Russ Mashmeyer's journey from fine art student to Meta's AI Product Design Director reveals the unexpected connections between creativity and code. Growing up in suburban Georgia where "everything was new," Russ found himself drawn to New York's layers of history, studying fine art at NYU before joining a touring indie band called The XYZ Affair. What began as building Flash websites for his band evolved into a career defining how we interact with technology.~

 Now leading Meta's Pathfinding team, he explores rapid prototyping of AI-powered products, helping designers and creators understand AI not as a replacement for human creativity but as an accelerant for bridging skill gaps.

Russ brings a unique perspective to AI development, viewing it through the lens of cultural history—from photography's disruption of fine art to desktop publishing's transformation of graphic design. He argues that AI models deliver "mediocre, average expected results" by design, making human taste, perspective, and cultural awareness more essential than ever. 

Key Takeaways

• Software as artistic medium requires the same creative intuition as traditional fine art, just expressed through different tools and faster feedback loops
• The best preparation for emerging fields is developing insatiable curiosity about how things work rather than mastering specific technical skills
• Cultural and historical context matters more than ever in AI development—understanding what resonated with people before helps predict what will matter next
• AI models are designed to produce average results, making human taste, lived perspective, and cultural relevance the differentiating factors in creative work
• "Naïve optimism"—the mindset of "how hard could it be?"—is essential for innovation, especially in spaces where no one is an expert yet
• Children benefit enormously from watching adults play, fail with smiles, and create together, normalizing creativity as a lifelong practice
• Every major technological shift creates cultural upheaval, but society consistently figures out how to metabolize new tools and elevate them to art forms
• The most powerful use of AI raises the floor of competence in areas where you have skill gaps while you remain exceptional in your core strengths
• Treating your first band like a startup teaches essential lessons about entrepreneurialism, feedback loops, and creating products people love
• New technology succeeds when it serves human intent rather than replacing human creativity, becoming a tool that unlocks what people inherently want to do
• Building in public fields where no one is an expert yet levels the playing field—depth of historical knowledge matters more than technical seniority
• The transition from planning perfect UIs in Figma to sculpting functional prototypes with AI represents a fundamental shift from drawing to three-dimensional creation
• Emergent capabilities in AI models—features that weren't explicitly designed—mirror how creative misuse of instruments led to entirely new musical genres

Daring Forever. Podcast with Radim Malinic

daringcreativity.com |  desk@daringcreativity.com

Books by Radim Malinic Paperback and Kindle > https://amzn.to/4biTwFc
Free audiobook (with Audible trial) > https://geni.us/free-audiobook
Book bundles  https://novemberuniverse.co.uk

Lux Coffee Co. https://luxcoffee.co.uk/  (Use: PODCAST for 15% off)
November Universe https://novemberuniverse.co.uk (Use: PODCAST for 10% off)

[00:00:00] 

Russell: I think people get really worried by the idea that AI can replace human cognitive function and creativity. But I actually think that that's not fundamentally what AI models do. AI models compute and deliver the mean the average of a particular request or topic.

And they trained to do that. That's, the goal of the model, is to deliver mediocre, average expected results. And that's just not what pushes culture forward, right? But if you have a particular area where you go really, like you're an incredible painter, but you're really bad at writing your artist abstract, well, gosh, you've got a great tool to help you get your words onto the page in a way that can make sense to other people supporting your intent, driving your painting work forward, but supported by ai, not [00:01:00] replaced by.

Welcome to the Daring Creativity Podcast, a show about daring to forever explore creativity that isn't about chasing shiny perfection. It's about showing up with all your doubts and imperfections and making them count. It's about becoming more of who you already are. My name is Radim Malinic I'm a designer, author, and eternally curious human being.

 I am talking to a broad range of guests who share their stories of small actions that sparked lifetime discoveries, taking one step towards the thing that made them feel most alive. 

 Let me begin this episode with a question. Are you ready to discover what happens when you dare to create? 

today I'm speaking with [00:02:00] RAs Mashmeyer, a product design director at Meta who bridges fine art, music, and artificial intelligence while maintaining an unwavering curiosity about how things work and what becomes possible next.

From working on early Facebook search technology to now being in Meta's current AI team, he works on products used by billions. He's always driven by the belief that software is an artistic medium, and technology should accelerate human creativity rather than replace.

In this conversation, Russ discusses the challenges of maintaining creative naivety alongside imposter syndrome, and we talk at depth about the reality of AI models and why human taste is more essential than ever. It's my pleasure to share with you my conversation with Russell Meyer. 

Radim: Hey, Ru, how you doing today? It's great to have you on the show.

Russell: Super good to be here. Really excited.

Radim: For those who may have never [00:03:00] heard of Russ Meyer how would you introduce yourself?

Russell: Well, I'm a product design director. At Meta. I'm working specifically kinda in the AI space. /

And just really exploring what's emerging as possibilities because AI is constantly expanding what it's capable of, and that unlocks like a lot of new possibilities, I think, for the kinds of software we can build. And you've had a ton of creative people on and and they make, different kinds of media, whether that's written media or visual media.

 I think of myself really more as an artist working in the software medium.

Radim: I like that answer because I mean, have a look through, you know, your breadcrumbs that you've got online and things that you do, and sort of some of the things that you promoted, you've created and. The, the artistry through code has been pretty much main, path of your life so far, right?

Russell: Yeah, it's funny. I mean, I, I started my adult productive life in college studying fine art. And so I [00:04:00] was deep into drawing, like very realistic drawing and painting installation art, performance art. Just did a lot of ridiculous, hilarious projects and interesting, you know, investigations as a, as an art student at NYU.

 And then when I left I actually found myself like not very excited about kind of the gallery world of fine art, and that was really the immediate direct path out of the undergrad program that I had been a part of in New York. And, you know, obviously there's no better place in the world, I think to study art.

 Than New York. Just because the density of, you know, networks and galleries and artists and incredible history there. And so that's really what I, what I went for. And I, and I really expected to come out of that program to become a fine artist. But instead along the way I actually kind of fell outta love with just the way art reaches the world and kind of how the world sees and interacts with art in, you know, lots of exclusive white rooms that, you know, you get side eye at for just [00:05:00] occupying, you know, if you look like you can't afford what's in there.

And I think that kind of engagement or that kind of, that mode of creating art and bringing it to the world just didn't appeal to me as much. And so when I left art school, I was kind of like trying to figure out what to do next. And I, I stumbled my way into becoming a member of a band. There was a, a band at NYU that was really popular as we were going through, the years of school.

And and I, I was super excited when, you know, in senior year they, you know, they were losing their guitarist and they came to a house party at, my apartment. They saw I had guitars and a keyboard and played music. And they were like, we're losing a member. Do you wanna join? And I was like, absolutely.

That's like a dream come true. So I, you know, I spent the next four or five years, you know, as a member of this band called the x, y, Z affair in New York. You know, we were kind of part of the Brooklyn indie scene and the early aughts not very successful, but it was a lot of fun and, and really what it.

Opened up for me is, is working in digital media, building websites for my band and really turning that into kind of [00:06:00] an art form of itself. This was the age of flash. So I was just going ham as an art student in flash and like of picking up action script and sort of learning some of that world of development, but mixed with really rich art and animation.

 that became kind of my playground was like our band's website and that led to doing other band's websites and then recording studio websites and then small brands and labels and other things. And really branched out from there to do, you know, kind of a full suite of design services.

 And so it sort of leveraged my early, like fine art skills to sort of enter into this world of design and, and software and, and the web, which was really exciting at the time.

Radim: I love that story. I absolutely love it because it's the power of music that kind of leads you onto the next becausewhat I wanna know about your to five years as a musician, Did you guys play gigs? Did you, I mean, as you 

Russell: Yeah, 

Radim: was, fairly famous,

Russell: no, I mean, we, we recorded a, full length album, a couple of eps. [00:07:00] We toured nationally. We unfortunately were not good enough or popular enough to get a tour manager which is what you, typically like work with in order to set up these national tours. But we, you know, we worked with other bands that we were friends with.

We made contacts in other cities and we set up our own kind of like national tour and, and hit South by Southwest and did a bunch of other, like really, really fun shows along the way. And it was incredible oh my God, like how many people get a chance to like, stuff themselves into, a conversion van with three of their best friends and just drive around the country, like goofing off and playing shows and trying to make something, kind of happen in the creative space.

It was, it was really exhilarating.

Radim: what was your life of music like? What was the thing that made you pick up a guitar? Was there any particular band, any particular song? What was that?

Russell: I've always really loved sort of more confessional, quiet artists, folks like Elliot Smith. I mean, I love Jeff Tweedy and Wilco. It's just like his voice has just been like a, a lifelong sort of thread in my life of just different [00:08:00] albums and different eras. uncle Tupelo and course Wilco and his, more recent solo stuff with Tweety and you know, his kids, which is man, what a lifelong dream to be able to like, be a musician, but then, you know, make music with your kids. I'm in my studio space today which is, you know, not only a, a space for me to build stuff and get creative, but it's, also kind of like a rehearsal and music space.

 you know, my son is, is learning drums and so we're, we're getting pretty excited about just rehearsing and playing and just having fun together with music as a family. I think it'd be really amazing. So some of that is drafting off of, I think Jeff Tweed's vibes and Aura there.

Radim: Oh, that's amazing because I had your wife on a podcast a while ago and she was telling me like she was learning to play drums 

Russell: And it's amazing because I think having our kids to see that flow of, of music and creativity from us is a gift because, I've spoken to so many people and some people have totally non-creative parents and, they're creative now, but.

Radim: [00:09:00] I think there's a little advantage when you just show them, real time that sort of opportunity of what they can do.

 just, just having fun and just appreciating that you never know when 

Russell: Yeah, 

Radim: pocket of creativity.

Right.

Russell:  I think kids really need to see adults play, right? They need to see that continuity, that that behavior that they just engage so naturally in as kids is okay in adulthood because so much of adulthood is about sort of taking care of priorities and getting down to business and and getting stuff done.

 But opening up that window to let your kids see you flub a line or fail, but do it with a smile and having fun and like doing something together in synchrony with other people that creates moments of, beauty and wonder and awesomeness. Man I wish I grew up a little bit more in a, in a space.

I mean, I was very lucky. My parents were very supportive of my, you know, my, my desire to, Study art and explore art and music. But but they themselves [00:10:00] were not, not very creative people. 

Radim: Where did you grow  

Russell: I grew up in Georgia so the American South. I grew up in a little suburb outside of Atlanta called Alpharetta. It is now, I actually haven't been there in 20 years. My folks moved to Florida in the mid ts and so I actually haven't been back to my hometown, but my friends tell me that it's just absolutely transformed as a place.

It used to be mostly like horse pastures and you know, when my family first moved there, I was born in Tennessee, very much born and bred in the American South. I went to NYU obviously when I was 18 for college, but until then, you know, grew up entirely in Tennessee and Georgia. and yeah, no, apparently my hometown is now unrecognizable.

It's just like luxury condos in the downtown and just every like fancy brand and store you can imagine. And it used to just be like a little greasy spoon diner and like an old bank and like the 75-year-old high school and [00:11:00] that was the whole downtown when I grew up there. So I'm looking forward to actually bringing Jess someday 'cause she's never been there either.

Radim: I mean, would you moan the demise of the old greasy spoon and a small bank? 

Russell: It'll just be a totally different place. I think it'll just be more of a curiosity visit than a, a desire necessarily to connect with. I think what used to be there, which I think is, you know, most likely long gone, but I've still got a number of friends, you know, from high school and and other things down there who are still in the old, you know, sort of neighborhoods I used to live in and sending their kids to the same school by name.

It's a new building, new place, but same name. And so I'll, I'll look forward to catching up with them when I head there.

Radim: so what was your jump like service? Do you go from. Quieter place into, into New York to study and change of scenery. was that effect on you like? how did that inspire you?

Russell: You know, I mentioned a little bit about my hometown. You know how when we first moved there as a family, it was mostly like horse pastures. It was on a crazy growth [00:12:00] trajectory as a population center, as an excerpt of Atlanta. It just became kind of the place or one of the. Two or three places that if you were an upwardly mobile person moving to Atlanta, which was exploding as a commercial and business scene.

 During, you know, the time I was growing up there were just a handful of places you would really move if you didn't wanna live in the city itself. And, and Alpharetta was one of those places, so it just grew insanely fast. I think we were, you know, our house was in one of the first large planned communities in Alpharetta, but then, you know, over the next, whatever, 18 years or whatever, 16 years that I lived there, probably 40 more, sprung up and there was like this whole mall.

And, but the, my point is the whole town was basically built during my lifetime. And it really didn't exist before then. And so my early childhood experience was spent in this place where everything was new. All the time, and that's really exciting and really fun. It's oh my gosh, there's this new mall, there's this new store, there's this new [00:13:00] restaurant.

But they're all chains. They were all the same places you could kind of like get everywhere else in the country. There was really no unique commercial history, business history, industrial history. And when I went to New York and toured NYU for the first time just to see what it might be like as I was touring, touring potential schools, the thing that struck me when I just walked into the streets of New York for the first time was how old everything was.

 And that was fascinating to me. Like I loved being in a place with so many layers of history not just in the creative sphere, but in every sphere. The buildings were old, the culture was old, the networks were old. The just like everything about New York. So much older than anything I'd experienced. And, and you know, as somebody growing up in the uk, Europe you know, you're kind of constantly surrounded by history.

 And that was really what I was lacking growing up [00:14:00] in Georgia. It was any, any real history to the place that I grew up in. And, and so that was what really enchanted me and sort of drew me to New York was trying to be a part of and learn from, you know, kind of that history and and just really enjoy it.

Live in that, persistence of that history.

Radim: It's interesting what you said about New York and what you said potentially about and for example, my experience of London. It's just once you start looking up beyond the shop windows and start seeing all the details and you know, when I said do you know, do you mourn about, the demise of the quiet pastures and you know, the horse trails because need to preserve some of those details because the modern architecture, 'cause it goes up so quickly and detail is not there. 

Sometimes people get it right, sometimes they don't. And you kind of thinking like, how much time did they spend on this building of this magnitude of that? Layers of that detailing.

' cause that's what we appreciate now. That's what, that's why people travel from different countries. Not saying Hey, you've got new highrises. Let's see those. No [00:15:00] one says that. They go Hey, I really wanna see that detail of that past. And when you start digging into the stories and as you said, the layers of history, it's fascinating.

Utterly, utterly fascinating. So, yeah. I love when you said, you know, it surprised me everything was old or coming from place. ' cause the quote you said where you were growing up that everything is new. When you think about what you do now through your love of action scripting, 

making band websites, you know, being into the code, where does it take you on a commercial path of a, working designer coder, and now a director of product?

So I, I had this kind of grounding in fine art. And then, you know, I, I, I sort of consider my band my first startup in, in a lot of ways, right? I mean, that's really what it is. You're trying to get an entrepreneurial effort that creates products that people love off the ground.

Russell: And so there's just a ton to learn through that [00:16:00] experience and that entrepreneurialism. And then, you know, somewhere along that path obviously I was, you know, kind of continuing to do. Website design, brand design, you know, this kind of work for a string of clients, both in small agencies and, you know, on my own sort of sharpening my skills, transitioned out of the era of flash into the era of web standards and WordPress sites and skinning WordPress sites and sort of building out this online infrastructure for small brands and businesses, you know, whether that's just like a marketing site or whether it's a small product they offer to their customers.

 Just doing really simple stuff like that in, in the early web when like applications were, were really starting to come to life for the first time in, in Web 2.0. and along the way there was Jess actually told me about a new program grad program at the School of Visual Arts in New York where we met and lived together.

I was in the band at the time but we were four or five years in, we had released an album, a couple of [00:17:00] eps. Like it wasn't really gaining momentum, I would say. when she shared you know, this program and interaction design at SDA, I went down the course list in the class list and I was just like, I am.

So excited to learn every single one of these things. It was just like right place, right time kind of thing. And, and the program hadn't existed the year before. This was like the first class. They were just spinning it up. Liz Danko was the chair of the program. She's incredible. I went and I met with her and talked with her about the program, but I was just so excited to learn about things like, just like really to level up my nerdery, learn about things like cybernetics strategic entrepreneurialism actual like development, software development but also lightweight UX r obviously, you know, core design principles, content strat.

Just like everything that kind of is core to what we do today within product design. That's really what the interaction design program was. Strategic [00:18:00] innovation. Just so many things that I was just so excited to get my hands dirty with after many years of mostly making like brochure wear for brands, I really wanted to make products.

Like I wanted to make things that people could use in their everyday lives that would. Unlock something new for them. And that's really what I've been kind of trying to do ever since is like getting into that space of new possibility and seeing, hey, this just came out last week, this new framework, this new API, this new model, this new AI model.

Like what now? Does that make possible, that wasn't possible last week.

Radim: That's amazing because as you were sort of reeling off the modules and the features, there'd be lots of regular listeners looking around the room. What do you mean lightweight UX r what, That is like literally you speak in different language to most of people, but where was the passion?

 when you got introduced to this course, which obviously didn't exist even a year before you knew what these [00:19:00] things were. And I wanna sort of find out what was that sort of guiding light inside you I want to know more about this because it's a part of my book, I'm writing Daring Creativity.

I'm thinking about a topic of now versus how, most of the things that we do, we don't necessarily start because we know how to do it. We wanna start now because we've got sort of the, a library of ideas, library of influences, inspiration, information, kind of put it all together and going, you know what?

I wanna start this now. I dunno how to do it yet. ' I knew, I don't think I potentially would be even interested. So in 

Russell: Yeah. 

Radim: that moment when you knew was the sort of hundreds and thousands of different sort of decisions coming together and saying, that's my calling.

Russell: That's a great question. I think it was a number of threads, you know, since I was a kid, I've been fascinated with the future and sort of like thinking about how to kind of predict the future. Like sort of like futurists do, you know, they're sort of like, they, they sort of keep track and tabs of, okay, well what's [00:20:00] emerging in different fields?

What do the experts think is like kind of the next wave and how do we like, think about what that's going to do in terms of, you know, what it unlocks for people's everyday lives, but also what the downstream implications to society might be and, all the ways we kind of built the world.

And I think the future, right, is like one thing to kind of pin on my like mood board of what drove me here. And then the other is just like a real curiosity. To understand how the world works. For some reason, since I was just a little kid, I have just, I have just needed to understand how everything that exists works.

 How does it get made, how does it work? Where does it come from? Who makes it I certainly don't, don't have comprehensive knowledge of everything. You know, I'm, I'm only 43. I've got a number of years left to go in lots of fields to like, investigate and learn about. But certainly the, the places where I've been really interested is software, computers.

 as a kid growing up in the eighties, [00:21:00] they were just becoming so powerful. Like the personal computer had emerged. I played video games and then the early web and then, but I didn't understand fundamentally how any of it worked. It just seemed to be this like magic. Sort of experience. And it was really important to me to fundamentally understand how humans made this thing I and how it works.

 And so that kind of insatiable curiosity to just mechanically understand the world has, has been kind of another one of those things that I'd put on my mood board as, a key pin for what, what drove me here. And, the program itself seemed like a way to open those doors to really understand how so much of what we use in this world, I mean, certainly around technology not just the software and the screens that we look at and the information convey there, but the hardware itself wrapped around it, the sensors and the networks and the cloud infrastructure that, that makes the experiences that we have possible is just been an endless source of fascination as just like an engineer engineering oriented kind of mind.[00:22:00] 

Radim: If I'm doing my maths right, most of this stuff, of your curiosity satisfaction was actually pre-internet. 

was your way to actually find that information? Because with fine art it's a little bit easy, you know, potentially you can pop into a gallery or buy a book. But some of the stuff that you described that you really wanted to know how it was made, who, who, who made it, and you know, where did it come from? That's a little bit hidden. not, that's not in a plain view.

Russell: Yeah, it's interesting. I mean, I think I don't know how much it came from the environment or just intrinsically I've always loved, there's there's a story from when I was a kid and my, my older brother had won a sort of early handheld video game console. It was called the Lynx.

It was made by Atari. You know, he brought it home and had some California sports game on it and a couple of other like games. You know, he played it and, and enjoyed it for a while, but you as a [00:23:00] kid, he had a bit of a temper and so he ended up like throwing the handheld game set, against a wall or just throwing it somewhere and it, you know, hid and kind of didn't work anymore after that.

And I was like, oh. And I was, I was all of seven at the time, maybe eight. And I was like, I don't know, maybe I can fix it. And so I like literally grabbed a screwdriver and opened up the back of this, microprocessor driven video game console as like a 7-year-old. And I'm like, oh, I can figure this out.

 and just looked at it, didn't understand anything about what I was looking. It was a bunch of capacitors and resist you know, I now have a sense of electronics and how those things work, but back then it was all Greek to me. but I like, you know, I saw a little capacitor that looked kind of folded over and I was like, oh, maybe if I like pushed that back up vertical, it'll like work again.

And so I like just fussed around a little bit, closed the backup, and I swear to God, the thing turned on immediately. And I think for me, even though I probably did literally nothing to fix that links video game system that experience [00:24:00] of opening up something very complex that was broken to tinkering around closing it and having it work was like incredible.

Like that fast feedback loop. And I think that's really what drove me to software as well, is just the speed of the feedback. Loop of creation is so fast. Like you make a couple changes, you look at it, you play with it, you can experience it immediately. And so few other art forms, barring, maybe photography, you know, kind of have that level of immediacy and that, immediate feedback loop.

Radim: I mean, you're talking about feedback loops and yeah, I think that dose of dopamine of being, fixer of consoles must be quite amazing at seven. But feedback loop also takes me back to music. Because I don't think anyone I've ever known would say that my band was my first startup.

when I think about it creatively, and when I, when you talk about know really fast feedback loops, you know, in music, if something works and if something doesn't, because again, that hit of dopamine when you start writing a song with your friends, [00:25:00] that's unrivaled, absolute mega load of dopamine.

' cause this is the best thing in the world. And then you play it through the room of people and they're like, what? And you're like, I've lost all the dopamine now. It was literally like, where did it all go? Because after this was great. So where do we go from, from SVA, from you completing the course and to, you know, taking your all 

Russell: Yeah. 

Radim: these days?

Russell: Yeah. You know, I spent two years in that grad program as, part of the premier class, the first class of the interaction design program at SVA and made incredible friends, like just every class ended up being such a banger, and it's such a different experience. You know, I was very excited about my undergraduate program and everything I would learn and do there, but it was so different to go out into the world and work and try to build something on my own as an adult, realize where my gaps are, what things I like, what things I don't like, and then to come back to school, [00:26:00] to an environment where it's just about learning and trying things with no fear of failure.

And there was something so special about being back in that environment, coming back from the world of, professional life. And I, I really dove in just headfirst you know, worked startup level, crazy hours because I was really passionate about the ideas I had and the concepts I was trying to build and the projects I was doing.

 Then after those two years chatted with a few companies small startups agencies, places like ideo, et cetera. And ended up talking to folks at Facebook, which was still relatively early. This was like 2009. Not early, early, early, but still relatively early. when I went and interviewed there I think the team, the design team was still 25, but that was like large to them.

That was, the largest they'd ever been at that point. And met people who just blew my mind with. What they were [00:27:00] thinking about the complexities and scale of the kinds of software challenges. They were trying to like, make progress against their openness to feedback and questions and trying to figure out what's the best idea we can get to together.

And that like core ethos that was really part of the core of Facebook in the early days, just was so immediately apparent in my conversations with the early Facebook team when I interviewed and man, I just got so excited about the idea of working with those really curious, really earnest really humble people on really big hard problems in a space where I felt like I could apply my skills as a designer, as a software.

Thinker, developer person, thinking about the future and how we can make the future like really exciting and better and, and incredible for all of us. And I went to work for Facebook in 2009 and started as a product designer there. I started on groups, [00:28:00] Facebook groups. We were just turning that into a mini Facebook inside your college campus.

So if you had a nyu.edu email address, you could be part of the parent NYU group and only meet and find other students at nyu. So it was sort of like seeding little college-based. Facebooks inside of the big, now mega sort of everyone is invited Facebook. That was incredibly successful.

Like we've managed to like seed and grow these incredible communities across pretty much every college campus in the country using Facebook groups. And then I rolled on to working on Facebook search, which was going through a huge redesign at the time. This was one of my first opportunities to really like, try and push on the state of the art for what an experience could be, you know, at the time, obviously, like Google was.

 the top name in web search, we were trying to unlock a completely different corpus, right? The content you and your friends share on Facebook, certainly only the things that are actually shared with you, not everything. And you know, your friends and you know, their [00:29:00] experiences and their likes and dislikes and sort of, try and extract some try and help people sort of like find value in the networks that they had built and the things that they and their friends had shared, places they checked in.

 And try and build something really powerful that you could query just like you do today with natural language to say, Hey, show me. Places my friends have been in New York or like the best coffee shops my friends have been to in New York City. And it's boom, we'll just go and we'll look at all the places that your friends have shared that they've been in New York City.

 We'll look at your profile and sort of rank those against, you know, our understanding of you and kind of give you a really great answer that's actually connected to your social network and the things that are relevant to you in your life. And that was a really interesting project, but unfortunately we were just too early.

 The, AI models, like large language models were not certainly not what they are today. We built a, a, really impressive, but. A little brittle and fussy natural language engine [00:30:00] that would parse anything you typed in, into a set of like logical components and then try and answer those logical components kind of one at a time.

But it was not as flexible and as robust as the LLMs that we have today. That can just take literally any text you put in, understand it and give you back like super relevant results. So we were a little early on that, but it was a lot of fun. I obviously learned a lot about the space that I'm now of course applying now that those models exist.

 And, and we can leverage those. And you know, a lot of my like early excitement about how to take what I had learned in grad school and apply it to pushing the state-of-the-art forward, kind of really started with graph search and thinking about natural language search. then you know, I worked on a, a number of other things and then was there for about six and a half years.

 I actually didn't expect to be at Facebook that long. And so after about six and a half years I was like, wasn't I only supposed to be here for two years, learn a lot and then go do a startup? And maybe I should get about that. And so after about six and a half years in 2008 I [00:31:00] left a new adventure.

Radim: before we talk about your next I wanna still stay in 2009 for a short while, because what you describing that's quite a move, that's quite a jump. again, as I told you at the top of, top of our conversation, I, there's a lot of creatives, artists, musicians, and there's a lot of people from that sort of sphere that I talk to who, would openly say, I've had an imposter syndrome, or I didn't feel like I was enough.

 needed to define my enough. maybe I was in too big of a room too early. You know, I, need to know, learn more about myself. I was going through stuff and you were like, Hey, I just arrived on Facebook and these people were crazy clever. And I'm like, how is Russ doing?

 how do you feel in this? Because these people are magical. But how did you feel yourself in the set? Did you feel like, Hey, I'm ready, and any sort of insecurities were going, aside or you do that? 

 I still very vividly recall there was so much happening in my life [00:32:00] at, at this moment, kind of exiting grad school interviewing at Facebook, which at the time. It was not a sure thing. Like all of my teachers were like, why are you like, do they have a design team?

Russell: What are you, what? and I had just asked Jess to marry me. I remember being in the parking lot of Facebook waiting to go into my interview that morning and like finally getting my dad on the phone sitting in the parking lot in my car to just tell him that I had proposed to Jess and that we were getting married.

And I walked in, had an interview. So there was, a lot of wild stuff, there was a lot of excitement in, in my life at this time and sort of changing phase obviously, working at Facebook back then meant moving to California, so it was a big decision for Jess and I to move out to California.

She was not having it at first. Sorry, to circle back to your question earlier, we talked about kind of two things, two drivers for me that, that we sort of pinned, right? One is like the future, the other is understanding how things [00:33:00] work. And I think you're kind of hinting at a couple of others that are, I think, really important.

One is just like a constant desire to learn. I get really sad when I'm not expanding what I know about the world or how things work or what I'm able to build and do and, and achieve in the world, right? So just that constant expansion and feeling challenge is like really important to me.

And so when I had these interviews with folks that I was just like, oh my God, like I, I hope I exit this experience that's smart and thoughtful and humble and insightful as you. And it just felt like a growth space, like a place where I could grow a lot and learn from really smart, really capable people.

And I did. and the other thing that I would put on this mood board things that were pinning drivers for me in my life is what you're kind of hinting at, which is a little bit of this naivete. And, there's two, there's two sides to this. On the one hand, there have been so many situations, potentially, constantly it's in fact it's probably more than norm than not that I feel under [00:34:00] capable of delivering what is being asked of me.

Like that imposter syndrome is real. No matter how far along you are in your career, you're always seeing stuff that other people are doing. You're like, oh my God, that's so good. I'm terrible. I'll never come up with anything that good. But you know, you do, you muddle along, you come up with some cool stuff along the way too.

 but there's a flip side to that, which is that sometimes you just sort of like assume things are gonna be easier than they always end up being. And I think keeping a little bit of that magical thinking of eh, how hard could it be? They did? It is like really healthy to keep with you because honestly, at the end of the day, so many things that are emergent that are just becoming possible, right? That are in the space of futures. No one knows anything about, no one's an expert in what that thing is. We don't understand what that thing is for the most part yet. it's a really interesting space to be both excited about learning new things.

 Feeling a little bit of [00:35:00] imposter syndrome and feeling like always a little bit like you have to prove yourself. But also having a kind of like how hard could this be, like kind of attitude about things I.

Radim: I think there's just the right amount of breadcrumbs that you can see of okay, these people have gone somewhere. They're leaving some clues, what they've created. I'll follow this. because it's interesting 

when you talk about imposter syndrome, just is it even imposter syndrome or is it just sort of that thing of am I just a few steps behind this person because they're already ahead of me? They're not necessarily better, they're not necessarily wiser.

They just done those reps, a few times before and that's what, you know, makes them look confident. But they might be feeling exactly like you because they're trying to prove themselves to somebody else or to themselves. So it just, thinking about it it's just a 

Russell: Yeah. 

Radim: but it's really, I think what's really interesting to this conversation that obviously you work a, slightly [00:36:00] field, yet the feeling and the, the perception is universal.

It just doesn't 

Russell: Yeah. 

Radim: you're not proving yourself with your brush strokes and, you know, choice of canvases and colors. know, the feeling of being enough doesn't matter what you do, because ultimately it's about, you know, how you perceive yourself within your environment.

Russell: I think one thing that has been really interesting to learn as I've explored different fields of creativity, right? Whether that's fine art, visual art music writing technology think you're exactly right. I think the same themes, the same tensions exist in every field.

 and I think my, my perception, my sense is that every field. Has a conversation that is progressing within the culture of that field, right? And it's some mixture of what do we want this to become? What do we, what's possible for it to become you know, whether that's [00:37:00] physics, whether that's painting, whether that's creative writing, whether that's software.

 And obviously it breaks down into all the sub niches of all of those fields. There's all of these individual niche cultural conversations happening. And, and I think one thing that I've discovered is that it's actually quite easy when you drop into a new field because of the information sphere that we built around the internet to come up to speed really quickly on where the conversation is in any field, to understand what has come before and where we are now and what people are looking ahead toward. And I think if you get really good at sort of like. Ingesting all of that context up to today, what's going on in the field today for any given field, almost immediately you'll have jumping off opportunities to say, okay, that's what came before. This is what people are thinking is happening next.

 Here's what I understand is eternal about human need, human [00:38:00] desire, human behavior. What do I think is the most likely outcome of the possibility set ahead of us? And how do I build something into that space that brings people joy, brings people value, helps them do something that's really important to them.

 And that, that core has been really sort of the driving force for me in building software and thinking about the future is what can we unlock next for people that they'll love?

Radim: That's amazing. Be beautifully said. By the way, I have to say thank you for clarifying the situation in the parking lot before Facebook interview, because for a second I'm thinking he didn't ask her to marry her in, in the parking lot just outside Facebook because for a second that's what he sounded like,

but 

Russell: was so much more ornate and complex than that.

Radim: okay. Well I'm glad it was next to your fiesta outside, you know, at headquarters. So you were at Facebook for six and a half years. You went around a few places through, you know, places like Shopify and then you come back and you're back at Meta now [00:39:00] at Meta as a director of product and obviously work on pie finding.

 when I spoke to Jess on this podcast nearly a year ago, she told me about the fact that. When you enter the role, which involves ai the situation for a few weeks was a bit frosty because it's a new situation, it's a developing situation, like it was, we are learning and, and I think what I'm actually displaying here, it was like from some of the previous guests, it was like people were scared about ai.

Like what is it, more than anything. What is it and what is it going to do? Because it's still being very much defined. Because as you and I had a conversation a few months ago when we met, we talked about the fact that there's different AI for different reasons, for different ways, for different products and all different, uses.

 That sort of shake up when Open AI just came out and said Hey, this is Judge GPT, off you go. And everyone goes, holy God. Like for, for the first time someone openly called the AI as a product, because we both know that [00:40:00] machine learning's been around for, you know, a couple of decades, 

 just that change of wording and slight tweak in usage and people go oh my God, what are we gonna do next?

 I believe you told me that you emailed Mark and said, I think we need to get on with ai. I think that was the sentence you told me. So he might be responsible for all

of it. So 

Russell: Well, well, no, no,

 no, no, no. Well, to clarify, I did not, I did not inspire Mark's foray into ai, I promise. No, the, the, the context there and the story there is actually when, when Mark kicked off sort of the, the founding of Fair within Facebook pre meta days, pre meta brand and hired Jan Laun from NYU to come and lead our AI research I got so excited because, gosh, you know, here's this incredible company that I work for.

I'm learning so much from, I'm doing things like natural language search and trying to push on the boundaries of kind of what's possible with technology. [00:41:00] And oh my God, we're, we're founding an AI lab here. Can I please be involved? That was my email to Mark I'm really enjoying search, but gosh, if there's any way or any reason you need design support with fair, please count me in.

 And he was really kind, you know, the company was much smaller back then, and he, he responded to my email, he was like, no, no, no, you're doing great work on search. Please stay there. You know, we won't need design here for quite a while. Which he was right about, you know, I was fundamentally researched for a long, long time until just really the last couple of years.

 But I'm back and I'm working on ai. So it ultimately came full circle.

Radim: So 

the work that you do on large language models and, AI implementation and, meta tools for your team how do you put yourself in a position of a user 

Russell: Mm-hmm. 

Radim: that's what we need to do.

 where can we make a difference and how can we actually make the future?

Russell: Yeah. this actually kind of connects to our earlier conversation about history a little bit, and I think one of the things that has been so important [00:42:00] for me to understand or or to be able to somewhat reasonably have a sense of where things are gonna go or what's gonna hit, what's not gonna hit when you put it out in the market and let people try it is having that sense of cultural history, what things that came before, worked really well, why did those resonate for people?

And, and the more you kind of study the past and study history and study breakthroughs and sort of what was happening culturally around those things. Have that become like the lightning in the bottle that it became you begin to see like a lot of patterns that feel really immutable for people over time.

And I think like coming back to the point around ai, it's like there have been many cycles of this in the history of, in commercial. Activity and industrialization, I mean, particularly in the art realm, like you, you go back to photography and Walter Benjamin and like his essays on just the cultural [00:43:00] uproar over photography as like a fine art medium.

 Or even just as a mechanism to reproduce what artists traditionally had to do a hundred percent of, they had to paint every panel that came to a printer or sort of found its way to the public in one form or another. And of course, photography comes along and mechanical reproduction and suddenly it becomes almost industrialized.

It's not something that an artist has to touch. And the, the emergence of the graphic designer becomes a, a reality or at least changes in phase and function based on these new tools. And the same thing happened with, you know, the rise of the personal desktop computer and publishing software and graphic design software on computers.

And, we've been through many of these kinds of phases with new technology and how it impacts how and what we create. And I totally understand that. This one feels a little different. This one hits a little different because before the, [00:44:00] every one of those things was clearly like some form of tool, but.

 Was culturally upsetting in, in one way or another. And of course, society, figured out how to metabolize and work with those new technologies and make them beneficial and end up lifting them up to the level of fine art. You know, photography, no one would argue that great, fine art photography doesn't belong in a gallery these days, but that certainly was the conversation when cameras first arose.

 And so I think AI is really similar in that sense, but people do get scared because there's a lot of conversations where people talk about AI being sentient and replacing humans and making humans obsolete. And I just don't think that's true at all. And it's certainly not the kind of tools that I wanna build.

I mean, I think when I see AI applied really beautifully, it's in service of a human trying to do something where they have a skill gap or they have a time constraint, or they have some other hurdle that. [00:45:00] Too challenging for them to get over alone. And AI becomes this really powerful accelerant to help real people today do the things that they just inherently wanna do, whether that's make music or write something or be more organized at work.

 Right. There's so many different jobs that you can hire AI for these days, but when it's in service of a real human intent, real human creative drive and it's acting as a tool in support of that human, I think, that's when people really light up and love to see what AI kind of unlocks for them.

 Just one other like finer point on this I think people get really worried by the idea that AI can replace human cognitive function and creativity. But I actually think that that's not fundamentally what AI models do. AI models compute and deliver the mean the average of a particular request or topic.

 And they trained to do that. That's, the goal of the model, is to deliver [00:46:00] mediocre,  average expected results. And that's just not what pushes culture forward, right? But if you have a particular area where you go really, like you're an incredible painter, but you're really bad at writing your artist abstract, well, gosh, you've got a great tool to help you get your words onto the page in a way that can make sense to other people supporting your intent, driving your painting work forward, but supported by ai, not replaced   by. 

Radim: It's an interesting that you said earlier and you said this one hits differently because when you give an example of a photographer and a fine artist, those are not the society at large. There are fringes. They're literally a groups of people doing this and doing that. Like whatever we be describing, it doesn't affect.

99% of the society whatsoever. I think through the amplification of stories, media, news articles, social media, now information travels fast and [00:47:00] sometimes people do not always, as you would know don't always look beyond the headline, you know, so it's easy to be triggered without having the full story, actually understanding what's behind the story 

 I think we talk about more about AI than we actually use it. there is a case for it, but there's also the misuse of the terminology.

It's a bit of a, more of a scale scale ing just for the click bait in a way. 

 But are some of those conversations just blown out of proportions, way too much. Because ultimately we create stuff as humans for other humans. 

Russell: I think there's definitely a lot of misdirection and hype you know, either from news outlets for the purposes of driving clicks or from builders of tools looking to like. Rise above the noise and create some notoriety for themselves, very practically, having put my hands on and played with so many of these models, these things really are at the level of tools.

They are not sentient beings sent here to eliminate the need for kind of human [00:48:00] independent thought and context and perspective. They really are just responding to the prompt you give them with an average response. And so if you wanna bring creativity. To AI as a human, you have to bring that creativity.

You have to have a perspective that's unique, that's lived in this moment, relevant to this moment which AI doesn't have currently. Right? You know, these things get trained periodically. They're not all knowledgeable about all of the conversations, right? happening in each of the fields where the state of that conversation is what that culture is getting excited about next.

 That's a real thing. It's immaterial, but it's a real thing that exists in the world that humans can get plugged into and draft on. And AI models don't have that context and so can't really produce stuff at the edge of cultural relevance without a human driving them to it. But what they can do is dramatically raise the floor for you as a creator in [00:49:00] areas where you have gaps in skills.

 They can take you from being completely. lacking any kind of knowledge in a particular area to being competent in that area, right? Not exceptional, but average. if you pair that with where you're exceptional as a person to create something larger than you could have created on your own, boy is that an incredible accelerant for creatives and builders and people who have intent and want to further those cultural conversations in each field.

Radim: How would you say the access to AI tools is almost driven by people's desire to know more or their curiosity? Because you and I had this conversation a few months ago about the fact that there's going to be different AI tools for different purposes. no AI tool can really do the one thing really well.

And thinking about it again, from this side of the argument, let's say I know that Photoshop is for, creative stuff with, bitmap stuff and you got, Illustrator and you got [00:50:00] InDesign. It's kind of, it's very well signposted, it's labeled, you got a list of features, you know, kind of what to do with and how, you got new AI models and you're like, it does this and it does that.

And it's almost like you're almost guessing what's underneath the bonnet. It's like buying a car, which is just a wide box and going, what'd you do with this? Where'd you drive it?

Russell: I think that's exactly right. I mean, I think even for the researchers who build these models, I think one of the things that's most amazing is you have a, kind of architecture, you know the different functionalities of what kinds of information, like the thing is tracking and able to sort of like segment out and understand about all the data.

You'll kind of feed it into kind of build its model. But even researchers discover emergent capabilities they didn't plan for that seem to kind of fill in between the gaps of the things they explicitly decided to design and build for. And so there's this really interesting phase with every model that comes out where they basically just hand it to designers and people to try to [00:51:00] do interesting new things that are adjacent to what they know it can do and just sort of see what else has it kind of figured out how to do.

And so some examples of this are like, you know, with video models, which are like gonna be I think a, a huge topic of discussion over the next year as they advance and become like the pretty incredible tools is you train a video model to take in a text prompt request and generate a series of frames that makes sense as a video based on that prompt.

 But you can also ask it to mask out certain things from a video that you hand it and it will do that. And it will just perfectly rotoscope and crop some object because it understands language, it understands how to process frames, it understands how to process pixels, and therefore it just learns how to just do that on its own.

 Which, we call those kind of emergent capabilities or emergent behaviors. And they're happening all over the place. And it's, it's fascinating, but it's [00:52:00] driven by the core data itself and sort of the connections it can make in the data that it already has and the capabilities that it already has, even if those things weren't necessarily explicitly planned for upfront, which is fascinating, but it's not the same thing as truly open human level logic.

It's what are the adjacent things that are also kind of included in the data that we just didn't really think about.

Radim: It's really interesting when you said that these tools are given to designers creative to try what to do with them. ' cause ultimately, as you talked about a sort of history and a cultural history, how many instruments or drum machines were bought for one purpose and then totally reinvented through a misuse in a way, or creativity or just use it for the purpose it wasn't invented for or created for. And what can I do with this? Where can I plug it here? What can I do with this? and, and find a sound that's not in there. And I, when you think for example, I'm thinking of the AO eights and the three oh threes, you know, and you, you get people like chemical Bravos like playing guitar pedals through keyboards, through drum machines.

 that's a beautiful chaos. [00:53:00] But no one can say Hey, listen, was for this particular use. you can't use it for anywhere else because that's where the magic really happens when you think about it. We, we've been given these turbocharged tools that really can unlock I'm now unlocking already sort of opportunities and ways of creating things that would normally be.

 of five or six different hardware, inventions away, if not 10. 

 this is defined. So I think this is, seeing this as an opportunity, I think is potentially a gift for someone because when you fear it, you only gonna be fighting it for a while. But until you really realize they're potentially, you can be really good for you.

Russell: I personally have ton of fun playing around with these things and trying to stretch them and and figure out what are they gonna be good for? why would somebody actively choose this tool or use this in their life? and I think that hacking that you're talking about is the best sign of a great tool is one that can be bent to new uses creatively by a person with [00:54:00] intent, with taste, with a sense of that cultural zeitgeist of, you know, what's going on in their field, what would be interesting to other people.

 think only humans still have that unique perspective.

Radim: I like to use the word taste. It's debatable or shall I, shall I say subjective? 

Russell: Well, I definitely don't wanna say that every video ever created by an AI model is of taste. That is not absolutely what is happening there. But 

Radim: no. 

Russell: human taste still exists and and matters in the open market. And if you're someone who makes stuff, it's still just as important to bring your lived human perspective and taste to whatever you make, whatever tool you use whether it's AI or not ai as it ever was.

Radim: I like for someone who says that, you know, you are always into the future, you are a futurist. Looking at the nature of how things are built. You are in the right place at the right time. I mean, I'm happy for you that this is, this is where you find yourself and you know, you, I feel more excited about learning and creating more than ever, ever before.

Well, more than ever. [00:55:00] But what I want to know for someone who's interested in the future, how do you see the immediate future? And if you would, better to guess where we see ourselves in five years or 10 years, 

Russell: Oh my gosh, that's such a big que you're asking me to tell 

Radim: how do you view, and what do you do from, from where you are?

 What do you reckon you'll be doing in five to 10 years and what you might be doing it with, because I'm not asking you to play God.

Russell: You know, as somebody who obviously has been passionate about software for a long time and, and even dabbled in, in before ai, you know, could do some of this for you, writing my own code and apps and, trying to build. My own artifacts, and release them into the world. But not really being a true developer, you know, always being a little bit limited on that end.

Having to rely on, you a larger team of engineers to build the ideas and, experiments that I wanted to try. And we have just phased changed into a new era where I can spend all day [00:56:00] in cursor building anything I want, immediately testing it out on my desktop, my phone, wherever, right away, having that fast feedback loop, right?

I'm no longer in Figma, sort of like, plan, okay, what's the perfect UI for this experience? devoid of any actual interaction with the software or the model and I can just go plug into a model, start building UI around it. Like really, it feels like going from a place of having to perfectly detailed draw, you know, a realistic drawing of something to sculpt. I feel like I can sculpt software now where I'm bringing rough pieces together. Core functionality, big pieces, getting them locked in, getting them functioning together, and then slowly refining and honing the expression of those, the input, the output, you know, the interactions once it's functional and working to get it really dialed into a place where it's satisfying and compelling and already fully functional.

 [00:57:00] And I think that's so magical. As a creator, as somebody who enjoys creating software as as an artistic medium in, in a sense being able to do that solo and have that immediate feedback loop. Not having to convince an engineer to take hours to hand code some dumb idea that I had in five minutes and drew out on a napkin sketch, but just to literally take a photo of that napkin sketch and send it to cursor and be like, here's what I was thinking.

Two minutes later, being able to play with a prototype of that and make a decision about whether it's any good or not that's what I'm gonna be doing for the next five years I think.

Radim: Well, I can only close up with three words that you said earlier and it says everything is new. When you were growing up, everything was new, and everything what you're doing now, still everything is new 

Because ultimately you see this as an artist, using tools and actually expressing yourself. Because as you said, when you feed into the machine or when you feed into the model, is your creativity, is your [00:58:00] soullessly, is your ideas. 

 it's what you've created with it, rather than hoping that it will create a next masterpiece by just pressing can you create a next masterpiece because that will never happen. 

 Russ, thank you very much for talking to me today. 'cause I really enjoyed yeah, learning more about what you do and I'm excited to see what the future brings.

Russell: Amazing. Yeah, thanks Rodin. Really great to catch up.

Radim: Thank you.

 [00:59:00]