Learn, Make, Learn
Learn, Make, Learn is two product geeks sharing qualitative & quantitative perspectives to help you make, better. Hosted by Ernest Kim and Joachim Groeger.
Learn, Make, Learn
The Perils of Crossing Over from Niche to Mainstream
We dig into the perils of crossing over from niche to mainstream. How can a brand built on a niche product or service successfully expand its appeal? Should this even be an aspiration?
FOLLOW-UPS – 01:24
“Low-Resource” Text Classification
78% MNIST accuracy using GZIP in under 10 lines of code
MNIST by zip
Clustering by compression
Kolmogorov complexity
Normalized compression distance
ChatGPT is a Blurry JPEG of the Web
Last week’s cover art
Apple’s LaserWriter Changed Everything (paywall)
Adobe, Stability AI & Canva are starting to pay artists
On Threads @LearnMakeLearnShow
THE PERILS OF CROSSING OVER – 12:23
American Fiction’s Big Literary Showdown
SCENE FROM AMERICAN FICTION – 14:27
THE CRUX – 16:00
Chris Wilcha Talks Gen X Angst
Hodinkee Cuts a Fifth of Jobs (paywall)
Maciej Ceglowski—Barely succeed! It’s easier!
Pinboard
CROSS-SUBSIDIES – 24:42
The car that saved Porsche (paywall)
Windup Watch Fair
When Movie Stars Become Brands
Adam Curtis Interview
KAIZEN – 29:25
What is Kaizen?
William Gibson interview
Stanley Cup Craze
Nintendo CEO’s refusal to layoff staff goes viral
FOCUS DOESN'T HAVE TO MEAN SMALL – 36:25
WEEKLY RECS – 39:05
On Film Grain Emulation
The “baseline” scene
CLOSING & PREVIEW – 44:50
(Image credit: Clair Folger/Orion Releasing)
****
Rant, rave or otherwise via email at LearnMakeLearn@gmail.com or on Threads @LearnMakeLearnShow.
CREDITS
Theme: Vendla / Today Is a Good Day / courtesy of www.epidemicsound.com
Drum hit: PREL / Musical Element 85 / courtesy of www.epidemicsound.com
Hello and welcome to Learn Make Learn where we share qualitative and quantitative perspectives on products to help you make better. My name is Ernest Kim, and I'm joined by my friend and co-host Joachim Groeger. Hey, Joachim, how's it going?
Joachim:I am good, Ernest. It got cold. it was warm. There was a nice moment where I thought Spring was here. But now it got cold again. I don't know how I feel about this. I wasn't ready. I was emotionally already letting go of winter and that was a mistake.
Ernest:Yeah, this has been definitely an unusual winter, being warm and incredibly cold, and then warm and now cold again, sadly, I, I think we're going to have to get used to these extremes because of climate change, but what can you do?
Joachim:Wow. On that happy note.
Ernest:All right. This is episode five, and our topic today is the perils of crossing over from niche to mainstream. How can a brand built on a niche product or service successfully expand its appeal to a broader audience. Should this even be an aspiration. If not, how can a niche product brand succeed and remain relevant in a highly competitive marketplace? But before diving into these questions, let's start with some follow-ups to our previous episode, when values and opportunity collide. Joachim, do you want to get us started?
Joachim:Yeah, I would love to, just two things. one I enjoy having the follow ups. I think it's great to be able to end an episode and then still pick up things because I'm always thinking about the topic as I leave it and go, oh, I wish I'd said that. So I'm very happy to have this little brief moment where we can revisit some of the things we were talking about. And I know that from last time we edited quite a bit because our conversation went for a good, good chunk of time. So I feel like I would like us to have another round of this, another version of this topic down the line. Yeah, I'm just springing this on you in this forum, Ernest, as opposed to separately do it on the podcast, then it'll be hard for you to say no, but it feels like there are a lot of things that we could still talk about. So for the future, I would like to put a pin in that and maybe we'll have a chance to revisit some of these topics. More directly connected to the topic from last time is, something that I touched on that I wanted to extend a little bit on. And that had to do with the constraints aspect. We talked about machine learning and AI and putting privacy as their primary value that they wanna stand by, and then allowing the technology to follow that principle as opposed to the other way around. And then from that we talked about how constraints could be very, very valuable and could help you design things in a better way. And then I hinted at the fact that a company like OpenAI is essentially completely unlimited. They have unlimited compute, unlimited funds. There's no discipline they can just do whatever they want to do. When chat GPT came out, some researchers published a paper on text classification it has this incredible title. It's low Resource Text classification, a parameter free classification method with compressors. Sounds incredibly dry, but it's really, really fascinating. One application is large language models to classify texts. So if you want to say, this is a news story, or this is gossip, I would like to have a machine classify text that I just start feeding it and tells me, okay, this is a gossip article, and this is a news article. What these guys have done is they said, well, you don't need all of those billions of parameters and huge server farms to do that. You can actually just take the text, run it through a compression algorithm, and then do that with your corpus of other data that you're using for training, compress all of that, and you essentially just look at what comes out of those compression algorithms and then you compare the texts in that, and then you can find similarities that way. And so when this paper was talked about on Twitter, there was a lot of controversy around it, of course, because it was very bombastic. They said they could do better than a lot of the large language models. They had made some questionable choices potentially in the way they executed on that. But the gist of it was you are able to compete with a large language model on your laptop without very much heavy computation. So, they're building off this idea that compute is constrained and, you know, maybe I can only access very simple algorithms, is there's something else that can come out of that that's very innovative. And then that inspired a whole group of people to think, well, what else can I do with that? And so someone tried to use it for character recognition of images, handwritten numbers and digits, and then running those images through a compressor and then testing if they can classify digits So really, really Incredible, and fun result. For completeness, this idea of using compression is quite a deep idea actually, it's really interesting. We'll link some stuff in the show notes about that. There's an old paper from 2004 where people were using these algorithms for, clustering data together. And so you can see it's a 2004 thing, right? Compute was limited. You had to be clever about how you used your processing power and the time to get to results. And here's a very lightweight, efficient way to do that. And we're rediscovering these techniques now and, and finding that they're incredibly powerful. So I love that as, if you impose a constraint on yourself saying, I'm not gonna use more than my laptop or something like that, you can start in creating these really powerful tools that you didn't even think were gonna be possible. I mean, at this algorithmic simplicity, you could imagine putting it on a tiny, tiny computer that's embedded in a device as well. So there's a lot of potential here. And also that reminds me, second follow up, I referenced it in our conversation, but it didn't make it to the final podcast, but I wanted to bring it up again. That this result kind of reminds us of the fact that LLMs are, to a certain extent, really big fuzzy databases. And I wanted to just reference Ted Chiang's great piece, from the New Yorker, where he discusses Chat GPT, and he describes it as a blurry JPEG of the web. And I think it's a really great piece of writing that captures the essence of what this technology is. And then, because you now think of it as a blurry compression of data, now it makes sense that a compression algorithm could perform quite well. So there's a lot of unity in these ideas, and it's that connection between all those things that I really love and I wanted to just pick up on. So constraints and having a constrained view of the problem space can lead to some pretty surprising and fun insight. So we'll link to all of the papers as well as a blog post from someone who did it as a little fun project to see how they could do on an image recognition problem. Absolutely fascinating.
Ernest:Touching back on something you said at the outset. I also appreciate this opportunity to share follow-ups I think it's a great resource for us. And um, I too spent a lot of time over this past week, thinking about last week's topic. And so I did want to share one follow-up. And it's a related to my use of the generative AI platform, Midjourney, to create the cover art for our episode last week. What we see happening today with generative AI platforms like Midjourney and content creators reminds me a lot of something I witnessed firsthand through the early nineties, which was the beginning of the end of the traditional prepress service bureau industry. Back in the days before the web, service bureaus played a vital role as intermediaries between companies that created high volume publications, like books, catalogs and magazines, and the printers that printed them. And for many decades, service bureaus thrived in New York. In fact, one of my first jobs was at a service bureau located near the famous Flatiron building, and my time there coincided with the mainstream adoption of desktop publishing. Thinking back, it's actually remarkable how recent this was, but the concept of desktop publishing was introduced in 1985 with the launch of the Apple LaserWriter, which was the first printer to feature something called postscript invented by a scrappy startup called Adobe. Later that same year, a company called Aldus introduced a program called PageMaker, and together these tools enabled non-expert users to create sophisticated page layouts, and print them at very high quality at work or even at home, if you could afford the$6,995 price tag of the LaserWriter in 1985 dollars which kind of puts the price of the Vision Pro in some perspective. Prior to desktop publishing, as a small business, you'd have to go to a service bureau or a design shop affiliated with a service bureau to have a sign, or brochure, or any other collateral materials designed, made ready for print and finally printed. Which as you might imagine, was a time-consuming and costly process. So in hindsight, it's probably not surprising that desktop publishing exploded in popularity. So much so that, within a decade, the writing was already on the wall for the industry, the service bureau industry. So I think the parallels with our contemporary context are pretty clear. I think there is a difference though. And it's that desktop publishing made existing job superfluous. Joachim, you had mentioned Schumpeter a couple of episodes go. And the advent of desktop publishing is, to me, a textbook example of creative destruction. But in the case of generative AI, while it might make illustrators and designers redundant, It continues to rely on their work to do what it does. And I find that to be a troubling dynamic. And that's because the technology is incredible. You know, it didn't just enable me to do something faster and cheaper. It enabled me to do something I never could have done myself, and did it far better than I could have imagined But none of that would have been possible without the huge corpus of creative work ingested by Midjourney. Work that was conceived and created by humans. But in the case of Midjourney, those humans were not compensated for their contributions. This came to a head in the music space just last week with universal music group, which has an enormous catalog, including work from most of today's biggest musical acts, pulling their catalog from TikTok in part over AI concerns. I'll just read a brief snippet from Universal's open letter explaining this move. Quote, TikTok is allowing the platform to be flooded with AI generated recordings, and then demanding a contractual right which would allow this content to massively dilute the royalty pool for human artists, in a move that is nothing short of sponsoring artists' replacement by AI, unquote. This issue is being litigated in the courts as well. And some platforms in the space are already paying artists for content used to train their models. Whether that's going to be enough to keep the artists whose work feeds those platforms afloat is an open question. And, unfortunately, the biggest players, including Midjourney, continue to insist that their approach to appropriating content is, at least in the US, protected under fair use doctrine. And yet, kind of making the counter argument here, in my case, I never would have hired an artist to create a bespoke cover art illustration. We neither have the time nor money for that to be feasible. So I could truthfully argue that my use of Midjourney did not take work away from an artist who otherwise would have profited from our business. Building on this, I could argue that Midjourney and its ilk satisfy a latent need that currently goes on fulfilled, enabling small to medium-sized businesses to access bespoke creative work to degree that previously wasn't possible, thereby enhancing their economic prospects. Is that a bad thing? Using Midjourney to create the cover art for our last episode, where we talked about the collision of values and opportunity, you know, I have to admit was something of an inside joke. But also a personal provocation. And after putting some more thought into this, following our episode, I've decided to hold off on using generative AI tools for the time being. I'm going to do some more research on the platforms that are paying artists today. And if there are platforms that are approaching this opportunity in a way that's sustainable for the people fueling their models, I'll come back to it. But for now, I've made the personal decision to stop using it. What do you think? Are Joachim and I, in in our misgivings towards generative AI, being simple minded, Luddites shaking our fists at the latest newfangled tech. Or do you share our concerns and agree that there are promising and less dehumanizing alternatives? Let us know by sending your thoughts to LearnMakeLearn@gmail.com or on Threads@learnmakelearnshow, all one word. Okay, now let's move on to our main topic. As I mentioned in the intro, today we're going to discuss the perils of crossing over from niche to mainstream. And I actually need to start with a warning, because, due to some of the references we're going to cite, this episode will include profanities. And speaking of our references, it's also going to include some dialogue from the film, American Fiction, that I highlighted in our recommendations segment last week. We're not going to give away any spoilers, but I wanted to forewarn anyone who wants to go into that film completely cold. Okay, so, what precipitated this topic was a key scene from American Fiction. In it, the film's main character, Thelonious"Monk" Ellison, played by Jeffrey Wright, confronts Sintara Golden, a fellow author played by Issa Rae. As explained in a great piece in Vanity Fair, quote, frustrated academic Monk is haunted by We's Lives in Da Ghetto, Golden's poverty porn bestseller that, to him, represents everything wrong with society's expectations for Black art. It's success inspires Monk to write his own, satirical book tailored to the White gaze. At first he calls it My Pafology, with an F. Then he changes the title to Fuck, unquote. The context of the scene is that Ellison and Golden are part of a panel of five judging a prestigious literary award. The three other panelists, who are all white, want to give the top award to Ellison's book, Fuck, which he wrote under a pseudonym, so the other judges don't know that Ellison was the author. The book was also marketed as a memoir rather than satire, and the three White judges take it at face value as a, quote, unquote, essential telling of the Black experience in America, because it so closely conforms to their expectations for Black art. During the debate over Fuck, Ellison is surprised to find that Golden shares his view that the book doesn't merit recognition. At loggerheads, the panel decides to break for lunch and Ellison has an opportunity to engage Golden directly. Okay, Joachim would you be so kind as to help me share a bit of the dialogue from this key scene by playing Sintara Golden?
Joachim:I will do my best. I haven't seen the film, so I'm gonna just read this blind, but I do want to see it. It sounds fantastic, but I'll do my best.
Ernest:Okay, so I'm playing Ellison and Joachim is playing Golden. I'll start. And so this is now in Ellison's voice. Please don't take offense at this. But how is Fuck so very different from your book?
Joachim:Is that what this is about? You think? My book's trash.
Ernest:No To be honest. I, I haven't read your book. I've read excerpts and it didn't seem so dissimilar.
Joachim:I did a lot of research for my book. Some of it was actually taken from real interviews. Maybe you've been up in your ivory tower of academia for so long, you've forgotten that some people's lives are hard.
Ernest:Your life. You went to an exclusive, Bohemian college. You had a job at a fancy publishing house in New York.
Joachim:So what? I don't need to write about my life, I. I write about what interests people.
Ernest:You write what interests White publishers fiending Black trauma porn.
Joachim:They're the one buying the manuscripts. Is it bad to cater to their tastes?
Ernest:If you're okay feeding people's base desires for profit.
Joachim:I'm okay with giving the market what it wants.
Ernest:End scene. Thank you, Joachim we'll have to save this for our submission to next year's podcasting awards.
Joachim:Thank you.
Ernest:In the film, the scene actually continues for a bit longer, but this snippet is the most salient to our topic today. And those retorts by Golden, which were so beautifully performed by Joachim, are what made me want to focus on today's topic: the perils of crossing over from niche to mainstream. To Ellison's challenge that she only writes what interests White publishers fiending Black trauma porn, Golden retorts,"They're the ones buying the manuscripts. Is it bad to cater to their tastes?" To which Ellison, with frustration mounting replies,"If you're okay feeding people's base desires for profit." Golden's come back to this is the crux of our topic today. She replies,"I am okay with giving the market what it wants." In another era, this might've been called selling out, and seen as a betrayal. But not anymore. I think Chris Wilcha, who co-produced the TV version of this American Life for Showtime, captured this shift really elegantly in an interview with POV Magazine last year. He said, quote, it is shocking how the notion of selling out as an, almost a nostalgic quaintness and as a gen X-er that was a generational obsession. We were worrying about what bands were signed to what labels or what brand was, or wasn't authentically targeting your demographic. I have friends who are musicians and writers, and that distinction is almost meaningless now. You want to sell. You want your song in that market. You want to sing for that commercial. I totally understand that. The hustle is so much harder now. The gen X armor of irony had to come down, unquote. And yet when it comes to products and consumer platforms, it seems that selling out continues to carry negative connotations. Especially for niche brands attempting to crossover into mainstream adoption. Joachim, do you see this dynamic as well?
Joachim:I think we started talking about this in the context of one of our joint passions, which is, funnily enough, the very elite world of wristwatches. It sounds ridiculous. It's what brought us together. So we, we should give credit to the hobby that brought the two of us together to make this podcast. But last time you had alluded to Hodinkee's shift from a blog for people who just like geeking out over watches... Hodinkee was this beacon for a niche audience to come together, share a passion, discuss something pretty silly and pretty trivial, but with no judgment and coming together as a community. So just the classic internet playbook of building a platform where a specific subculture can emerge and find each other. And of course, recently they've, maybe a couple of years ago, they received huge investment. LVMH injected money. One of the early guests on their website who was a big fan, John Mayer also injected money into the platform. And I think that fundamentally changes the nature of what it is that they were doing. That's kind of the context that we were thinking about. And referencing Stafford Beer again, you know, the purpose of the system is what it does. When you're injecting money into a thing, you want more money to come out of that thing. It fundamentally changes the nature of what the product is and what it is meant to do. Before it was okay for it to be a niche audience. And maybe that audience was never gonna get very big in NBA speak. You know, the total addressable market was just that number and that's it. So when it becomes an investment vehicle, it has to do something different. And so when we talk about selling out, we start realizing that it's no longer a thing that just needs to cater to the people that enjoy that product or that service or that music. It now has to generate a return. So that's kind of where I think the fear comes from, is that we understand that its purpose has now shifted
Ernest:Yes, you know, on that point of having to generate a return, it's remarkable how many people I've talked to, particularly people who've started up new businesses, who've told me that the reason their business failed was that they had taken too much money at the outset. And that might sound, you know, just unfathomable to a lot of people. But just as you had noted, Joachim, the challenge is that investment comes with an expectation of a return. And if you take too much of it at the outset, you're forced to really turn that thing that you were starting as a passion into something that could become just completely unrecognizable to you because you find that, you know, to deliver on those expected returns, you have to try to cater to a much broader audience, and then that'll dilute that original nugget of the idea that you were passionate about in the first place. So that's definitely something that I'd call out as a warning to folks you know, who might be getting started with a new startup. You know, funding might seem like this incredible thing, but it comes with strings attached. So you really just have to be aware of that and the effects that that might have on your business.
Joachim:Yeah, this reminds me of a talk that I heard, and we'll link to that again by Maciej Ceglowski. And his talk is called Barely Succeed and Maciej runs a website called Pinboard. Pinboard is a web caching app, so you in install a browser, add-on, and you can store web pages on his service and. His goal was to make it a small business. We don't call things small business. everything has to be a startup. This is only step one and it needs to become astronomical growth, right? But Maciej approached his business as a small business. This thing needs to be sustainable for me to get an, to get an income. And so he built it from the ground up to be niche to a certain extent. And he got to the point that he could basically live off that and he didn't need a full-time job anymore. That sounds incredible, right? That's a very focused thing that says, I just need to sustain one person. I need to run this thing as one person. He has the skills, of course, luckily to be able to do all of that, but it's essentially the modern equivalent of a small business as we would recognize it. And so I really like his pinboard model. I think that's a great counterpoint to what Hodinkee did, or they could have just been. Self-sustaining. You know, there's three passionate people who want to talk about watches. Okay, let's run this to make enough money so that these three people can earn a comfortable living off of that. I think that's a totally fine alternative and the way you generate growth, quote unquote, is by becoming more efficient, getting good at it being able to generate the content and increase the quality of your output without having to put more money down. You know, I think we're very obsessed with literally using debt a k, a leverage to, you know. Jump up to some other scale from zero to nothing. So I found that your discussion reminded me of the question, is there a counterpoint to this constant growth and the VC fueled hype beast and the pump and dump dynamic, right? This perpetual growth approach works really well if someone at the end is willing to take the asset off your hands. You kind of have this hot potato dynamic that's pretty unhealthy so I, I have to say, I've, I really enjoyed Hodinkee when it was focused around telling me about watches and now it's very hard to tell the difference between content and advertising. So many pieces are now co-branded with other people. So I'll start reading something about a G-SHOCK But you start realizing the tone is a little bit off. They seem to really love g sh I mean, we all love G-SHOCK, obviously, but well, obviously it's a, a partnership with G-SHOCK and sneaking an article. As a review, that is basically an ad. They're eroding their own brand in the process of turning it into an investment return vehicle. And if you're thinking about just having this be your. Alternative income source, your goals will look a lot more healthy, right? You don't need to look at your income and every month you'll be thinking, or every three months, you'll be saying, well, we better get 30% increase on my salary now. That, that that's not something that we expect so. Then if you're running the business to sustain yourself, you wouldn't have that as your goal. You'd just be saying, in five years time, I hope there's a little bit more there. So that's something that comes to mind when I think about this. The shift away from niche to something more extreme.
Ernest:Now on this topic of running a business to to sustain yourself that you just mentioned. Something that we had talked about offline as we were preparing for this episode was that there could be instances where you find that, in order to sustain your business, you do actually have to extend your product offering in a way that reaches a more mainstream audience. And a classic, maybe the classic example of that is Porsche. You know, ah, when Porsche introduced their first SUV, the Cayenne, Porsche enthusiasts, who tend to be 911 enthusiasts, just hated that concept of a Porsche SUV. But, that was essential for Porsche to do to enable them to continue to create the 911. Porsche without an SUV business is no longer a business in today's hyper-competitive automotive marketplace. So, you know, that's something to be mindful of as well, is that you may find that in order to enable you to continue to focus on that core thing that you love and that your audience loves, you may actually have to build on your existing business in a way that allows you to continue to invest in that core Now to your point about growth through other means like say, efficiency, that brings to mind another example that comes back to that watch space and Hodinkee Um, and another brand in that world, which is a website called Worn and Wound. Historically they've focused on more of the entry level a portion of the watch market. And the way that they were able to grow was not by broadening their audience, but by offering more to that same core audience of watch enthusiasts. And they did that by expanding their existing website with events real-world events. They featured some, you know, relatively big brands but for the most part, they really focused on that core of entry level brands, but also independent brands that was really where they got their start in terms of the focus of their website. So I think that's another really nice example of a smaller brand, a niche brand, being able to drive growth without diluting the core of what your brand stands for and instead building on it in interesting ways that allow you to maintain that connection with your core audience.
Joachim:I love this notion of the cross subsidy, I think, in the arts world, that's a very common thing, right? Someone as an actor will do the big movie so that they can then do the small movie that they love doing. This reminds me of a YouTube video from a video essayist who talks mainly about movies. Patrick Willems, again, we'll link. He did a whole thing about. How movie stars, some movie stars have become brands upon themselves and even their movies feel like extended commercials for their own personality and their brand. So the two people he singles out are Ryan Reynolds and the Rock. Patrick Williams in his video points out this interesting shift away from. Commerce in order to create more art to subsidize your art. And he contrasts that with George Clooney, who famously was an early investor in a tequila brand Casamigos, which then was sold to a huge conglomerate, and he made an ungodly sum of money. But as a result of that, he's openly said, now I'm able to take risks. I'm able to innovate. I have the safety and security to. Push the boundaries of what people wanna see in the cinema. I don't have to worry about making money from that. I like the worn and wound example as well, they just expanded the community into the real world and they knew exactly people would be ready to participate with that. It just feels like an extension of the website into the physical realm. In this era of commerce and overexposure and all of those things, it really feels like the most radical thing that you could do as an artist or as an innovator is to basically be completely anonymous. I. Right. A documentary filmmaker called Adam Curtis, who is with the BBCI, I strongly recommend his movie. They're very mesmeric and quite strange. He has a very particular view of the world, but I. In an interview. He said the most radical thing that you could do today is make art and tell no one about it. If art is supposed to be subversive, then that probably is the only thing that you could do is be completely invisible.
Ernest:Now something else you had mentioned in our offline conversation about this was this concept, this Japanese concept of secret brands. Could you talk more about that?
Joachim:I was thinking about secret brands in the context of a William Gibson novel. Another reference to him, he just seems to have his pulse on these things in, in a way that's a little bit ahead of the curve. And the book I was thinking about is Zero History which picks up the story of one of the main characters from the second book in that group group of books called Spook Country. But in zero History Hollis the main character has been asked to investigate a brand called Gabriel's Hounds, and they make denim, incredible denim. It seems like from the description in the book So he or she is describing it. It's what the Japanese call a secret brand, Hollis said only more so this may or may not have been made in Japan. No regular retail outlets, no catalog, no web presence, aside from a few cryptic mentions on fashion blocks. eBay Chinese pirates have started to fake it, but only badly the minimal gesture. If a genuine piece turns up on eBay, someone will make an offer that induces the seller to stop the auction. So there's this incredible vibe surrounding this brand because it's so secretive. And that's, I think, really magical in our era now of oversaturation. I think something that comes to mind that's pretty topical now are the Stanley Cups, you know, super saturated collabs. Crossovers. I mean, absolutely crazy. You know, Crocs is another one that comes to mind, you know, that were initially a weird niche ish product, ugly, gross, slightly practical, and then the captures, the zeitgeist. Perfect. We, in fashion, for some reason, I'm, I'm struggling with that one, sorry. But, I like the idea of that there could be a space for someone to just make a damn good product and not have to, you know, make a big fuss about it. I really en enjoyed that idea and it's a long run horizon. You're not trying to milk the market now, give the market what it wants now, and then walk away taking a long run perspective, you would want to build a brand in this way. You want the market to grow organically, and you don't want people to be forced to purchase something.
Ernest:Now this isn't a secret by any means, but, what you just said a second ago about, taking a longer perspective, taking a long run perspective, brought to mind something we saw recently You know, as you might be aware Microsoft attempted to acquire Activision Blizzard and the FTC tried to block that. But as part of that process a whole slew of emails from Microsoft leaked out, and many of those were from Microsoft's gaming CEO, Phil Spencer. And in those uh, messages, he made it very clear that he wanted to, he hoped to acquire a Nintendo. He said in one of his emails, getting Nintendo would be a career moment. But what was really interesting is that in another email, Spencer noted that,"the unfortunate or fortunate for Nintendo, situation is that they're sitting on a big pile of cash, and they have a board of directors that until recently has not pushed for further increases in market growth or stock appreciation." So, you know, you could see in the tenor of his emails that for Spencer, he just can't comprehend the idea that you wouldn't just want infinite growth and you know, want more money. Uh, The idea that you could possibly have enough was just completely foreign to him. But I think that's something that you see you've seen very consistently in Nintendo is that they're playing this much longer-term game. And there's also this sense that, you know, there is a point where enough can be enough. And that they prioritize things beyond just infinite growth. I think Nintendo is also a great example of building around constraints, you know, something that you talked about earlier as well. Back you know, a couple of generations ago, Nintendo just didn't have the capacity to compete with Microsoft's X-Box or Sony's PlayStation when it came to just the pure horsepower of their machines. And so within those constraints, they innovated and came up with the, Wii. And then I think you saw that again with their release of the Switch. You know, they can't compete on those same A speeds-and-feeds grounds. So instead they, embrace those constraints and also, you know, go back to their values of centering everything they do around fun. Um, And that's enabled them to come up with these great innovations that sort of, zig where everyone else in the market is zagging. So I, I, I just love Nintendo as an overall example of, you know, maybe not being a secret brand, I think they're pretty well known, but of taking a very long-term approach and showing that that can drive great success for a company.
Joachim:Oh, man, I love that. It's, I've, I'm, as you mentioned, Nintendo and these connections to Japanese culture that I'm thinking. Oh no, maybe this is my half Japanese background that's kicking in, that's making me feel comfortable with, you know, humble and not over the top growth, not explosive growth. I said it before, but you know, I said the way that you could still grow. And still make more money is you get better at what you're doing and you get more focused and you get more efficient. And of course the Japanese have that idea of kaizen continuous improvement, it's never good enough. Keep finding those improvements. So I really like that idea of just continuously looking at what you're doing and asking yourself, is there a way to do this better? Nintendo plays the game of you know, let's make this sustainable. And when we sell a device, that device will make money. So yeah, we cannot give you the, the craziest graphics, but we're gonna make money on this thing. And what we'll center is fun. because we're gonna focus on the job to be done for the game, which is to entertain you. Put another way, Kaizen stops you from trying to take out debt to fuel the growth. It makes you look inside into the process and improve that process. I. It's a constraint really, if you think about it.'cause you're saying, I'm not gonna bring in new stuff. I'm gonna just focus on what we're doing and we're gonna get really, really damn good on this. Again, I can call back that little example of those compression algorithms, you know, I'm going to figure out how to use this thing and we're gonna milk it and we're gonna get something very, very powerful out of that. So yeah, maybe that's gonna be the, from my side, kind of the heuristic, I'd like to stick to Kaizen and continuous improvement to generate growth.
Ernest:To your point about focus. I wanted to cite just one more case study that I was directly involved in to share an example of how focus can enable you to drive success at a pretty big scale. Yeah, I can't mention any specifics here, but I was involved in a program where I saw this kind of Uh, boom, bust cycle play out in real time. Um, Just like you'd mentioned with like say uh, the Stanley Cup and Crocs. But we had a product that was originally created based on a very niche insight. It was very true to, that insight and I think because of that authenticity, it became pretty popular and we decided that we would broaden distribution. And you know, what happened is that it did become even more popular, reached a very mainstream audience. But it also then fell off a cliff pretty quickly. You know, it really um, went through that boom bust cycle. And when I came into the project was when it was really at the bottom of that bust cycle. And the way my team was able to bring that product back was by coming back to that focus on that original niche, and the needs of that audience. Um, We really focused in, on, you know, what they were looking for again. But also finding those things that they wanted improved. You know, they loved that original concept, but there were things that they thought we could do better about it. So, you know, we really took a step back from that mainstream appeal, focused on the niche and then relatively quickly that product you know, was successful with that niche audience, but it came to be a very broadly successful product, selling in over a million units per year type volumes. But all of that success was predicated on us coming back to a focus on that niche. So I just wanted to emphasize that. Focusing on a niche, doesn't preclude you from pretty big levels of success. But you just have to remember that that success is predicated on staying true to the needs of that core audience.
Joachim:I agree. Yeah.
Ernest:All right now that you've heard our perspectives. We want to hear from you. Have you ever in your work, creating products faced this tension between niche and mainstream? If, so how did you resolve it? Let us know at LearnMakeLearn@gmail.com. Now let's move on to our recommendations of the week. Joachim, what has you excited this week?
Joachim:I revisited an old blog post from Steve Yedlin. Steve Yedlin is the long-term cinematographer of Rian Johnson, Steve Yedlin has been Rian Johnson's cinematographer since I think his early films, including Brick. And Yedlin runs this blog that is just fascinating. And one of the blog posts that I think is really interesting for me is that titled On Film Grain Emulation. So Steve Yedlin has worked tirelessly to get digital cameras to behave and respond to light and action. Like an analog film camera. Digital behaves very differently from film, but we love film. So how can we get those elements that we love about film? Onto digital really deep thought, which is why those films look great and you wouldn't think they're shot on digital. They don't feel cold or. Like a Marvel movie. Sorry. Everyone who's been working on Marvel movies, no offense but Lin has taken his time to figure out these techniques and one of the things that he's really obsessed about is film grain emulation, getting the graininess of a digital image to feel like analog film. And this piece describes his approach. I had been stumbling around with a data science problem where we had really high resolution time series data, meaning multiple signals coming through over time, and it was being stored in a huge database that no one had looked at in a long, long time. And I was just fiddling around thought, you know, I think I'm gonna try and summarize the data at a high enough resolution that there's a way to essentially reconstitute, almost a generative model, so I had done that and then I read this Steve Lin piece and I go, wait a minute. He stumbled on the same thing, but he's applying it to film grain emulation, and I was absolutely taken by that. I thought, oh my goodness. This is fascinating, absolutely fascinating that there are these parallels these connections between the context I was working in, which was just totally quantitative. And then here's this guy who's a cinematographer, and he's thinking about these deep concepts to get. Grain to look right. So I recommend reading this thing. It's really interesting. I've been kind of keeping it a secret because he stumbled on this very interesting approach. But, you know, maybe if he ever listens to this, he'll know that I was interested and, you know, can we please talk, Steve? I just think there's something going on here that I, we should be thinking about more, but a really fascinating piece and again, one that speaks to. Kaizen continuous improvement. He wasn't gonna accept that digital is digital in the way it looks. I love that, that whole philosophy that surrounds it. But this particular blog post, I just has a personal resonance with me. So yeah, that's my rec for this week.
Ernest:Coincidentally my recommendation this week is also a blog post. And it's also related to film. It's a blog post by a woman named Andi. And she, this was actually from about a year ago. She wrote this incredible post about blade runner 2049 and a scene from that movie that's known as the baseline test scene. And I won't spoil anything about the movie. If you haven't seen it, it's just incredible. Um, Such a richly textured, beautiful movie. But I remember. Being intrigued by this baseline test scene, but I just didn't really understand that there was something about it that. Caught me, but I didn't quite. Grok it, I guess I'd say. And then I read this blog post and it just, everything clicked. This a story that you told about the Genesis of that scene. And its intent and what it's conveying. All absolutely kind of put the puzzle pieces in my brain together. And gave me a richer appreciation for the film so much so that I went and rewatched it this past week. And I just thought that this was such a great example of the wonderful qualities of the web. That it gave this person. The space to write this deeply researched piece on this very specific topic. That was of interest to them. Without really having to be concerned about whether it was going to draw an audience. And then thanks to the hyperlinked hyper-connected nature of the web. I was able to find this article, thanks to a thread posted by Jason Kottke. So it just struck me as one. I love the subject, the writing, but also, you know, Coming back to these topics that we've been talking about, especially around generative AI. That it shows the incredible things that are possible When you have this human to human connection, that isn't mediated by large language models. Instead you have a human who was engaged by this work of art, which inspired them to write this fantastic piece. And then just thanks to the, you know, simple hyperlinking technology of the web, I was able to find it. So um, I just wanted to highlight that and, ah, we'll, you know, we'll include links to both of these blog posts in the show notes as well. All right. I think that does it for us. Thank you so much for joining us here at Learn, Make, Learn. As I mentioned, we want to hear from you. So please send any questions or feedback to LearnMakeLearn@gmail.com. And tell your friends about us. In our next episode, we're going to discuss the role of experimentation in product innovation. Experimentation and AB testing have become the standard at all tech-driven enterprises, but as this a good thing? Are we losing discipline when we assume we can just fix it later? Reid Hoffman of LinkedIn fame summarizes this perspective as, quote, if you're not embarrassed by the first version of your product, you've launched too late, unquote. So, is it better to be embarrassed or to launch a product that you can stand by? We'll share our own experiences and perspectives on this topic on the next Learn, Make, Learn.
