Subscribe
Share
Share
Embed
AI is the biggest technology shift of our lifetime. This show is about how to profit from it together. Each week I talk with the founders and CEOs closest to AI and Content, the ones figuring this out in real time. I’m also building an AI content business myself and share the lessons I learn along the way.
WHAT WE COVER
THE TITANS: How companies like OpenAI, Anthropic, Meta, and xAI are moving, and why their decisions matter.
THE INCUMBENTS: How content giants like Disney, News Corp, Universal Music Group, and Reddit are responding to AI, and what it means for creators and publishers.
THE PLAYBOOK: Real lessons on AI business models, content strategy, IP licensing, distribution, and getting paid.
ABOUT YOUR HOST: Rob Kelly has interviewed Steve Jobs and Bill Gates, helped pioneer early web content licensing, and built multiple companies with more than $100 million in total sales. His work has appeared on CNBC, CNN, TIME, and Entrepreneur.
Beyond business, every episode explores what AI means for jobs, creativity, families, and the next generation.
If you want clear thinking based on real experience in AI and media, Media and the Machine is your guide
Thanks! -Rob
I'm Rob Kelly, this is Media in the Machine, a show about the biggest technology shift of our lifetime and how to profit from it. Each week, I talk with the founders and CEOs closest to AI and content, the ones figuring this out in real time. I'm also building an AI content business myself and share lessons of what I learned along the way. You know, life's funny. I began my career lucky enough to interview leaders like Steve Jobs and Bill Gates.
Rob Kelly:Then I went on to be a three time founder and CEO, driving a $100,000,000 plus in revenue and some failures too. And now I'm back at the table, interviewing this new world's current and future leaders. This isn't only a business story, it's a human one. So every episode ends with me asking my guest what AI means for our jobs, our families, and the next generation. We'll figure this out together from the inside.
Rob Kelly:Welcome to Media in the Machine. My guest today is Ty Ahmad Taylor. What I love about Ty is that he transcends media and tech. He is founder and two time CEO, and most recently a product VP at both Meta and Snap. He's also on the boards of GoPro and SFMOMA.
Rob Kelly:He started his career at the New York Times, then went on to found and sell his sports startup Fan to Samsung. And now he's getting his hands dirty with AI. He shares how he just rebuilt a startup that once took him four and a half years in just five minutes. I asked for Ty's take on media in the new AI era, such as why Yahoo Sports is showing up everywhere in AI products while The Athletic for New York Times is almost invisible, what that means for content companies deciding whether to block or embrace the AI bots, and the AI business model he's most excited about for media companies. Tai also breaks down how leaders should think about plummeting traffic from Google and what replaces that old distribution playbook.
Rob Kelly:He also shares insights from a recent workshop he led with OpenAI, Perplexity, Delphi, and Listen Labs all in attendance. And he lays out his three vectors framework for AI product development. A special thanks to Jen Collins for introducing us. Oh, and a quick note before we dive in. Ty and I talk about Apple not having its own LLM and SFMOMA not doing anything in AI.
Rob Kelly:Well, since we recorded, Apple and Google announced a deal where Siri will now be based on Google's Gemini, and SFMOMA launched the latent space, an interactive AI installation. Just shows how fast things are moving or how slow I am in getting podcasts out. Anyway, let's get into it. Please enjoy my conversation with Ty Ahmad Taylor. So I want to get your take on The New York Times.
Rob Kelly:You work there. They sue OpenAI, but then they do a deal to license their content to Amazon. How would you think about that if you were still at the New York Times?
Ty Ahmad-Taylor:So I think for a company like the Times or any content generating company, the the set of challenges that occurs is like, how do I not enable the LLMs to sort of steal the chickens from the chicken coop, so to speak. That if you don't optimize for LLM consumption, you are going to be trumped by media companies and by content generators who do. And so I ran my own startup which was called Fan Feeder from 02/2012 before it was acquired by Samsung. That startup, we spent four and a half years trying to build a sports aggregation site that would aggregate all of your favorite sports news and information and distill it so that you didn't have to go hunting and pecking across multiple websites or applications to find the best news and content around your information. I've rebuilt that entire thing with a single prompt.
Ty Ahmad-Taylor:My four and a half years worth of work took me five minutes, fifteen years later.
Rob Kelly:You you built this recently just for yourself as for kicks? Or
Ty Ahmad-Taylor:I just I just built it for myself. It is my solution for keeping up to date. I like the Warriors. I like the 49ers. And so I get at 6AM and 5PM wherever I am in the in the world, I get automatically pushed updates.
Ty Ahmad-Taylor:A delta from the last push that basically replicates the features and functionality of the app that I was building with my team for four and a half years.
Rob Kelly:What'd you use to build that?
Ty Ahmad-Taylor:A prompt that I put into ChatGPT, Gemini, Claude, and Perplexity. My point in all of this is that and the reason for giving the background is as a result of doing that, I see a lot of Yahoo content. And I wasn't spending a lot of time at Yahoo because Yahoo's optimized for LLM consumption. What I don't see is a lot of New York Times content, specifically from The Athletic, which is their sports destination. Because they're either gatekeeping or not allowing the LLMs to access their content in a full and robust manner.
Ty Ahmad-Taylor:And so while you may be waiting for payment from LLMs from training on your content or access to your content, what you're doing is denying yourself an opportunity for content ubiquity and then making money either through ads or through affiliate fees. And so that's the trade off.
Rob Kelly:So you rebuilt fan feeder which took four years and how long did it take you?
Ty Ahmad-Taylor:Five minutes.
Rob Kelly:I love this stuff. So you're getting more content from Yahoo Sports than The Athletic because your guess is The Athletic is stopping the bots from scraping Precisely. And Yahoo is being more open. I'm just curious, which side would you do if you're running Yahoo or The Athletic?
Ty Ahmad-Taylor:I would be more of like Yahoo Sports. I'm not critiquing what The Athletic is doing. They have reasons for doing what they're doing. They wanna make sure and guarantee that they're being paid. But in this messy transition period that we're in right now, I think a lot of companies are looking for I wanna make the money that I have been making.
Ty Ahmad-Taylor:And again, my sense of things is that that has changed. So what are you gonna do with the new ground truth? I think the larger question facing content companies is that they are in the middle of a transition from being ad supported to what I believe is gonna be a new revenue model or economic model, which is to be affiliate supported. So how do you provide affiliate links to the things that I'm interested in based upon my demonstrated interests? And subscriptions and and ad based models, how they have been making money.
Ty Ahmad-Taylor:But what I tell any company is that, you know, how you make money today is no guarantee of how you make money tomorrow. Because product market fit today is not a guarantee of product market fit ad infinitum. And so evolving product market fit means also evolving your revenue model. And I think things are transitioning with the advent of the frontier LLMs into an affiliate revenue model. And I don't think any of the content companies have really figured out how to do that.
Ty Ahmad-Taylor:But what Jim Linzon, who I used to be on the board with at GoPro who's running Yahoo is doing, is that he's taking a ubiquity play so that Yahoo content is unavoidable. I'm not telling you that I know that for sure, but I'm saying that that's what I found in in the course of getting aggregated news and information.
Rob Kelly:If you're a company right now, content media company, who's seeing a huge plummet in traffic from Google, how do you think about using AI to replace that traffic?
Ty Ahmad-Taylor:I think you have to think about how does your brand and how does your content show up in the LLMs and understanding again this transition between zero click results and affiliate generated results. If I were running a company, I would say how do I optimize that I'm I'm showing up everywhere and I'm showing up in ways that drives my subscriptions, drives my impressions, drives my click through, and drives my affiliate revenue? Not necessarily in that order. That's the opportunity in front of every single content company. It's a different model though than what's existed to date.
Rob Kelly:In the case of if we kept going down that rabbit hole of Yahoo in this case, and let's just say Jim's on board with it, what's an example of affiliate revenue that Yahoo and Jim can generate using the example of your personal fan feeder? This isn't a new business, by way. This is just Ty's special little AI that's hooking him up while he's traveling.
Ty Ahmad-Taylor:It's for personal service. And then for some of the devoted fans of the fan feeder site, I sent them the prompt. They can swap out my team names with their team names, and they can get the same service on Gemini.
Rob Kelly:Oh, so I can use fan feeder. Excellent.
Ty Ahmad-Taylor:You can have your own fan feeder. Yes. Exactly. And I'll share the prompt with you.
Rob Kelly:Very cool. So back to the Yahoo. What would an example be of some affiliate revenue they might make from you getting an alert on your favorite team?
Ty Ahmad-Taylor:Sports memorabilia and merchandise and things of that sort. They have an opportunity to sell me 49ers lightly worn apparel with athlete sweat on it, no. That would be gross for me at least. But things of that sort.
Rob Kelly:You would watch that?
Ty Ahmad-Taylor:I would decrease the value almost immediately. I'm not interested in lightly worn apparel. I'm not the target market for that. But there are people who are.
Rob Kelly:In that case, would the affiliate ad, if we call it that, would that have been on the AI or would it have been on Yahoo Sports?
Ty Ahmad-Taylor:The answer is yes. It should be on both.
Rob Kelly:Some combination of both, you can see that happening. And that is just another form of advertising. Right?
Ty Ahmad-Taylor:Correct. But, you know, what you're finding is that there are a lot of zero click searches as a result of the adoption of the Frontier LLMs. But when people do click through, it's a much more highly qualified click, especially in the realm of e commerce and things of that sort. So while you might not be getting the traffic that you did before and looking at things from a volume perspective through advertising, when you do get traffic, that traffic is more highly qualified because they're showing greater intent. And so that's the transition from a traditional ad model to more of a affiliate model that I see going on.
Rob Kelly:And if we go down that thread of AI for growth, for customer acquisition, mean, I what's fascinating about you is you've been through a number of different platforms, you know, whether it's TV, whether it's social, whether it's movie theater stuff, movies. You've seen all of this. Is there any reason to think of AI as not another distribution platform if you are Yahoo Sports or The New York Times or The Athletic? Is it for sure inevitable that's gonna be distribution?
Ty Ahmad-Taylor:I think it is. But I think that the challenge is is that the revenue models are not set. The rules of the game are not set. What you see is right now is some gatekeeping and pay for play, and I don't have any great suggestions besides the affiliate revenue opportunity. And so that's upsetting because those are folks who are saying I used to make a lot of money this way or I had money that was mildly steady this way and now that's been upended.
Ty Ahmad-Taylor:And with all things in life, human beings don't like change. And so it's gonna require change.
Rob Kelly:Can any of the AI frontier models succeed without advertising as we know it?
Ty Ahmad-Taylor:I don't know. I think that they're gonna have to think of it relatively quickly. They're gonna have to develop their own sort of advertising ecosystems to support what they're doing in that space. And I think that that's very, very nascent on all of the frontier platforms.
Rob Kelly:Jen Collins, one of my favorite BD folks, she introduced us, said that you were one of the smartest folks about AI that she knows, and that you had just come back from teaching product folks about AI. What were the types of folks who were the students in the workshop?
Ty Ahmad-Taylor:Oh, they were not students. They were high level product leaders from various companies who came and heard companies like Perplexity, OpenAI, Factory dot AI, Delphi, Listen Labs, and others describe how to use their tools and how to structure their companies in such ways to sort of very much liberate and and open up the frame of thought around how you actually engage in the product management, product development process.
Rob Kelly:And you were the one kind of bringing this together?
Ty Ahmad-Taylor:I did it working in partnership with two groups. One of them was James Buckhouse, who's the design partner at Sequoia Venture Capital, and the second is with A Crew Venture Partners. And working with both parties, was able to construct a a full day in San Francisco, which were repeating in New York to enable people to essentially understand the vertical amplification of the product management function as a result of enterprise adoption of AI, which is a mouthful. My thesis is that there are basically three types of AI.
Rob Kelly:Is this what you called the three vectors?
Ty Ahmad-Taylor:The three vectors of AI. AI number one is replacing entire verticals within the industry. So instead of having a legal department or an HR department or a finance department, you maybe have one human being and that human being's work is augmented using AI. That's not the distant future. That's that's sort of where we are now.
Ty Ahmad-Taylor:And you can replace entire departments with AI. And that's not to dim the hopes and prospects of those in the field. I'm just saying the nature of their jobs are going to change because you can amplify the amount of work that one person would do. So that's what I would call vertical enterprise AI. Then there's AI that amplifies the work of a given vertical.
Ty Ahmad-Taylor:So this is also vertical enterprise AI, but it it's not meant to replace people. It's simply meant to up the productivity of the people who work in that discipline. So you've got products like Vercel, Factory, Cursor, Replit that essentially amplify the work that can be done by engineers within the engineering vertical, so that the output of a given engineer is multiplied several fold. The same thing occurs for product managers, which is that there's a set of tools that are adopted by the enterprise that enable product management. The product management is a discipline to be amplified in the amount of work to be done to be amplified.
Ty Ahmad-Taylor:That's vertical amplification, is different than vertical replacement with AI. And then there's horizontal tools. And those tools come from companies that essentially allow for agentic work across the enterprise so that any worker, not a specific discipline, is able to develop agents or is able to engage with ChatGPT or Anthropic or any of the other frontier AIs like Lama or Grok and is able to amplify the amount of work that they get done. The last bucket is really AI deployed as an end product for consumers or for businesses. For those businesses and those companies, they're deploying AI directly in their products to make everybody else's lives easier.
Ty Ahmad-Taylor:Does that make sense? And so that's what we expect during the day.
Rob Kelly:One of the unique things about you in my view is having worked across content as a and tech as a through line, both hardware and software. And I wanted to get your take on, number one, Apple, specifically with AI. Most folks would say that they're just behind and lacking vision direction. But how would you think about combining software and hardware if you're Apple in this new world of AI?
Ty Ahmad-Taylor:So I think the first part of the question depends upon whether they're deeply invested in using their own chips. And what I've seen today is that they wanna use their ARM chipsets, the m series of chips that they've created for anything that happens AI so that they can do more on device rather than in the cloud. That's my rough understanding of what they're doing. Given that that's the case, there's an opportunity for them, for business customers who use their devices, primarily I'm thinking about laptops and to a lesser extent phones and tablets, for them to be more productive using the Apple suite of products than they would be otherwise. And then for consumers, it's essentially how do you deploy AI directly in the consumer products that Apple has that makes people's lives easier, that simplifies the steps that they have to take?
Ty Ahmad-Taylor:A classic example of something that would be aided by AI that that is not done yet is I had a meeting on my calendar yesterday, and I did not have the context for how I this meeting had come on my calendar. And so I searched my email, nothing. I searched LinkedIn, not connected to the person. I was like, this is the most bizarre meeting, and I don't know why it's on my calendar because I did not recall the form in which I'd been communicated to about the meeting. It was done in text.
Ty Ahmad-Taylor:But there's no linkage between the calendar invite and the fact that I've been introduced to this person in text, is not a natural introduction mechanism for me at least. And so I was like, let me go search my text devices to find the context for why I created this meeting. And what you would want is that you'd want an on device AI service that basically acted as a a knowledge base for like, you're meeting with Susie. Susie was introduced to you by text on this date by this person, and this is her LinkedIn profile. Or here's some other context about her.
Ty Ahmad-Taylor:And so you have AI enterprise agents that are trying to stitch that together, but they don't have access to some of my text messages. I mean, you've got people who are doing smart things in text and AI, but they don't have access to my email messages. So stitching all this together seems broken. Apple's got clients across the spectrum. It seems like something that they could just do naturally.
Rob Kelly:Does Apple have to have or own an LLM specifically to succeed in the new world of AI?
Ty Ahmad-Taylor:I've heard noises in outlets like the the information that they're interested in buying Perplexity or potentially another service. And Perplexity is an unusual hybrid. It's both a service and not truly a frontier LLM. It sits on top of the other LLMs, but has some really great products in the marketplace that you climb at their browser and then their their core service. I don't know that they need to do that, but I I think that they have to use typical things, which is do customer discovery, find the jobs to be done, and then just determine how AI products can either make those things more productive or make those things easier.
Rob Kelly:And what about OpenAI buying Joni Ives' company?
Ty Ahmad-Taylor:Oh, that's that's really fascinating. Right?
Rob Kelly:And, actually, you can talk about they're not the only one combining hardware and software in the AI space. Obviously, you got your old employer, Meta, with smart glasses, arguably, in my opinion, the by far, the most useful AI piece of hardware I've used. But back to OpenAI, yeah, what's your take on on that move?
Ty Ahmad-Taylor:What enables them to do is to think about working in the voice category. And so my view of product and software development is that it has to work 95% of the time to get to the customer satisfaction levels that that make people happy. I've got two different sets of voice assistants that I've used historically. And I won't say which companies they're from because we're not trying to offend anybody. But, you know, those work about 60% of the time.
Ty Ahmad-Taylor:And I have a friend who has a tagline, David Temkinny has a tagline, Everyday Computers Are Making People Easier to Use. I have to change what I say and how I say it to be understood by the voice assistants. And so the Joni Ive marriage is like, how do you get voice to be in the 95% it works sort of realm? And voice as an input control suddenly destabilizes the operating system layer that sort of historically has been a three player game between Google's Android platform, Apple's iOS platform, and the Windows platform. That's now destabilized if voice is the primary input tool because anybody can create devices that have voice input and it's delinked from the hegemony of the three sort of like platform controllers.
Ty Ahmad-Taylor:And so I think that they'll be working in voice control in various ways that we have seen, either through glasses or home devices, in ways that we haven't seen that I can't conjure up on the moment, but I, you know, I think that they'll be thinking about. The hard part is that Meta just invested in Luxottica, which is the lion's share of the glass market. And that creates an advantageous lock in for them for glasses that others are gonna have to get around. Joni Ive and his brilliance notwithstanding, they're gonna have to think about if they're gonna break into the glass market, how would they do so?
Rob Kelly:Can we talk more about Meta? What I'd love to know is I'm a believer in the trickle down effect, especially at the founder level, and I can't think of a larger company that's founder led than Meta with Mark Zuckerberg. Based on your experience with Meta, how do you think Zuck is thinking about AI?
Ty Ahmad-Taylor:I can't get into his head and I wouldn't presume to get into his head about it, but I do know that he's been very hands on from the things that I've read in terms of their approach to what they're doing in the AI space. And there's almost not a day that goes by where their strategy around AI has not been dissected or or looked over. I think that he sees that as an opportunity for the company and he's been very good at at identifying opportunities for the company and platforms that it can take advantage of historically, whether it's Instagram, WhatsApp, or or anything of that sort. So history has shown that his vision for what he wants to do and what he views as being the next platform has been very on the nose. And when he tries things as all product people do that don't work or don't make sense, he's quick to pivot.
Ty Ahmad-Taylor:So, you know, I'm bullish on the entire sector, but I'm also bullish on the opportunities in front of them. More broadly, I know it's not a question that you asked about, I apologize, but I think we're poised at the cusp of the next industrial revolution. And I don't have a vested stake in promoting the type of change that's going to occur. It's gonna be displacing for many people. But there's not gonna be one winner.
Ty Ahmad-Taylor:There are gonna be multiple winners. And so placing bets thoughtfully I think is gonna be the opportunity. And no one really knows what those bets look like at the moment.
Rob Kelly:Can you think of any CEO of the Frontier models that could outsmart Zuck?
Ty Ahmad-Taylor:Oh, I wouldn't hazard a guess. I think there are gonna be multiple winners in different categories and in the same categories. People don't buy their cars from a single car manufacturer is is an example.
Rob Kelly:How about also the most recent big company you worked for, Snap? What's your take on how AI is impacting Snap's business? How are they doing with it?
Ty Ahmad-Taylor:During my tenure, we integrated a ChatGPT backed chat agent directly into chat so that you could converse and ask questions as you would with ChatGPT. SnapGPT? Yeah. Yes.
Rob Kelly:I'm joking. I have no idea what it's called.
Ty Ahmad-Taylor:It was called MyAI. And as ChatGPT got better with each release, those answers got better with each segment. I used it to research bakeries in Paris, you know, when it first came out and found that to be really useful and helpful. And the answers that it's given for the bakeries now are even better than than when it first came out.
Rob Kelly:How about GoPro? You're on the board of GoPro. They did something really interesting, which is asked any creators of video footage using GoPro devices if they wanna opt in and upload their content to help train AI. For user generated content companies, how would you think about this new world of AI?
Ty Ahmad-Taylor:Well, I think there are opportunities. GoPro announced publicly in late July that it was offering incentives for users to allow for their content to be trained. So you you have a GoPro, you have footage that exists in the cloud if you're a subscriber. That footage in the cloud has an opportunity to help train some of the AI learning models about what is the most important moments in your life. And so for GoPro, I think AI is really an opportunity to take the footage that's on your camera, automatically edit it, and then distill that into moments that are shareable, that are the highlights of your life.
Ty Ahmad-Taylor:And the GoPro cameras capture material in a different way than what you could capture on your phone, so it's a different market. And so that that's the opportunity in front of that company.
Rob Kelly:I heard you say that your true line in your career has been connecting people with the information that they need most.
Ty Ahmad-Taylor:Yes. Can you
Rob Kelly:just share your take on that?
Ty Ahmad-Taylor:Yeah. What I did until I got to Meta was that I connected people to the news, information, sports, entertainment that mattered to the most, whether it was at the New York Times, Comcast, Samsung, Viacom. When I got to Meta, that flipped on its head and I became deeply enmeshed in how do you connect businesses to the right people, to the people that matter most. So an inverse of what I've done historically. And I did the same work at Snapchat.
Ty Ahmad-Taylor:To be clear, didn't do that myself. I did that with the team and I did it under the auspices of what both companies needed. But when I think about that, it gave me an interesting perspective about these are the things that consumers want and then these are the things that businesses want. And the space where the two overlap, that's fertile territory to sort of you know, there's a gentleman who's CTO at Meta named Andrew Bosworth, who goes by the nickname Bos. And he said that the the whole mission or gambit is like how do you make ads suck less?
Ty Ahmad-Taylor:That very much stuck with me. Like, how do you make marketing not get misrepresented as content, but how do you make it function in the same way that content does so that people get the same value that they get out of content? And I think that's very smart.
Rob Kelly:Like the sweaty jersey.
Ty Ahmad-Taylor:Like the sweaty jersey. And if you think about Jonah Peretti at BuzzFeed, he was doing the same thing. He's like, how do you make ads behave like content so they're delivering the same amount of value? So it's a recurring theme among some of the brightest minds in the industry. Therein lies the opportunity.
Ty Ahmad-Taylor:I don't know that it's been solved yet, but a lot of people have come very close.
Rob Kelly:Can you talk about the things you've invented and patents you hold?
Ty Ahmad-Taylor:Yeah. So the bulk of them came when I was at Comcast, and I was doing things that I was learning on the fly like set top box design. But some of them were also around user interface components for the visual display on a TV set top box. And Comcast has really greatly evolved since my time there, and they have the Xfinity platform that has some really beautiful, elegant user experience directly on the set top box. But that was that was the bulk of my patents.
Ty Ahmad-Taylor:I think I have one that I got with Samsung in Korea that is you're playing games on a set top box or on a TV set. It notifies you that that your friends are playing games. There's like a notification system for games. But the invention process, I think, is really fascinating to me. It's something that I've not done a lot of in the later stages of my career, and I got away from it to my regret.
Ty Ahmad-Taylor:But I think there's a lot of opportunity in this space, and I love inventing things. I have a friend, sorry, as a side note, I have a friend who we'll unnamed, but he and his daughter every night before bed would come up with a new invention. And I was like, what a fantastic way to spend time with your child. My children are too old for that. I don't think they wanna spend that time with me.
Ty Ahmad-Taylor:But I love that as a as a device to really engage your younger children and and foster their creativity.
Rob Kelly:And AI makes that endless now.
Ty Ahmad-Taylor:Correct. You can visualize it.
Rob Kelly:You went to cooking school. Can you just share that experience?
Ty Ahmad-Taylor:Sure. I joined my first job in 1996, and that company was did very well until it didn't. It was a company called At Home, and we brought a search engine called Excite and then became known as ExciteHome. So I was just working seventy plus hours a week and I I got burned out. So cooking and photography were two things on my radar where I felt like I did not know enough and I just learned about photography from the head of product at GoPro actually.
Ty Ahmad-Taylor:He opened my eyes to f stop and and aperture and ISO in a way that I did not know before. Well, I went to the Prue Leaf Cooking Academy. She's now known in the Great British Bake Off, but she was not well known when I took the courses in The UK.
Rob Kelly:Are you using AI for recipes?
Ty Ahmad-Taylor:No. I I use the New York Times cooking recipe app and Bon Appetit. Those are my my go to sources. They're reputable, and, you know, I like the taste of things that have come out of it.
Rob Kelly:Well, it's funny. When you take a scientific approach to it, you don't want to cut corners that AI definitely can do in that case.
Ty Ahmad-Taylor:Could you imagine a hallucinated recipe? And it's like, oh, you want Thomas Keller's chicken. This is how you make it. It just makes up the recipe? That sounds like a disaster.
Ty Ahmad-Taylor:And also, a potential case, salmonella. So let's not do that.
Rob Kelly:Tell me about the Frankie Knuckles story and including the artist who owns the collection.
Ty Ahmad-Taylor:Frankie Knuckles was a seminal house music DJ from the nineteen eighties and nineteen nineties, and perhaps his most famous song is called the whistle song. And Theaster Gates is an artist based in a living artist, to be clear, who's based in Chicago who acquired the Frankie Knuckles archive. And what I've been doing is helping them realize Theaster and his team's vision for how to make the Frankie Knuckles archive manifest in a public way. And I won't talk about the archive itself, but what I'll talk about is my product development process that was used to take what would typically be one to three months worth of work and truncate it down to around fifty five minutes worth of work. Step one was, is I listened to requirements using a automated AI note taking tool called Granola.
Ty Ahmad-Taylor:The key three questions that were asked of the team were, what is the problem that they're trying to solve? Who is the audience for that solution to the problem? And then the last component was, what does success look like in eighteen months? I asked those three questions over the space of twenty five minutes with follow-up questions. Granola takes an entire transcript of the call and edited it down into two essential pages of notes.
Ty Ahmad-Taylor:And then step two is taking the two pages of notes and deploying those into a custom built LLM called chat PRD, which is a tune to generate product requirements documents. So it's chat PRD, which stands for product requirements documents. The product requirements generation process can take anywhere from two to eight weeks, roughly, depending upon the the scale of the project. And it was able to spit out a PRD in about ten minutes. The PRD was a little bit under 20 pages long.
Ty Ahmad-Taylor:I took the PRD directly and dropped it into Lovable, which is a vibes coding platform from lovable.dev based out of Stockholm. And that generated a functional prototype in about ten minutes. So we're now at the forty five minute mark. It took me about another ten minutes to drop in a Spotify widget so that you could listen to some Frankie Knuckles music. And in so doing, I was able to visualize the artist's intent for what mister Gates was intending and looking forward to seeing.
Ty Ahmad-Taylor:It's not fully functional. It is a prototype, but it gave him a sense of the art of the possible. And that effort really sort of opened my eyes. You know, doing something under an hour that used to take one to three months is really an amplification of productivity that we have not seen in my lifetime.
Rob Kelly:What did Thyaster think?
Ty Ahmad-Taylor:He liked it quite a bit, and he accused me of engaging in some kind of alchemy. But I think it was a compliment.
Rob Kelly:Tell me about Norguard.
Ty Ahmad-Taylor:Norguard was an effort that I had to basically learn more about AI tools. I basically in 2015, I wrote a short story. It's about a search and rescue team that's based above the Arctic Circle. That because they're above the Arctic Circle, they can essentially conduct search and rescue explorations across a bunch of northern situated countries around the world simply by flying over the North Pole.
Rob Kelly:And you wrote this story?
Ty Ahmad-Taylor:I wrote this story. It ends with this note. That's when everything changed. And it's essentially they tripped over a body and it turns into a a vampire sort of a mystery story. Cause the vampires like the fact that the North Pole is very dark for half of the year.
Ty Ahmad-Taylor:And so they take advantage of that. The only thing is that it's slim pickings for vampire fodder. This is pre sinners and pre some of the other things that you've seen in the in the cinema. But I I wrote the short story. It's unpublished.
Ty Ahmad-Taylor:A couple of friends have read it. I did nothing with it. But what I did was is I took that roughly six to seven pages of a short story. And I went and I used a product that's called storyborder dot ai. What story border dot ai does is it helps you take any written material and it automatically took my story and generated a multi page screenplay with the characters and scene instructions, the things that you would look for in a screenplay.
Ty Ahmad-Taylor:And my mind was sort of blown about the auto generation of the screenplay. It then extracts the characters from the screenplay and allows you to edit the characters as you see here to maintain visual consistency for each one of the characters across different scenes. One of the many opportunities that exist within AI is memory, which is having consistency from moment to moment and remembering previous chats and chat history. Because not only does it have hallucinations, but it's got a Dory the Goldfish problem where it also forgets what you've programmed it with. So for each one of the characters, you can edit the description based on either the description reference or the image reference.
Ty Ahmad-Taylor:And in this case, it's a middle aged African American gentleman with locks wearing a thick parka and gloves. You can then save that so that you have character consistency across the screenplay. From that, storyboarder.ai automatically generates a shot list, which includes a shot description, the time of the shot, the shot size, the perspective, the movement, the equipment, the focal length, and the aspect ratio. A friend of mine named Linden Barwad identified for me the fact that any film is nothing but twenty five hundred five second segments that are strung together. It's two thousand to three thousand five second moments that are strung together.
Ty Ahmad-Taylor:This is not something that I knew intuitively, but once you see the shot list, you get an understanding of what actually can be done. And from the shot list, again, this is all via AI, you can then generate a storyboard. You can change the art style of the storyboard, but the storyboard is based on my screenplay upon my characters and the shot list that I just told you about. And with that, you can edit each one of the shot lists as you see fit. And you can edit the description, the text around it.
Ty Ahmad-Taylor:And so these are all AI generated images. But we're not done. From that, you can then generate an animatic, which is also known as a pre visualization of what actually transpired during the course of the film. So I did all of this in under an hour, copied the story, generated the screenplay, synthesized the characters, generated the shot list, generated the animatics. And what you have is a a visual overview of the storytelling that I hope to engage in.
Ty Ahmad-Taylor:So with this, I can then export this overlay an audio track, you know, sort of a visual mood board. But it brings to visualization the actual work that I was thinking of. And that for me is really a masterstroke. And what it's gonna allow for is it's gonna allow for a lot of naive creators to come up with films of their own. And that output is something that I'm really looking forward to seeing.
Rob Kelly:Had you written a screenplay before or written a story?
Ty Ahmad-Taylor:No. Creative writing, yes, I've done over the course of my life, but I've never written a screenplay or a film or anything of that sort before.
Rob Kelly:So you had written things like Norgard before for yourself, and did it just stop at the point of being a blog on Medium or, you know, in your Word doc or Google Docs?
Ty Ahmad-Taylor:I never even posted my short stories. Like, the last time I wrote creatively at length was in, you know, was when I was at college and university. And so I spent time as a visual reporter for the New York Times at the beginning of my career. And so that was a lot of storytelling, but it was storytelling about real events. And so creative expression, creative storytelling, I just had some time and decided to write this short story.
Rob Kelly:You're on the board at SFMOMA. What are they talking about AI wise over there?
Ty Ahmad-Taylor:Nothing so far. I mean, you know, there's not a lot of AI generated art. I think it would be dismissed by and large. I'm a broad art consumer. Like, I try and see art every day if I can.
Ty Ahmad-Taylor:It doesn't happen every day. I'm in New York this week, so I walked over to the Guggenheim right after lunch and saw Rashid Johnson and Faith Ringgold show just today and remain in awe of their sort of like broad talent. I don't agree with all of their art or like all of their art, but like I I love the exposure to We've not had any large conversations about it, so I don't have any smart insights there, unfortunately. I wish I did.
Rob Kelly:Don't you think it's just a matter of time given it's San Francisco that SFMOMA will have a wing just for AI generated art?
Ty Ahmad-Taylor:Oh, I don't I I would not preclude it. I would not preclude it. I definitely see that that as an opportunity. Just three months ago, I I was trying to generate a placard and I wanted it to mimic a living artist. And the LLM knew better.
Ty Ahmad-Taylor:And it was like, I cannot do that. You're asking me to engage in copyright violations. I was like, oh, okay. I don't know. I I rubbed up right on the edge of, like, what's permissible, and the LLM knew better than to generate what I was looking for.
Ty Ahmad-Taylor:I said, basically, generate this type of scene in the style of in this case, there's an artist that I greatly admired named Kinsura Davis because I was gonna use them as a placard for for a dinner party, like little name tags. And it was like, nope. Can't do it. And I was like, oh, okay. Yeah.
Ty Ahmad-Taylor:I guess that makes sense. That makes sense. I had thought about it.
Rob Kelly:It's a great topic. How much of your Norgard story, let's say it goes into full on screenplay, maybe even a short film, how much of that is copyrightable?
Ty Ahmad-Taylor:I wrote the original story. I don't have a copyright on that. And then going through the LLM process, to your point, I probably could generate a short film based upon it. Every character in the film would be synthetic. The scenes would be synthetic.
Ty Ahmad-Taylor:The camera motions would be synthetic using the process that I've been using. That's a question for lawyers. I don't know. I I didn't pay any actors. I'm just using compute horsepower to generate these images.
Ty Ahmad-Taylor:I think the larger question is like, is there liability because it's based on some preexisting image that it's using as source material that I'm somehow unaware of?
Rob Kelly:Would you describe yourself as sort of semi retired? No. No. No. Not at all?
Rob Kelly:Just not No. Are you active in management besides being on boards?
Ty Ahmad-Taylor:No. I'm consulting with a lot of companies, three companies specifically. And I'm also on the GoPro board, and I'm relentlessly dedicated to advancing the careers of people who have less experience than I do. And so I spent a lot of time doing that.
Rob Kelly:Maybe I'll ask this a different way. The the projects with Frankie Knuckles and with Norgard seem like passion projects not for money.
Ty Ahmad-Taylor:Correct.
Rob Kelly:Is that fair to say?
Ty Ahmad-Taylor:Yes.
Rob Kelly:I was just wondering if that's are we in this new era where when maybe you're doing more sitting on boards and consulting, but less full time at Meta or Snap or THX or New York Times. In this new world, are people gonna be spending more of their time on these AI passion projects like the two you shared?
Ty Ahmad-Taylor:I think there are two ways of thinking about it. So one is that people are gonna get a lot more productive, like a 100 x more productive. Not five x, not six x, but a 100 x more productive without running them into the ground hopefully. And that's my distinct hope and desire. In that vein, there is no more leisure time.
Ty Ahmad-Taylor:And then there's another view that says that if people become 10 x more productive, not even a 100 x more productive, does that give them more leisure time? Do you just issue forth the same amount of products? I don't know who's right, but what I do know is that there's an opportunity for greater experimentation. And for me, I wanna remain au courant with the tools that are out there that are being used so that I can speak to them effectively, so that I can counsel others about how they should approach things. And so it's important for me to to be hands on with those tools so I know how to use them, what they do, the time savings that they incur, and that I can then write that out or or share that out and then, you know, rent and distribute so that other people are also amplified.
Ty Ahmad-Taylor:It is my distinct hope that part of the human condition is is helping others be better at what they do and also be better at enjoying life and understanding life. And so that's sort of my goal with the things that I've been doing.
Rob Kelly:And a couple questions on maybe a little more humanitarian type stuff. How are you guiding your kids about AI? What are you telling them?
Ty Ahmad-Taylor:Man, it's been a it's been a road of mistakes. Like five years ago, it's like you have to major in computer science. And I think I was wrong about that. And then two years ago I was like, you both need to take SQL pull courses so you know how do SQL pulls. And I was wrong about that.
Ty Ahmad-Taylor:So I've got a child who's studying math and physics and the opportunity for that person is, you know, potentially going to work at a hedge fund maybe, but also going to go work in AI. And so it's my view that anybody who can communicate well and understands human needs is gonna flourish in the in the new economy that we're that we're transitioning into. And those who are not multidisciplinary are gonna find it a little bit of a harder slog. And so that's the counsel that I've been giving them.
Rob Kelly:So you see an edge in being multidisciplinary and not singularly focused sort of ten thousand hour rule?
Ty Ahmad-Taylor:The ability to understand human needs and to communicate well will serve anybody in any work capacity is my view. You could be a product leader, you could work in service, but those core human traits I think are gonna serve Gen Z and Gen Alpha for the rest of their careers and their lives. It also helps just understanding people and being empathetic with people. I think that that's desperately needed in a skill set that's not ever gonna go out of style. Empathy.
Rob Kelly:Are you gonna create an avatar of yourself for your family?
Ty Ahmad-Taylor:No. I find it creepy and weird. But other people have done it, and I I don't look down my nose at them. I think that scaling one's impact through a digital clone is super valuable, but the edge cases where stuff goes awry or the things that concern me.
Rob Kelly:What about for when you're gone?
Ty Ahmad-Taylor:For when I'm gone.
Rob Kelly:When you're no longer with a beating heart.
Ty Ahmad-Taylor:Yeah. I don't have the ego bounce up about being larger than the footprint that I have. The people who know me hopefully view me in a positive light, and I believe my my children are great people on the planet, and and that that's enough for me.
Rob Kelly:And even just for them to talk to you when you're gone. Hey, dad. Got a problem. How would you help me on this? Less ego in that case.
Rob Kelly:Do you know what I mean?
Ty Ahmad-Taylor:Yeah. Yeah. I I haven't thought about that deeply enough to consider. I would love for them to be able to find solace and comfort, but I also trust their decision making at their relatively young ages. They're both adults.
Ty Ahmad-Taylor:I don't have my father anymore, and I miss him, to be clear. But there was a with nothing but love and respect, I think there was a point where I outgrew his advice. I would want the same for my children.
Rob Kelly:If you could have an avatar of your father, think you're okay with without it?
Ty Ahmad-Taylor:What I'm left with with my father is nothing but fond memories. And in so doing, I believe that I elevate his spirit and his character to a place where he's held in high esteem, And that is the most important thing that I can bring to bear with regards to where he is now as opposed to when he was on the planet with us.
Rob Kelly:Alright, Ty. Anything that I did not ask you that I should have?
Ty Ahmad-Taylor:Just I commented that human beings don't like change. What I've historically done is I've been a futurist. And sometimes being a futurist and thinking about what the future holds, I make mistakes. But the thing that I would leave everybody with is that the change system that we're in now with this transition to AI and potentially to a general AI, which is self intelligent and aware, it's gonna require a lot of change. Change in the work that everybody does, change in how we interact with our peers and colleagues.
Ty Ahmad-Taylor:But with that change is going to come a lot of opportunity. And for those who are willing to change their personal workflow and to adopt the tools and to experiment with them and make mistakes and learn from the mistakes, I see nothing but a super bright future. That's the seed I hopefully have planted with my children. But with adults who are in the workforce, I would say the same opportunities exist.
Rob Kelly:Thank you, Tai.
Ty Ahmad-Taylor:It's my pleasure. Thank you for having me.
Rob Kelly:Well, this is Media and the Machine. A few things about you and me. If you wanna hear about the next new episode, make sure you hit follow on the show in your podcast app. If you wanna go a little deeper, head to mediaandthemachine.com and subscribe. When you share your email with me, you can see handcrafted transcripts, read the essays in my newsletter, and be the first to hear about who the guest is on the next show.
Rob Kelly:You can also email me directly from there, maybe you wanna recommend a guest. I'll give you a shout out if you do. I love paying it forward. From time to time, I also open up office hours and host small meetups for subscribers, just to meet, talk, and build things together. If you're creating something of your own or thinking about it, I'd love to help.
Rob Kelly:Maybe you've got a podcast in you. Finally, I don't have a marketing budget for this show. So if it's finding you and others, it's because someone like you passed it along. I'm genuinely grateful. If you have a moment, an honest rating helps me make this better for you.
Rob Kelly:You can just go to the show page and click one of the stars. And I'd rather you give me a low rating than no rating at all. I mean it. It pushes me to get better. Thanks again, and see you next time.