March 20, 2026 – Furniture.com Co-founder Alex Seaman and AI Transactions Attorney Spencer Rubin

0:04 Intro 1 : Broadcasting from am and FM stations around the country. Welcome to the Small Business Administration award winning school for startups radio where we talk all things small business and entrepreneurship. Now here is your host, the guy that believes anyone can be a successful entrepreneur, because entrepreneurship is not about creativity, risk or passion. Jim Beach,

0:26 Jim Beach : Hello everyone. Welcome to another exciting edition of School for startups radio, we got a fantastic show for you today. First up, the founder of furniture.com is with us. Alex Seaman, we actually grew up about a mile from each other and had a lot in common, and Spencer Rubin is with us, an AI legal expert, and we’re going to learn a lot about all of the changing rules related to AI and what we need to be aware of. So it’s a great show. Thank you so much for being with us. We’re going to go ahead and get started right now. Very excited to welcome to the show. A local Atlanta grew up here in Atlanta, and now lives in New York. Please welcome Alexandra Seaman to the show. She is the Senior Vice President and co founder of furniture.com she comes from a long generation, four generations of furniture sales companies and some companies that you may know of. Her family also started, can I tell the other name too? Alex, sure. Room, please. No. So her family started rooms to go as well, and furniture.com she holds a Master’s in contemporary art from Sotheby’s and an MBA from Wharton, and was recently named a 40 under 40. You don’t get any cooler than that, Alex, welcome to the show. How are you?

1:43 Alex Seaman: Thank you so much for having me, Jim. I’m great. How are you?

1:46 Jim Beach : I am well. Before we recorded, Alex and I were talking about the high schools that we both went to just down the street from each other. And I think I know what house she grew up in, and I’m sure our paths have crossed many times, and I think that I know her father as well, but we’ll have to research that. But anyway, Alex, welcome. Tell us about, tell us the story. Jump in, wherever you want to go. Great, yeah, yeah.

2:13 Alex Seaman : I’m happy to start. Let me, let me take you like 100 ish, years back, probably more Honestly, my great grandfather, actually. So Jim mentioned, you mentioned that I’m fourth generation, kind of in the traditional furniture retail industry. And my great grandfather started a very small Discount Furniture Store in Brooklyn, New York, and he passed away quite young. And my grandfather, who was at the time, in the Air Force, and his brother came in to run it. They were like 2020, and 17, or something like that. And so my grandfather, along with his brother, and then later my father and his cousin did then grow semen furniture to be, I think, the biggest furniture retailer in New York, but regionally in New York, and when they sold it in the late 80s, early 90s, you know, my dad, with with, like, a lot of support from my grandfather, decided to start a kind of similar but different concept called rooms to go, which, if anyone’s familiar, it’s really about packaging up room spaces so the shopper is able to sort of sidestep the interior design process and again, still like value furniture, very mass market, big box, lots of stores, traditional retail, but that’s kind of the the not, I mean, I would say the industry I grew up around, because that’s sort of just been, I call it like In my DNA. Like furniture retail, it’s in my DNA. Retail is in my DNA. But I myself, took a bit of a different path. So, you know, post college, I got a master’s in contemporary art. I live in New York, and I went on to work at Sotheby’s auction house. So I was in the art world, but then I did work at a fashion startup, at which point in my mid 20s. I, you know, I think a lot of people in their mid 20s are like, Oh, no, what do I do? Do I go into the family business? Do I do this? Do that? Sort of, kind of, let’s say, horizontally, to this at the same time, the ownership of rooms to go. My really sort of adventure arm of rooms to go. Had acquired furniture.com back maybe eight, nine years ago, and just sort of was sitting on it. Knew it was like a very powerful domain, valuable, but wasn’t really putting much of a business against it. And in 2022, we decided that, you know, furniture.com is it? So obviously its own company. It’s a separate company. It’s a totally different type of business model. It’s a.com that had been around and gone through, you know, several hands since, really probably the late 90s, early two thousands.com boom and bust. And we decided to really reimagine it in this very new way that we feel sits kind of at the center of, you know. The furniture shopping experience between, you know, retailers and shoppers. And, of course, you know, I can give you maybe a little bit of context on what furniture.com is now and and how it evolved. Yeah. So it began as an aggregator, but, you know, very quickly, it takes a couple of years, I would say, to build up the foundations of these, like tech, data, type of companies, but at the core, and we did just relaunch, actually, about a month ago, which is really exciting. At the core, furniture.com is kind of beyond a marketplace. I would say it’s more of a decision kind of layer, or a platform that fits at the decision layer of shopping for furniture. And so we partner with at this point. I think we’re at, I think we just hit 80 this week, 80 furniture retailers. And these are, these are national, reputable, trusted furniture brands, right? These are not necessarily like white label overseas brands like you might see at other furniture marketplaces. And I think that’s really, really one of our huge differentiators. And I I think, like, you know, coming from the furniture background, we know we understand the furniture shopper and that that person, you know, furniture quality and brand trust and the customer service and the delivery and logistics and white glove like, there’s so much that goes into it, but it’s such a fragmented industry and business, and with sort of the first e commerce landscape, and then now, where that paradigm is now shifting, with AI and agentic shopping and how people are shopping, we really felt like it was a good opportunity to work with these furniture real retailers, and take the industry and elevated and sort of be a partner to it in this next age of digital shopping, which is not coming. It’s now here. And so we’ve been thinking that way the whole time. And so furniture.com I would say, like in a nutshell, it’s, you know, the one place, the one tab, as opposed to the 20 tabs where you can seamlessly, quickly, discover, decide, and help narrow, and then actually buy a multi merchant cart, all on furniture.com and we do that through different layers of AI, you know, agentic shopping

7:15 Unknown Speaker : And AI.

7:17 Jim Beach : So I’m on the site right now, and there are products, and underneath the product, there’s the price, and also it says lumens, light and living, or one Kings Road, that’s the supplier that you have the deal with, right? 80 of those,

7:34 Alex Seaman : Exactly 80 of those. And we have, like, pretty, sort of very close partnerships with those retailers like we, you know, we get, have their permission, and a lot of times very much joint effort in, you know, getting their product feed. I mean, I think now we’ve evolved where we have the capabilities to sort of do all that work without, without their intervention, which is great, because it’s easy for them. But yes, we have 80 of those product feeds. Many of these retailers are not digitally native. So those that you mentioned, those two actually are. But you know, if you think about the Raymore and flanagans, the rooms to gos, the, you know, those types of retailers, they’re regional, they’re, you know, very successful businesses, multi billion dollar businesses, or, you know, a lot of the retailers, I would say at least a third of our retailers are top 100 in the in the country. And so there’s so much opportunity there. And I think there’s a lot of opportunity for these folks to capture this kind of aging, sort of millennial online audience that’s now becoming a large chunk of the main furniture shopper today in how they’re shopping, because they’re shopping differently, you know, they want to shop between brands they, you know, have, I think the app, we did a bunch of research in house, and then the average furniture shopper has something like 14 tabs open, spends 13 hours researching, you know? So, like, it’s like, I don’t know, like a third of your work week, right? Like 13 hours researching has like, four or five tools open. So whether it’s Google Sheets, whether it’s like Pinterest boards, whether it’s social media or, like, shared notes with their partner, and so how do we take all of that and create a really integrated, kind of centralized page with products, and then a product detail page and a really, really super advanced conversational search, and now we’ve just recently launched Dottie, who’s our AI agent. So how do we bring all of that, all of those sort of tool sets understanding furniture shopper behavior, understanding what they need in order to solve this, like heavy, heavy decision fatigue, which is where we usually find our shopper, that’s really the problem we’re solving. Is this, this decision fatigue, because there’s definitely enough options out there, but how you make that choice and how somebody helps you through that curation? Process that was a big opportunity.

10:03 Jim Beach : All right, I’m so sad right now I’m on the website just looking around. I am so out of style that it is mind boggling. I doubt that 1752 guy over here, you

10:17 Alex Seaman : Know, there’s something for everyone on furniture.com. I said there’s something for everyone on furniture.com.

10:24 Jim Beach : No, not me. I just cannot stand the upholstered headboard, which is 100% in now, you know, and my style is just very different. I like the head.

10:40 Alex Seaman : Let me ask you a question, what can you describe your style or even like without knowing what it is? Can you describe the type of furniture you like?

10:48 Jim Beach : Queen Anne antiques? So 1750 1800 English.

10:54 Alex Seaman : Try certain so open you see the chat search chat box at the bottom. Click toggle over to Dottie, which is our agent like, Agent mode.

11:05 Jim Beach : And, yeah, we’re that. I don’t see that.

11:11 Alex Seaman : Oh, you closed it. Oh, so if you look at Oh, I think it after this, I can tell you, I don’t want to take too much time, but yeah, or you can, if you see in the left tab there’s, like, a search icon. If you click it, it’ll it’ll pop back up. But if you do have time to do it, and you search exactly what you just said in that sort of conversational language, the way you said it to me, you really should start to get some results that work. You might have to prompt back and forth with it a few times the way you’re used to with other llms. But this is our, you know, in house, LLM, but it’s trained solely on furniture, data, furniture, product data, furniture, you know, merchandising expertise, interior design style expertise. So that’s Dottie for you. She’s like your furniture merchant, your in store associate, your stylish best friend, your interior designer. All all in the palm of your hand, on your phone?

12:03 Jim Beach : Yeah, I did. It’s working very well. I love the the Dotty, it’s it’s just not, it’s still too modern for my taste. But anyway, that’s again, as I said, I’m the old one. You know who music died for me, Alex, in 1977 stopped all music stopped at 97 so I’m sort of there with my furniture as well. I just, that’s just me and my personality, I guess.

12:35 Alex Seaman : So fair, totally fair. Yes, I need to see after we get off, if I can find some of that antique, English style woods and through Dottie, and I’ll send you the link. I feel like I probably can. Let’s see. It’s a good challenge for it’s a new, it’s a totally new feature on the site. So let’s see how well she works. I’m confident, yes, yeah.

12:58 Jim Beach : Well, a Queen Anne by definition, has no fabric, though, except for the seat, and so just by definition, I see one here that’s close, but again, it’s still a very modernized version, though. Let me tell you, I’m

13:16 Alex Seaman : Gonna have to talk to our I’ll talk to our head of sales, see what we can do about getting more of that style on. For you, oh,

13:24 Jim Beach : Don’t do it. For me, I’ve got better furniture now than I can handle, though. Share fair. I inherited all of my parents stuff that, you know, they bought antiques only, and so we grew up in a house that, you know, you weren’t allowed to touch the furniture Right, right. And lots of ivory too. My parents collected ivory and put tons of money into ivory, and now it’s illegal and I can’t get rid of it. You know, literally, I’m not allowed to give it to someone, yeah, so, yeah, I don’t know what the hell to do with it. Now no one wants you know. Again, my style is very different. Let me go through some of the questions. I know the guy that owns wine.com and I know what he paid for it in 2001 which was $6 million so I would assume buying furniture.com would roughly be 345, $10 million how am I doing?

14:20 Alex Seaman : You’re in the ball park, but you’re not quite there. It’s not, it’s not something we well, it’s not something we publicly disclosed, but it was a pretty penny for sure, right?

14:32 Jim Beach : Yes, well, it deserves to be. It’s one of the best URLs on earth. You know? It’s just an incredible URL.

14:39 Alex Seaman : So I think for us, the opportunity, you know, as we think about the furniture retailers side, like these, again, it’s a super fragmented industry, but it’s a very successful and very, very cooperative industry. And even so, though it’s and I think most furniture retailers would agree. It’s highly regionalized, and it’s for how big an industry it is, it’s a little bit technologically archaic. And so I think for us, kind of creating a platform where these sort of trusted retailers, brands with quality product, quality service, and everything that comes along with it could compete with some of the, you know, giant marketplaces and really do so in a way that’s kind of, I guess, like optimized for the shopper that needs, really needs help being confident in what they’re buying. And so they need to be able to have choice, and they need to be able to have, you know, AI curation, like, how do we put those retailers in front of those shoppers? Because that, to me, like, our most pinnacle, like, most important thing is have the best shopper experience. That’s always what we’re working towards. That’s always like in any any question, that’s the tiebreaker. Is it the best shopper experience? And in my opinion, those retailers in that product, but being able to shop between them and do it easily, and do it in a way that feels native to the way people shop these days, with chat and, you know, all the social media and all that like, that’s what we were trying to do, and that’s that’s why it’s so valuable to have that URL, because we felt we needed that in order to sort of jump off towards towards our goals, right?

16:29 Jim Beach : Well, the site is beautiful, incredibly functional, very easy to use. You’ve got the right left tools that you know the range and prices and things like that, you can adjust how much you’re willing to spend and turn down the bar so that you don’t see the $5,000 items that you can’t afford. So, I mean, the site is great, awesome. Y’all did well with this.

16:53 Unknown Speaker : Thank you. We’re excited. So

16:55 Jim Beach : How important is social media and SEO? You know, how do you market and build awareness online for the new brand that you are?

17:09 Alex Seaman : Yeah, it’s a great question. I think we have two jobs to do right now as marketers, as a company, which is, you know, first and foremost, most is like top of funnel awareness, right? So these are, we’re starting that rollout heavily on social media, right? Paid social, obviously, organic, and then influencer, I think, is a big part. Like, we’re really focusing on the channels that really, I think, cultivate Trust, which is a key thing with a big, giant ticket item, like furniture that’s tactile, that’s expensive is, you know, that’s visual is you really want to feel confident. You don’t want to have to return, trust me, you don’t want to have to return like a sofa. It’s very it’s not the most fun experience. And so, you know, I think some of those social media and earn channels are really important. I think we’re also heavily focusing on our own brand, which I think is crucial. You asked me about SEO as it relates to, you know, first of all, building a trusted brand, but also and launching one, but also, of course, as we think about the AEO, kind of landscape, which builds on SEO, but really, really heavily relies on brands that have a clear message, that have good reviews, that have a lot of user trust, that have, like, really strong engagement. You know, it’s a little different now. And so if you want your and we, of course do like we want our products, which are really our retailers products, and our company to our brand, to show up in these llms and these new channels. And so we want to make sure we’re positioned for that. And so I think brand itself is a key channel, and that’s something we’re super focused on. And then, of course, the lower funder channels are helping us. I guess I’d like to say, like, get water through the pipes. So we are currently rolling out, but over the last five weeks, we’ve been rolling out the the multi merchant cart. So the single cart, and it’s honestly been so awesome to see, like, every single day is our highest day of, you know, amount of purchases with almost no marketing. Like, we just turned on some of the lower funnel, Google Shopping, search media channels. We just turned those back on in more of a kind of directed way. And so, you know, and we’re seeing a lot of folks shop between retailers, which is another really cool thing to see. So it’s, we’re starting to validate pretty quickly the behavior that we were hoping. And so, you know, now, I guess it’s just to the to the moon, but I think I really, when you ask me about channels, they’re kind of all important right now, but I we definitely do have a big, a big job to do as it relates to building this brand and building that upper funnel and then continuing to use a lower funnel as like a, you know, sort of we’re very, very, very fast based business. We are built on, and our processes rely on. Very heavily on AI practices. And so how do we just, like, iterate and optimize quickly? And I don’t mean in a sprint, like a two week sprint, like a technology sprint. I mean daily, hourly at this point, which has been really, really fun.

20:14 Jim Beach : What about delivery rooms to go specializes in really quick delivery sometimes that day, certainly within a week, are you using the postal UPS systems? Or you how you do your delivery?

20:30 Alex Seaman : So this is something where this is really cool, actually. So I’m thrilled you asked me this. So there were, you know, we always wanted furniture.com to evolve into this multi merchant cart. There were a lot it was hard at the beginning. Now we’re able to do it because we actually have an agent, like an agent power checkout, an agentic checkout. And what that means is the shopper is checking out in one cart on furniture.com like they would any other, you know, web experience, but in the background, we’re actually checking out on each retailer’s individual owned website, like on the shopper’s behalf. And so when they click Buy, we’re doing the same on, you know, on each website that they have products in the cart from, and so, and when I say we’re I mean, AI and LLM and AI is doing the same. And so it’s doing it simultaneously. Our, you know, our, our, like cart, shopper experiences is dynamic. So it pulls in those retailers, delivery calendars. The shopper can select a day, select, you know, the type of delivery, a time, all the same things that would happen with the retailer. And then when they press buy, you know, we make it very clear, you get an order confirmation from us, but you also will get confirmations, and then, you know, shipping and delivery follow ups from each retailer. Because, you know, those folks, the retailers, like, they’re the experts. They’re amazing, like you said, rooms to go does, you know, a lot of times either same day, next day, certainly, same week, delivery, and it’s true of a lot of our retailers. They also have amazing customer service. We don’t want to try to we don’t want to try to insert ourselves in that, because I don’t think that would be the best shopper experience. So we really want to let the retailers do what they’re amazing at. And it’s also a really, really cool bottle, because, you know, then the retailer, as well as us, like we both sort of then retain that shopper, that customer, in our own brand funnels. And so, you know, now let’s use the example of rooms to go, but let’s say it was a multi merchant cart. Let’s say it was rooms to go and One Kings Lane and Raymore and Flanagan, even though that’s in a different region, like, then each of those retailers would also retain that customer, which is, which is valuable for them. And so, yeah, it’s really cool. It’s, it’s really opened up, I think, like, you know, the last year or so, with the advancements that have come in, AI, certainly, there’s probably some negatives to that, if you think more meta, but, but for our business, it’s really helped us create a model that’s super, super, I think, attractive to the retailers and and really good for a furniture shopper, like really, really helps them. And so we’re excited about it.

23:17 Jim Beach : Well, you should be. It’s a great, great step forward aging Millennials prime market. You said they have any furniture, education, any style, quote, unquote, that is targeted toward the aging millennial. I have sort of borderline kids that age, and they have absolutely no understanding of furniture. They’re just buying whatever clicks and, oh, there’s the buy that. Yeah, you know, whatever the four roommates like or something. You know, it’s just, there’s no logic to purchase it. Yeah, I

23:58 Alex Seaman : Think you hit the nail on the head. Like, I think it certainly varies person to person. When I say aging millennial, I mean that now the folks who you know, so the average furniture shopper is a woman in the US is a woman. She’s 40, you know, like she and so now, if you think about millennials, we’re quite, and I myself am a millennial. We’re like quite in that bucket. Those folks are, you know, having kids and and moving into homes, and it’s just a big time for furniture purchase. They also have more disposable income. And so, you know, we’re thinking about their needs. We’re also thinking about kind of, I think, you know, specifically our shopper, we want to make sure our end, by the way, this is a huge segment of people. Like, there’s more than billions of like, of billions and billions of dollars and in opportunity with this segment. But she’s really, she or he is really somebody who is shopping for furniture online. First of all, and there still is a lot, or at least there. Doing some degree of their furniture, discovery, decisioning online. Maybe they’re not purchasing and and, you know, because in store, it’s still really important too. And they are that person who sort of taught at that moment of decision fatigue. So they have a lot of tabs open. They’re trying to shop across a lot of brands. Maybe they have an idea of their style. Maybe they don’t, maybe they just kind of know what they like. They don’t know how to articulate it. But we really want to help them at that moment. And so@furniture.com we find the majority of people coming in are really coming in. And we did a giant sort of consumer insights research project, or project, really project, I guess, around this for five months, while we were sort of at the inception of replatforming and rebuilding our shopper experience around this time last year, and we found that people are really starting, the people who are using furniture.com like the people who were solving, You know, a part of their furniture shopping journey, they are really searching, like, first and foremost on like, visuals, right color, style or a version of that and category, like, I want to a white bouquet sofa. I don’t know if I like modern or contemporary and like that way. I just said it. That’s how people search these days. And so we knew we had to be able to one have semantic conversational search and chat environment, but to have the expertise of it built in that, first of all, very quickly, like search has to be very fast, has to be really fast, integrated into the PLP, has to be very visual. And then, you know, they use this, the semantic search, the filters, which are, by the way, like they’re they, they’re persistent throughout and if they use Dotty, then it becomes context aware, like you would see in another LLM, like a GPT or something. And so very quickly, narrow down. It’s okay, this is my price range, like you said. And maybe, okay, maybe now I recognize some of these brands, but I didn’t realize that they had this style I liked, or I didn’t even realize that my style was called, you know, Mid Century Modern or I didn’t, you know, even think of this brand when I was shopping for my sofa, but I’ve heard of it, and I trusted, and I know that they’re like a trusted retailer in my area, or whatever, and so that’s sort of how we see people shopping. It’s very different than, you know, people who are going directly to a brand, who are brand loyal already, and with for our retailers, of course, that comes up like, that’s come up throughout this process in the last couple of years, as we onboarded a lot of these folks and like, those are not the shoppers we’re going after. Those are your shoppers. Those are their shoppers. Like we’re trying to find the ones that need help figuring out what they want. They have an idea, maybe, or they can just come in and start discovering and start chatting and, you know, looking. We also have non product content, like something you might see in a social media feed or a Pinterest board. We have that which also helps guide people, helps inspire people in their design journey. So it really is the gambit, you know, of folks who come in, but it’s certainly not like somebody with a, I mean, it can be somebody who’s like, very directed, and they do a very, very like, you know, long, long tail search I’m searching for, I don’t know, like a leather and wood mid century modern sofa that seats five that can deliver before my in laws come in town next week. Like you might have that, but you also might have the person that just knows they want, like a wood, maybe modern, not sure, sofa for for, you know, three to six and so. Or you might see someone who’s, I need a sofa. My roommates and I move in next week, like, these are the types of searches we see. And so it’s really kind of dentist. I mean, furniture.com is broad. We certainly are solving for somewhat of a specific shopper in the furniture industry at large, you know, online. But I think it’s, there’s, there’s a lot of different types of shoppers and needs that we see come through and we, we hope to solve pretty seamlessly.

28:59 Jim Beach : Well, I think you’ve done a great job so far. It looks like absolutely the best site. And you know, I know this space fairly well. I got kicked out of Georgia Tech architecture school about five days before graduation.

29:12 Unknown Speaker : Alex, no way.

29:16 Alex Seaman : Oh, so you’re definitely not a design novice at all. I think the two go super hand in hand,

29:21 Jim Beach : Basically a Masters of architecture just three days. Yeah, you know, but right? Fantastic. What you’re doing. I’m very impressed. And thank you so much for being with us. What’s the URL again?

29:36 Alex Seaman : Furniture.com everybody should check you out if you want one place to discover and decide and then purchase across all the trusted retailers. It’s your place. Lots of fun things coming, honestly features every day. So we’re really excited to get it out there and for folks to start using it to help them design their homes well.

29:57 Jim Beach : Thank you so much for being with us. Great. Great success, and we’d love to have you back in a year and get an update. I would love

30:06 Alex Seaman : To thank you so much, Jim for having me. This was really fun.

30:08 Jim Beach : Thank you, Alex, and we will be right back. You.

30:20 Intro 2 : Olis, well, that’s a, that’s a, that’s a wonderful question, actually, oh my gosh, I love the opportunity to do this. Thank you, Jim, wow, that’s, that’s, that’s a great one. You know, that is a phenomenal question. That’s a great question, and, and I don’t have a great answer, that’s a great question. Oh, that is such a loaded cleft, and that’s actually a really good question.

30:43 Intro 1 : School for startups radio,

30:46 Jim Beach : We are back and again. Thank you so much for being with us today. Let’s go and talk a little bit about AI and the legal aspects thereof. I’m excited to have Spencer Rubin with us. He is an attorney at grellis Shaw legal firm. They are based out of San Francisco, but he is in the Denver office. He specializes in all things legal and AI, transactions, startups, global brands, AI, governance, IP strategy, anything that would interest us. He is the lawyer for Spencer. Welcome to the show. How you doing?

31:22 Spencer Rubin : I’m doing great, Jim, thank you so much for having me today on school for startups radio. That was a lovely introduction, and everything you said was completely accurate, right? So I don’t know if you asked chat GPT or Claude to give you that summary, but you did much better than they would have ever done.

31:36 Jim Beach : Am I right? This is what your publicist sent me so well. There you go, chat GPT as well, right?

31:45 Spencer Rubin : Stands to be determined, but yes, I think I am. So we’re here today to talk about AI and the law. And there’s lots of different directions we could go in, but one of the things that’s been really salient to me recently is going to be how folks are using their name, image, likeness, in the law in contracting, right? If you have the news coming out recently, they’ve been able to now essentially take a bunch of celebrities, name, image, likeness, and turn them into a bunch of parodies. And I believe it was a company associated with bytedance in China and a bunch of the studios in LA sent them cease and desist letters saying, like, Wait, what is this parody of Titanic that we’re seeing like run amok on Tiktok or Instagram? And it was very, very, very realistic portrayals, for example, of Leo DiCaprio and Kate Winslet, to the extent that they were like, This totally violates these celebrities, name, image, likeness, but also is likely going to infringe on, for example, the studio’s IP so that’s been really salient, happy to talk about that or anything else today, really.

32:44 Jim Beach : I appreciate you bringing that up. I mean, I’ve been to China many times. I’ve done business there. They’re just going to give you the finger, aren’t they? I mean, they don’t care if we violate their IP, or, yeah, we don’t. They don’t care if they violate our IP.

33:00 Spencer Rubin : Well, generally speaking, you’re correct. So the only leverage, I think, for example, that US studios would have over bytedance is the fact that bytedance is sort of spun off Tiktok, but obviously there’s still going to be some financial consideration on the back end for them. I don’t think they totally divested their ownership. So you can put a little bit of leverage to the extent that bytedance still holds any portion of Tiktok ownership, right? But I obviously now they’re not the majority owner due to sort of what was put out in the public, at least as national security interests, but correct other than the bytedance scenario, yes, it’s very difficult to enforce your intellectual property ownership and rights in China with regard to AI one, because it’s just difficult to track down who the would be infringers might be in China, there’s a little bit of a jurisdictional control issue, meaning that when you’re trying to search out potential infringers where they’re using AI technology or otherwise, it’s hard to enforce a judgment against them. It’s hard to find them they might be on the dark web. The Chinese government does not really cooperate with the US government on these matters. I mean, think, for example, too, of like trademark infringement with regard to the gray market and black market for knockoff goods, right? I mean, if Louis Vuitton and Gucci could enforce their trademarks adequately in China, you wouldn’t have any of these factories that make all these goods and distribute it worldwide. You walk through any airport in the US, you see one out of three women carrying a Louis Vuitton or a Gucci handbag, please, like most of those are fake, right? They’ve all come from China. So if even the largest fashion brands in the world cannot control trademark infringement or counterfeits, it’s going to be, obviously, equally as difficult for the studios or any copyright holders, for example, in the US, to enforce infringement on their their movies, their their listening, sorry, not their listening, their podcasts. Right celebrities to enforce their rights and their images. But there are things that you can do you know, outside of the infringement, to help control and monetize your personal brand, with regard to AI, which is. Is sort of a pivot here, but also really important for people that are looking how to how to incorporate AI into their content, but also make sure they’re compensated adequately.

35:08 Jim Beach : Tell me more about that. How do I use my personal brand with AI?

35:13 Spencer Rubin : Well, this is a great question. So first of all, if you want to negotiate on your behalf by yourself, Godspeed. More power to you, generally speaking, whether you are an influencer or a celebrity or anyone in between, you want to have a lawyer that’s going to help you negotiate this right. And the moral of the story is, there are two things you can do to really make sure that if someone’s going to use AI to generate content using your name, image and likeness, that you are putting these guardrails up on these two items. The first is the scope of the license, and the second is the monetization of the content. So with regard to the scope of the license, if you are going to license out your personal brand, your name, image, likeness to anyone that’s going to manipulate it and commercialize it using AI, you need to make sure that there are restrictions on what your output from the Gen AI can be used for. So first of all, the company that’s maybe manipulating your brand or your name, image likeness, let’s just say, for example, you are a celebrity, and you’re doing an endorsement for an insurance company, right? We see a lot of those out there on TV, especially during football season, and you are in negotiations with the celebrity and the insurance company, and they say to you, we own the output from any Gen AI that we create from you, and we have the right to make further reproductions of your name, image, likeness for future commercial spots. Maybe you film one time, but we’re going to create six separate commercials with it. And if that’s all the contract says. You’re screwed, because if they say, We own the output from the Gen Ai no other limitations on it, then who’s to say that they can’t sub license that downstream, that they can’t sell it, that it can’t be used in some kind of compilation down the road, right? So as the talent, as the celebrity, as anyone that’s trying to manage their their brand, their name, image, likeness, when AI is going to manipulate it and potentially make derivatives, they make derivatives. You have to one on the license scope say you may own the output, however, you agree that you will not use the output for X, Y and Z. So you cannot use the output and sell it to a third party. You cannot use the output and create any content beyond what is listed in the scope of work for this contract, and you cannot own the output and essentially walk away scot free create a derivative that may hurt my own personal brand or my own personal reputation. I think the word in the in some of the legal arguments tarnish, so you have to put in there specifically you will not make any use of the output that will tarnish my reputation. Otherwise, if you do, it’s a breach of contract. I’m going to come after you for damages. So a lot of times when you’re working with technology companies, specifically ones that will implement AI and they just say, flat out, we own the output, we can train our model with it. You know, sorry about it. That may be fine for certain uses. If we’re just talking about like a data analytics or a regular Gen AI that I’m going to put in prompts to help me do my math homework, but it’s way different than when we’re talking about celebrities, personal brand influencers, name, image, likeness. So that’s one the guardrails. You cannot have them tarnish your image or use anything downstream that would be outside the scope of the contract, or sell it to third parties, and then you have no control. You don’t know who’s going to use it. The second thing that you have to be aware of is the monetization. So again, a lot of Gen AI companies are used to dealing with just raw data, prompts, things that are low end in terms of monetization capability, right? But when you’re talking about using someone’s name, image, likeness to help promote something, and you might want to use it further downstream to help monetize your brand, again, we can go back to the insurance and the celebrity example, they might say, hey, look like we’re going to pay you $100,000 to come film this spot for Nationwide Insurance, and we’re going to cut you the check after you film it once, but we have the right to use your name, image, likeness, downstream and as many number of commercials as we want to. Thank you so much. But you know, you just got six figure check, and you’re never going to have to come in more than once to film. Sounds like a sweetheart deal, right? Like, Oh, I lifted a finger one day and I got six figures. It’s not a sweetheart deal. And the reason being is, if they have unlimited right to use you down the road for whatever iteration of your name, image and likeness, they could be accruing multiples and multiples and multiples of revenue based on using you as a celebrity without adequately compensating you. So let’s take the reverse example. Before AI, let’s say you’re the celebrity. You go into film for Nationwide Insurance, and you’re going to have to go in every three or four months and film a different commercial spot. Maybe for each of those commercial spots, you were going to get paid a flat fee of $50,000 right? But like, if you’re going in four times a year, you. Your $50,000 has now become $200,000 whereas if they said, Hey, you come in once and we’re going to film you, but we can create as many derivatives using our AI as we want to. And here’s $100,000 you’ve just made 50% less than you would have actually made had you gone in three more times than the one that they asked you to come in. And so you actually see this sort of law of diminishing returns as a celebrity or an influencer. If someone is going to try to tell you, we’ll pay you a lump sum for your endorsement, but we have the right to use AI to manipulate you, to put you in you know, subsequent advertisements, subsequent marketing campaigns, Omni channel spreads. This is going to really impact the celebrity in terms of what they actually should be comped. Now, again, you and I are sitting here. No one’s necessarily asking us to, you know, endorse Nationwide Insurance and cut us a, you know, a $50,000 check or $100,000 check. I wish, right, but this really is important for folks that you know, are making this type of money that could potentially lose out on, you know, significant amounts of income from AI taking over. So what you have to do, if you’re in the celebrity shoes or the influencer shoes, protect your purple brand, is you say no, if I was going to make X number of dollars doing this in the traditional way, where I come in four times a year, and film each spot. You are not going to use AI the brand in order to essentially cut down on what my anticipated income would be from doing this promotion, because I don’t use the celebrity you want to say, I don’t want you to cheapen again my name, image, likeness, and I’m not trying to give you essentially access to me, a celebrity, for you to just be able to pay me, you know, a little bit, and then monetize all of this profit off of me that you wouldn’t have normally been able to. So in the contract, what does that mean? Basically that means, again, like we had said, for example, okay, you know, we’re going to put some guardrails on the scope, meaning, if they own the output. Here are the guardrails on subsequent use. Here are the guardrails on who you can sell it to. Here are the guardrails on not tarnishing my image. You put those same guard rails on the number of times you film and what the actual compensation is. So one of the things that would be very useful is, if you are used to a lump sum payment for an endorsement. Do not agree to that right out of the gate if they are going to be using AI to create subsequent marketing materials. Of you. So maybe, for example, what you have to do, and this is just starting to come to the forefront, so I don’t have the panacea, cure all for it, but one of the things you can do is you can say, okay, look, instead of a lump sum payment for me to endorse this brand, I’m going to actually track it more on a traditional advertising basis, wherein, for example, you see clicks on ads, and sometimes revenue is based on the number of clicks an ad gets. You want to have the same thing be done if you are going to be monetizing your name, image and likeness. So how many viewers did I reach, if we’re talking about traditional Nielsen, how many clicks did my ad get? Whether it’s a video or some kind of like image based ad, start getting monetized off of that, right? Because if it’s if it’s lump sum and it’s aI game over, because they’re going to create as many derivatives as they want to if they have the right to. So you want to make sure that you’re not locking yourself in for a lump sum payment upfront, which looks really juicy and awesome, only then to not be paid down the road when they are continually using your name, image and likeness through all of these derivatives that Gen AI can spit out of you.

43:35 Jim Beach : All right, a ton of great information there. Spencer, I want to go kind of off topic for a second, and then we’ll come back to AI in just a second. But one of the thesis points of this show is that you don’t need to have an original lightning bolt idea to start a business. You should just go copy, borrow or steal someone’s idea and make it better. Of course, we are going to respect IP trademarks, copyrights and patents, but if you don’t have an idea, just go copy someone else’s idea. Give me your legal opinion on those three or four sentences.

44:08 Spencer Rubin : Well, that’s a great way to phrase it. If you don’t have your own go copy someone else. I would phrase it a little bit differently. I would say, if you don’t have your own idea, go learn from someone else. Right? Because the scope of copyrights, patents and trademarks is only so broad in the world, right? Like, if I see a phone being sold at Verizon and it’s the Apple iPhone, I’m like, I must create my own smartphone, and I do it from scratch. And yeah, I’ve hired some technology developers. Yeah, I’ve hired the same manufacturing plant or a similar one in China to actually make my device. But I haven’t actually gone in and copied the apple code and copied the design specs and really, just like, ripped every little bit and piece of it apart so that I’m making my own. But it’s really just a complete copy of the apple. Then really, in that scenario, it’s. An idea, and I’ve learned from their product, but I actually haven’t copied something in a way that would infringe a patent or infringe a trademark. Let’s talk about that was a broad example, right? And it’s salient because you’ve actually seen Samsung and Apple suing each other over patent infringement. But let’s talk about it from the startup perspective, right? Like, let’s say that you want to go out there in the world, and you want to make some sort of AI service, right? It’s going to be some sort of platform as a service. It’s going to be, you know, available to consumers, and you’re going to basically offer them some kind of, you know, product through the AI platform. So let’s say it’s really high end image generation, right? There are tons of softwares out there already that can do that chat. GPT has Dolly. I think the Claude can do pretty high image, not drafting, but generation. For you, there’s something called Flux Dev, which a lot of people in the movie industry use to help them generate videos and images. It’s really high end. So if you want to do something like that, great. It’s actually not that hard. Go hire a contractor if you have the budget. If not, hire the contractor and give them some equity in your company once you take off, right? But make sure that when you get yourself and your team composed, that you’re not actually, in this example, ripping the source code out of whatever model you think you’re going to base yours on, right So learn from what it does. Figure out how to code something that’s going to do high end image or video generation, but don’t rip the source code a lot of times too. You can build a model on top of a pre existing one, as long as you have the right API license. So a lot of my clients right now build the backbone of their AI model off of chat GPT because they have the API license with open AI. That’s totally fine, but no one’s going in surreptitiously into open a eyes code and figuring out how chat GPT runs and then building their own baseline model based on that source code. So in AI, for example. To your point, absolutely, go out there and learn from what other people are doing if you don’t have your own original idea, but don’t copy because, specifically in source code and the AI model world, if it can be proven that you have actually just ripped source code and it was not open source, meaning freely available to the public, feel free to copy it. Then, if it was not that, you are going to be an infringer. So sometimes just kind of have to look, and you have to touch and you have to find people that can help you do what you want to do, especially if you are trying to figure out how to pay them. You don’t have the right amount of money as a startup, that’s okay. You will figure it out, but almost kind of like the car analogy. Look at the car, feel it, test drive it. Take it for a run. Decide how you want to build your own. But don’t open up the hood and start ripping components out and then putting them in yours.

47:53 Jim Beach : That’s what you in trouble, yeah, but copy, borrow steel is a lot sexier than what you just said. It’s easier to remember. All right, we really need to move on. Spencer, we’re almost out of time already. I have a new addiction. I am writing a new book, and I put each chapter into chat GPT and ask it to give me a grade, and it does. And now I’m addicted to getting a 9.7 for every chapter in my book out of 10, and it’s a little you know, you get suggestions. I incorporate the suggestions. The chapter gets a little bit better, and eventually the whole book might be a 9.7 out of 10. I have asked chat GPT, Can I quote you on that? Chat GPT, can I say that chat GPT gave this book a 9.7 out of 10, and chat GPT doesn’t really know how to answer that question. I think it gives all sorts of answers that are all over the place. What is your thought good on the cover of my book, I put chat GPT 5.2 says this book is a 9.7 out of 10

49:02 Spencer Rubin : That is funny. And my answer to you is no, it’s funny. First of all, because I think a lot of people have this idea that what chat GPT tells you specifically is the gospel, right? But if everyone is receiving the gospel from chat GPT, then what is the actual gospel, right? There is, there is an objective truth that’s out there for everything, right, or close to an objective and then there is a subjective and chat GPT, along with other chat bots that are, like, commercially available in the market, that are really, you know, highly infiltrating the population, right in terms of usage. So chat GPT, Claude, co pilot Gemini is a big one my clients use. These are all built, and you can tell, because you know based on the answers it’s giving you when you prompt it, they are all built to appease the user. They are all

49:54 Jim Beach : Got that wrong. Yes, it does a very good job of that.

49:57 Spencer Rubin : Yes. So when you say to. It. Oh, please, you know, let me know what my score on this chapter of the book is. And, you know, let’s say you get back something originally seven, right? And you’re like, Okay, well, is it really a seven? Is it closer to a nine? Tell me why it’s not a nine. You can even talk the algorithm into giving you a higher score based on the prompts that you give it, because it’s trying to appease you. So putting an endorsement on anything that’s a chat, GPT said, This is great. Well, I’m sure it said it was great in some iteration of the prompt you gave it. But how do we know that the first time it spat out an answer to you, it actually said it was good, you could have manipulated the answer to get what you wanted. Also, it’s giving you confirmation bias. It’s trying to tell you that you’re good so that you’ll keep using the app, so you’ll keep using it

50:43 Jim Beach : As a movie critic, giving it five stars. I have to take that with a grain of salt, knowing that they also sent that movie critic to watch the movie in Cabo st Lucas for a week. And that’s buyer, you know. I have to trust whatever you know. What’s the difference between trusting, you know, USA, today’s movie score, versus chat, GPT, you always have to do that and judge whether it’s a legitimate source or not.

51:11 Spencer Rubin : That’s fair. But let’s think about it this way, right? And I’ll give a really good concrete example, Cisco and Eber, or then it was whoever, after one of them died, then it was Roger and Eber, right? Whoever it was, there were two. There were two dudes, right? And so, yeah, they could have been flown around the world and wined and dined by the producers in order to give them, you know, a really good two thumbs up score, right? But at the end of the day, there were only two of them. And as each individual, there was one brain in each of those bodies, right? There is not one chat GPT brain. There is not one anthropic brain. There is not one Gemini brain. There are instances of this platform on every single person’s computer, and the more that interacts with the user, the more that instance of the model is going to be accustomed to giving that user the answers that they want. Because it is, it is coded to give you confirmation bias. So whereas with Siskel and Ebert, or Roger and Ebert, sure, they could have had a bias because they were induced to give a good rating because someone took them on $100,000 trip to the Caribbean and show them the movie at their, you know, palatial villa in St Kitts and Nevis great. But like that’s just coming from two people, one brain each chat, GPT is giving every single user the answer that things they want to hear. So where is the unified source of truth in that however biased, it’s not because you can manipulate your own instance to give you a more unified version of the truth that you want.

52:36 Jim Beach : All right, damn it. I’m gonna have to redo the cover of that book. Spencer, thank you so much for being with us. Great information. And boy, did we blow through our allotted time quickly. And we’d love to have you back, because we have a lot more stuff to go over. How do I get in touch with you? Hire you and help with my AI legal issues or my startup issues.

52:54 Spencer Rubin : Appreciate that, Jim. First of all, thank you so much for having me today. It was lovely to be here and talk sort of waxed philosophic, so hopefully all the listeners will be like, okay, yeah, come on, more AI, because we want more people talking about it and to be excited about what’s to come. So if people want to get in touch with me, it’s very easy. You can google our law firm, gorillas. Shaw, LLP, out of San Francisco, California. Folks can email me. It’s just s r@gorillas.com and or, if people are feeling very adventurous, they can text me at my phone number. Now, I don’t know if you want me to leave that here. There you go. Great. Okay. Well, you know what? They can just find me on my website or email me, as I said, fantastic.

53:35 Jim Beach : Spencer, thank you so very, very much. And I am serious, we’d love to have you back soon and dive deeper into this and thank you so much for being

53:42 Spencer Rubin : With us. Thank you so much, Jim. I really appreciate your time.53:46 Jim Beach : We are out of time. Thank you so much for being with us. Hope you have a great day. Take care and go make a million dollars.