of it was, frankly, some of the people who did it went about it in such a way that was incredibly off-putting. If you weren't a restaurant that was putting food online and then somebody shows up and waves a phone at you and says, "I'm here to pick up a takeout order." And you're like, "I don't do takeout." And they say, "Well, I have this order." Like, that's not exactly a way to ingratiate yourself to an industry. - Sure, sure. - So I think there were certain things about how it was done that were tricky. I think there are other things that, like, we are slow. We are slow, generally speaking. I think the economic pressures are making us faster. And I think also just, I think we sometimes underestimate how influential Moore's Law actually is. Like, it changes everything. Like, the adoption curve of everything is radically faster. - Can you just tell our listeners what Moore's Law is in a sentence? - Yeah, then we don't need to get into like the basics of it, but it's basically like computers get twice as smart every 18 months. - Right, right. - Don't come at me if you're an actual technology person. - No. That's what's though a sentence. - It's your size. - They're all from here. It's fine. - They get wildly faster. And so like, if you think about like the speed of adoption of, you know, server based point of sale versus cloud based point of sale, the pace at which toast got from zero to 150,000, like, a lot of that does relate back to Moore's Law and speed of networks and speed of computing. And that changes everything. And I think it's also changing the pace of which restaurants are will consider adopting things. - I think that's fair. And I think the other thing Brandon about AI specifically, and I don't want to be like, this time it's different. But, you know, the way that it can be conversational and friendly and it's just the language is different now with this particular kind of technology. Like, the way that it prompts, the way that you do that. It's just, I think that this is maybe a little like, it just, it feels like the right kind of technology for a group of hospitality people that are perhaps averse to tech. - Yeah. And like the positive feedback loop you can get from it, which is like the flywheel of I invested time and I got an outcome. I think really speaks to restaurant people when they actually get that first aha moment of like, so one of the, one of the chefs who came to the hackathon reached out to me and said, hey, I built my first custom GPT. And like he had, you know, multiple restaurants that are all sort of Eastern Mediterranean Greek really highly regarded restaurants in the Boston area. And, you know, he was able to, he created a custom GPT that allowed his CDC's and your chef de cuisines at the restaurants to interact with the GPT to think about fall menu development in the way that he would to help accelerate their confidence in talking to him about dish ideas that they had. The fact that someone can go from, I didn't know what a custom GPT was to, I built one to my chefs are using it to it's influencing how I'm thinking about my fall menu change. Like, Jesus. - Yeah. - It is different. - That is different. - Yeah. And to my earlier point, I would love to talk to that person about whether they like, I don't know, use like some of them modern like recipe stuff and whatever in their kitchen because I would be, I would be surprised if they did. I would think most of the time it's a no. And like there are these people who are now, as you both are saying, like, oh wait, AI is different. This actually works really like I, like it should and can have an actual impact on me and impact on my business. Well, and that's like that to me is the key phrase. It works like it should. I guess what I mean when I say that technology is caught up to restaurants is like, there's just no margin for error on either money or time. And now that I'm saying that, it's a little weird 'cause these things hallucinating get shit wrong all the time. But the fact that it can do it what it does, just is really conducive to this industry. Yeah. Yeah, that's funny. Last week, the Microsoft co-pilot put mine into Excel or something, right? And then in Microsoft, like put out a statement that was like, don't use this for like, as the final word on anything. Yeah. It could be. Oh, it had lots on everything. Like it's so fun when you want to like AI product. I have raged about this because it's funny that like the errors that are allowed by the computer versus the errors that are allowed by the humans. But it's fine. Oh, yeah, that sweet green thing you post. That's a sweet green thing, man. Yeah, yeah. Yeah, I asked Google how many sweet, I asked Google, just Google how many sweet greens they were in America. And it returned an extremely wrong answer in the AI summary that is now at the top of Google search results. The reason that it did it was that it was pulling the data from all of the sweet green outpost locations. But I don't know where it was getting. I don't, maybe it was using maps. It didn't say where it was getting its data. No, no, it didn't. It was like a website that was also scraping that was then selling that data back to people. And it was just so wrong. And then when I asked Chachi PT, how many sweet greens they were in America, it gave me the same wrong answer. But then it said, wait, there's a discrepancy. This other source, which was sweet green, I think, says it has this many. Oh, this is why there's a discrepancy. And it figured out the answer. And as Google is pulling traffic away from publishers and websites and content creators and all of the things that I do, it's giving also wrong answers. And I was very upset about it. Well, I wish I could screenshot this and show this to our listeners. But it's still to this day, it says, as of September 15, 2025, Scrape Hero reports there are 904 sweet greens with 180 in New York. Meanwhile, the first link on Google says, today, this is a sweet green's website, treegreen.com. Today with over 140 locations, I mean-- somebody needs to teach Google's AI about primary sources. God. Yeah. Anyway. Anyway. Still not fixed. I know, right? I know. Well, yeah. I love the whole story about the chef. What are-- I mean, if you don't mind sharing, what are some of the ways that you're using AI to better your operations? And what were some of the things that maybe you should suggest to other full service-- small multi-unit operator, single-unit multi-unit operators? Is it OK to swear a little on here or no? I think so. Yeah. OK. All right. If you have kids in the car, you didn't know. Are we supposed to say something like that? So why? I would really like, when I tell our teams and when I tell people at the Hackathon and I wrote about it, it's like, I just have these two acronyms. One is bite, not playing to the choir here, but bring it to everything. Oh, I like that. Oh. Wow. So-- No, no. Brandon's going to steal that. He's going to commercialize the shit out of that. And the other one is fails. Fucking around is learning shit. And so we do literally bring it to almost everything. Whether it's the most popular thing I ever wrote was about trying to fix long ticket times at brunch. And the role of AI in that, I use it for like, we had a Yankees game a couple of weeks ago-- or, sorry, no, it was a Jonas Brothers concert. And was using it to say, hey, look back over the last three months, show me everything every day that's over X amount of dollars and sales. Then look at my product mix. And what do I need to get my bartenders and my cooks ahead of? On our busiest days, what should we expect? And it was making recommendations around like, hey, if you're thinking about batching cocktails, you might want to think about batching these. If you, like, given your product mix, you might want to think about how to get tortilla wraps steamed ahead of time. Like, it's just there's almost no limit. And I think one of the things that we're on sort of a mission to try and do is plug everything we can in or send ourselves up to. So like, you know, and without compromising, in any way, our relationship with guests, our relationships with our team. So like, that's, I think, a really important part is like, how do you get the objective reality of the restaurant and the subjective narrative of the restaurant in a place where you can use these tools, but not at the expense of privacy and surveillance and creepy shit? Yeah. When you say you're saying that it as the AI, is it using chat GPT of mix of stuff? I use a bunch of different things. OK. Primarily chat GPT and cloud. OK. So non restaurant specific tools, which is also a very important point here, I think, because these are just general tools. Yeah, I mean, in fact, and a lot of them are, I think, some of the most useful things that I've come to use are actually not even remotely adjacent to restaurants. There's an amazing product called chat PRD. That's a product manager, a brilliant woman named Clare Vo who created this product that is meant to write product requirement documents for technology. That's its intent. But it's really good at-- if you say you can put into it, hey, I want to design a menu offering or I want to design a marketing campaign for the ex Brooklyner who now lives in Boston and has friends from Brooklyn coming up. And it wants to impress them without spending a ton of money. And it will walk you through it the way that a product manager would think about product development. And it's actually wildly interesting. To think about these frameworks that exist in other industries, and I think that gets back to the welcome conference thing, I didn't get to go this year, but my business partner did, and our wine director did. And Adam Grant and the Rory Sutherland ones were the two back just talking about ad nauseam. And Rory Sutherland was so good as well. It was so just interesting. It made me not speed on my trip to Cornell this week. No, yeah, because you won't get there faster. Because it's thinking about the difference between going 70 miles an hour and 80 miles an hour is one minute of an hour. And I was driving for four hours, and I'd get there four minutes earlier, yet I know the implications of if you hit something at 80 miles an hour versus 65 or 70, it's a difference between getting hurt and dying. Yeah, four minutes, not worth it. Anyway, he was great. I know Claire. I'm going to tell her after this that you use it for this really awesome purpose. My husband used to work with her. Oh, OK. Yeah, I agree. She's brilliant. But also, I think that is the thing. It is this-- that is what I was trying to say about AI. It's like this major unlock for restaurants, not because it's restaurant-specific. It's just-- it's hospitable. It's hospitable technology that helps in very specific ways. And you just-- and you already have everything you need. It just helps you think faster and better. And I just-- I love that a general GPT told you to steam for a T.S. It just-- that's red. I'll be curious to see how the bigger players incorporate this over time. And it's really concerning to me how they will, frankly, like-- Bigger players in restaurant technology. Well, restaurant tech, in particular. Featureizing this to increase your average revenue per user is not a recipe for it to actually have the impact that it should have. But that's inevitably where it's going to go. Not to be overly cynical. But that's absolutely where it's at it is. Here, give me $29.99 a month. And then you'll get this canned prompt product. That, again, many restaurant tours, it'll probably be the best version of AI that they can use without putting in a ton of time and/or know-how into trying to get it. This is the problem with, like, scale is that, you know, I don't know how many restaurants are on DoorDash or on, I don't know, toast, right? You said about 150,000. But like, it's just going to be a button that they can click and be like, use AI to optimize blank. And it'll just be sitting right there and easy to press in 20 bucks a month, rather than doing the work that you're talking about. Maybe. Yeah. It's going to be that. But I think if anyone-- I know I imagine there's a lot of listeners from the restaurant tech space. And if anyone's listening to this and is willing to hear someone out, like, just-- you're going to do that. You're going to try to increase your RPU and your LTV through AI. Just don't forget that it's also a thought partner. It's just not about writing crappy emails. Like, use it. Try to build into the product, like, something that elevates the user and, like, the restaurant manager and teases out from them their abilities. That's what we're using. I think we're trying to train our managers to do is, like, it knows everything except what you know is one of the things we talk about a lot in the restaurants. It knows everything about data. It knows everything about whether it knows everything about everything. But it doesn't know anything about who's sitting at table 36, which servers are confident with the wine list and which servers aren't. And how you tease-- how you get them to be more confident about it. So how do you bring these two things together? And the people building the technologies need to think about that, like, that these should be amplifying human capability, not just replacing it. Yeah. I have a question. Do you have any concerns about privacy, proprietary data, like, any of this, like, feeding this-- like, figures and stuff into, like, a chat GPT tool that is not necessarily private? No. OK. Not at all. I mean, to Brandon's point, and you guys summer wrap up here, like, I have concerns about DoorDash and my data. I don't have concerns. I frankly have much more concerns about what restaurant tech players are doing with my data and what OpenAI is doing with it, frankly. Yeah. Yeah. At least they're not withholding it from me. Right. OK, there we go. The AI does listen. It's a repeat back what we've talked about for the last five months, they'll just send it to you. So one of the sides of AI that I think has been-- well, we haven't explored it yet in this conversation. And I think it's actually like the right hook that's coming that people aren't-- I don't think people are paying enough attention to-- is discovery of restaurants. Because as consumers get more comfortable using AI, and AI is put into the tools that they use to do everything from discovery, there is this black box for restaurants on how to be discovered within the systems. And so while there is this beautiful productivity enhancements that you can get out of using AI to enable your humans to give better hospitality, there's this whole other side of it where I'm really worried about how discovery is going to be affected. And now it's pretty straightforward. There's Google AdWords, and you can pay for ads, and you can have high scores on Yelp and this. And obviously, you know 1,000 more hacks than I do as to how to be recognized or Instagram likes or whatever. But it's going to go into a black box for a period of time where we don't know how the restaurant is going to choose the itinerary on where to bring your friends when they come to Boston or whatever. Does that concern you? Or do you have thesis and comments from your perspective there? Yeah, I mean, I have a lot of thoughts on that. I think it is really concerning. And what's mostly concerning about it is that those with the resources are going to-- I think they're calling it like-- it was search engine optimization, and now they're calling it geo-generative engine optimization. And those with the resources are going to figure it out fast, just like SEO. And the independent guy is never going to get onto the field. And so that's really, really concerning to me, but there's also a lot of history of that's just like-- This is what technology does. It favors those with and creates a wider gap between those with and those without. So that's really, really concerning. I think one of the things that I'm trying to do and we're trying to do is increase your surface area as a restaurant, is one of the ways I think about it, is like whether it's creating content or just put yourself out there for algorithms to see you and catch you as many ways as you can. That's not necessarily the business we're all in, but I think it's increasingly a necessity. I will say that the thing that's scariest to me about it is that we never taught computers to understand restaurants. That's the scariest part about it. Like if you think about what all of these models are going to be trained on, it's going to be fast food, fast casual dining, fine dining, ultra fine dining, $1 to $5 signs, and ethnic cuisine type. Because that's the history of the internet when it comes to understanding restaurants. But that's not actually how anyone who loves restaurants or anyone who works in them or opens them thinks about them. And so I think we need a whole new classification system for thinking about restaurants. The metaphor I make here is that I feel like we're where Blockbuster was, and we need our Netflix moment. So Blockbuster was horror, comedy, drama, action. And what Netflix figured out was like, no, you don't like horror. You like Japanese anime body cult horror. And then they were able to use that classification and that really interesting kind of micro genre to make K-pop demon hunters. To make K-pop demon hunters would let's go. Yeah, like, oh my gosh, my six-year-old is having a K-pop demon pop, Hunter's birthday party. Just sorry, excited. Yeah, of course, this is what it sounds like, of course. So like, if you think about, like, I teach it to you often on one of my favorite classes to teach is around restaurant classification and segmentation. And just to talk about how terrible it is. Like, what about pizza? Like, how many pizza places do you have? I have like eight. Yeah, like they're different categories. But they're crazy. They're Japanese anime body cult level categorizations. Like, yes. And like, when I asked my students, how many of you have drunk high late-night pizza? They all raise their hand. Okay, how many of you have drunk high late-night pizza out versus drunk late, drunk high late-night pizza home? Delivery, right. Versus first date pizza, versus parents are in town pizza. That's how we actually think about restaurants is. Yeah, this is a commission. Yeah, this really cool nuanced, wild thing. And nothing is set up to get models to understand them that way. And I went down a really deep rabbit hole about a year ago around this and kind of emerged with my own version of an algorithm around it, which was pretty fun, that turns out the same guy, his name's Tony Jabara, created the algorithm for Netflix and for Spotify, the Discovery algorithm. He did Netflix first and was like, oh, movies are too easy. I want to do something hard. I want to do music. And so just another sort of AI piece to the story here, I went into Claude and said, let's act as Tony Jabara and create an entirely new system for classifying restaurants. And something really cool kind of came out the back end of it. I haven't figured out what to do with it. But it's really fascinating. And like basically what it's based on is that it trains a computer to think about restaurant nuance differences the way the computers think about color. And like, because if you think about the finite gradations of color when you look at this code or a hex code, like that's millions of nuanced differences. And that's what we need restaurants if we're going to actually trust them for Discovery and recommendation. Right now they suck. Yeah. I write, talk and think a lot about the flattening, the algorithmic flattening of culture restaurants, whatever, you know, as written about by Kyle Jacob, I reference this book once a week. And this feels like a piece of technology that can fight that appropriately. Yeah, it just, and you know, if you've ever been in the product world, like we, we think a lot about in our restaurants this concept of the jobs to be done. And like, that's a clay Christian thing. And it's like you didn't, nobody downloaded the Uber app because they wanted the Uber app. No one wants the Uber app. They want the job that the Uber app does. It's point to point on demand transportation. And I think the same is true of restaurants. Like sometimes you're hiring a natural wine bar to identify yourself as someone who hires a natural wine bar. Sometimes you're hiring a natural wine bar with food. Sometimes you're hiring a natural wine bar for just snacks. Like there's 11 different reasons. You might hire a natural wine bar. And there's a different one for all of them. But we've got to figure out how to get computers to actually understand that. Or maybe we don't. And that will, in doing so, we'll just preserve humans. Humans, but, and what he's gonna do it? I'd rather see someone do it with an altruistic bent than like an extractive bent. Always my hand out to it. Sounds like fun. Eli, I was gonna say, the fact that you have and started a restaurant tech company yet is, I mean, it sounds like that's something that should happen at some point. I did, I failed miserably. And whatever, who cares, failed once. Get back in the horse, let's go. It's part of the person. No, in candidness, I am like talking to a few different people both at the startup level and the more established to see if there's a way to work together. Just 'cause, nice. There's only so much I can do. And within the walls of Shybird, very little. Frankly. (laughs) Well, hey, I mean, again, like I said up at the upfront, there are very few people who think about restaurant tech as much as you do and are as smart about it. And still an operator and still running restaurants. Today, like this was chock full of just like a ton of, even more of that welcome conference stuff, that framework stuff about how to think. And I'm super appreciative of you to take the time and come on. We have to check back in with you every once in a while and see how things evolve. 'Cause that's the best part of this is it'll be like a very quickly evolving conversation. No, and thank you guys for creating a platform for different conversations about restaurants. I think that's what we need more of, more than anything else is like places to have conversations that aren't just about the new hotspot or what's happening with bear season four. Like, we've got to get, well, I've never watched the bear. So, you won't get that here. You won't get that here. Conversation. Nope, nope. Not on this podcast. No, I'm sure it's a great show. Well, I'm sad that it's, we're at time because it's always a delight to talk to you, Eli. Thank you. Thank you guys. Really appreciate it. Thanks for listening. For info on these trends and more, sign up to receive my newsletter at expedite.news. And if you love the podcast, rate and review wherever you're listening. We'd love to hear from you. Thanks so much. [BLANK_AUDIO]