I worked on this problem for more than a decade, as co-founder of what became the Center for Humane Tech, and in other roles.
One way to break it down is different users, but as Ivan notes towards the end, we all have Marl in us. So, another way to get at the same thing is to exclude certain engagements from the metrics — the things we click on but would not reflectively endorse as meaningful. You get a lower engagement number that represents meaningful choice, rather than just "revealed preference" / engagement.
This is what I work to align LLMs with at the Institute for Meaning Alignment[1], and Ivan is helping! I also have a paper[2] on the difference between revealed preference and meaningful choice.
(It's also worth noting that this process of enshittification doesn't just happen in software. Markets and voting also have this revealed preference vs meaningful choice problem. So, making this distinction is a chance to upgrade all of our large-scale systems.)
2. Hobbyist builder that do sell out (at a good price) can use the capital to build even more hobbyist tools, instead of becoming a VC or retire on an island somewhere.
Hobbyist builder that fails both 1 & 2 is just another kind of Marl.
Doctorow has studied this phenomenon under the term of "enshittification". His analysis is different though. Instead of look at the psychology of the individual user he takes a more systemic view and looks at the economic incentives. In a nutshell, an online service can either serve the user well or maximize revenue but not both. Would be interesting to take a deeper look at these seemingly opposing but probably complementary analyses and to compare "Enshittification" with "The Tyranny of the Marginal User".
I think Apple did a pretty good job of developing and evolving their software while serving pretty much everyone, including the "marginal user". Maybe the article is more about social software than software in general?
I would kill... KILL to resurrect OKC with some developers - only instead of marketing and binary questions, do so with UX Research question formation in practice. Call it OKKupidating or something. You think 2016 was cool? You'd have loved it in 2010.
The questions were shit though, and the user generated ones were awful. But get a few dozen UXRs on it? Mercy - what we could do!
Just (finally) read this, which touches on many of the same things. It's a good read, and overall aligns with Doctorow's observation that an online service can either serve the user well or maximize revenue, but never both.
Personally, I've just been turning off more and more of these services and figuring out that life is pretty good without most of them. In fact, it's better when I'm not spending a good chunk of it trying to squeeze my brain into Marl's preferences. It doesn't solve the larger problems, but at least I'm happier...
"Since [...] economic incentives operate on the margin, a company with a billion-user product doesn’t actually care about its billion existing users. It cares about the marginal user - the billion-plus-first user - and it focuses all its energy on making sure that marginal user doesn’t stop using the app. Yes, if you neglect the existing users’ experience for long enough they will leave, but in practice apps are sticky and by the time your loyal users leave everyone on the team will have long been promoted."
Implicit here is a choice of time unit. Do I care about the marginal user today or the marginal user(s) over the next year? Choosing a larger time unit those who will leave the platform come into focus again.
Thus, afaiu, the bigger issue is that the economic incentives at work here discount the future too much. This problem of discounting the future too much is a much larger problem. For example, it is at the heart of why we do not tackle seriously climate change and biodiversity loss.
The industry has taken us back to mainframe computing of the 60's, this began in 1997 with pc games with ultima online, lineage and everquest, they started removing multiplayer networking code and rebranding pc games in order to steal them. You're 26 years too late buddy.
They are removing plaintext root access to our computing devices, when the iphone was launched in 2007 and android shortly thereafter, apple and google made sure to transfer ownership of the device and software to themselves to jack up software prices.
Fantastic commentary on a painful topic that I would suggest sits uncomfortably with us all. TikTok is a horrific example of the way UX is heading, and I can only hope that humanity has some sort of awakening to less ‘easy’ content.
This has been such a painful death blow to the Internet. It went from a fun place that we could come together and discuss ideas or share funny memes.
Now it's just stale, forced, oversaturated and over socialed. Too many "content creators" are just normal people desperate for attention who never touched the old internet and got the original concept. Internet, to them, is just a place to get money and attention from.
Wow. Well put. The scariest thing is, this translates even to domain-specific apps such as Navionics Boating. I use it every time I go out, because, somehow, they've not yet managed to touch the charts and rendering and it just works, better than any of the competitors. But, the rest of the interface is like a Fisher Price toy. You want to add a waypoint based on a specific lat/long you got out of a pilot book? There is no such thing as "Add waypoint" in the UI, nooo, you enter the lat/long in "Search" and then tap on something or other to add it as a waypoint.
This attitude manifests itself throughout the application's UI, as if, indeed, the application is optimized for "Marl’s tolerance for user interface complexity is zero.".
Loved everything about this article, and I see this happen with so many things. The things most vulnerable are the things that take armies of people to create. One person can create music, and so while there are the Taylor Swifts of the world (who I would argue is the music for your Marginal Users), there are hundreds and thousands of more authentic musicians with more complex, challenging ideas that also experience success and get attention. But start to move to something like software, like Google products, it takes a vast amount of resources to accomplish that. And then you start getting subject to the forces you discuss.
The Tyranny of the Marginal User
I worked on this problem for more than a decade, as co-founder of what became the Center for Humane Tech, and in other roles.
One way to break it down is different users, but as Ivan notes towards the end, we all have Marl in us. So, another way to get at the same thing is to exclude certain engagements from the metrics — the things we click on but would not reflectively endorse as meaningful. You get a lower engagement number that represents meaningful choice, rather than just "revealed preference" / engagement.
This is what I work to align LLMs with at the Institute for Meaning Alignment[1], and Ivan is helping! I also have a paper[2] on the difference between revealed preference and meaningful choice.
(It's also worth noting that this process of enshittification doesn't just happen in software. Markets and voting also have this revealed preference vs meaningful choice problem. So, making this distinction is a chance to upgrade all of our large-scale systems.)
[1] https://meaningalignment.org/
[2] https://github.com/jxe/vpm/blob/master/vpm.pdf
1. Hobbyist builder doesn't have to sell out.
2. Hobbyist builder that do sell out (at a good price) can use the capital to build even more hobbyist tools, instead of becoming a VC or retire on an island somewhere.
Hobbyist builder that fails both 1 & 2 is just another kind of Marl.
Doctorow has studied this phenomenon under the term of "enshittification". His analysis is different though. Instead of look at the psychology of the individual user he takes a more systemic view and looks at the economic incentives. In a nutshell, an online service can either serve the user well or maximize revenue but not both. Would be interesting to take a deeper look at these seemingly opposing but probably complementary analyses and to compare "Enshittification" with "The Tyranny of the Marginal User".
I think Apple did a pretty good job of developing and evolving their software while serving pretty much everyone, including the "marginal user". Maybe the article is more about social software than software in general?
I would kill... KILL to resurrect OKC with some developers - only instead of marketing and binary questions, do so with UX Research question formation in practice. Call it OKKupidating or something. You think 2016 was cool? You'd have loved it in 2010.
The questions were shit though, and the user generated ones were awful. But get a few dozen UXRs on it? Mercy - what we could do!
https://news.ycombinator.com/item?id=37509507
Comments on hackernews
Just (finally) read this, which touches on many of the same things. It's a good read, and overall aligns with Doctorow's observation that an online service can either serve the user well or maximize revenue, but never both.
Personally, I've just been turning off more and more of these services and figuring out that life is pretty good without most of them. In fact, it's better when I'm not spending a good chunk of it trying to squeeze my brain into Marl's preferences. It doesn't solve the larger problems, but at least I'm happier...
https://www.amazon.com/gp/product/B093G9TS91
"Since [...] economic incentives operate on the margin, a company with a billion-user product doesn’t actually care about its billion existing users. It cares about the marginal user - the billion-plus-first user - and it focuses all its energy on making sure that marginal user doesn’t stop using the app. Yes, if you neglect the existing users’ experience for long enough they will leave, but in practice apps are sticky and by the time your loyal users leave everyone on the team will have long been promoted."
Implicit here is a choice of time unit. Do I care about the marginal user today or the marginal user(s) over the next year? Choosing a larger time unit those who will leave the platform come into focus again.
Thus, afaiu, the bigger issue is that the economic incentives at work here discount the future too much. This problem of discounting the future too much is a much larger problem. For example, it is at the heart of why we do not tackle seriously climate change and biodiversity loss.
The industry has taken us back to mainframe computing of the 60's, this began in 1997 with pc games with ultima online, lineage and everquest, they started removing multiplayer networking code and rebranding pc games in order to steal them. You're 26 years too late buddy.
They are removing plaintext root access to our computing devices, when the iphone was launched in 2007 and android shortly thereafter, apple and google made sure to transfer ownership of the device and software to themselves to jack up software prices.
https://cs.stanford.edu/people/eroberts/cs201/projects/trusted-computing/index.html
Friggin Moloch I swear...
All my homies hate Marl.
But it isn't Marl who is making decisions about the product, you said it yourself: it's management and ultimately stockholders.
Fantastic commentary on a painful topic that I would suggest sits uncomfortably with us all. TikTok is a horrific example of the way UX is heading, and I can only hope that humanity has some sort of awakening to less ‘easy’ content.
This has been such a painful death blow to the Internet. It went from a fun place that we could come together and discuss ideas or share funny memes.
Now it's just stale, forced, oversaturated and over socialed. Too many "content creators" are just normal people desperate for attention who never touched the old internet and got the original concept. Internet, to them, is just a place to get money and attention from.
I miss the older days.
Wow. Well put. The scariest thing is, this translates even to domain-specific apps such as Navionics Boating. I use it every time I go out, because, somehow, they've not yet managed to touch the charts and rendering and it just works, better than any of the competitors. But, the rest of the interface is like a Fisher Price toy. You want to add a waypoint based on a specific lat/long you got out of a pilot book? There is no such thing as "Add waypoint" in the UI, nooo, you enter the lat/long in "Search" and then tap on something or other to add it as a waypoint.
This attitude manifests itself throughout the application's UI, as if, indeed, the application is optimized for "Marl’s tolerance for user interface complexity is zero.".
Loved everything about this article, and I see this happen with so many things. The things most vulnerable are the things that take armies of people to create. One person can create music, and so while there are the Taylor Swifts of the world (who I would argue is the music for your Marginal Users), there are hundreds and thousands of more authentic musicians with more complex, challenging ideas that also experience success and get attention. But start to move to something like software, like Google products, it takes a vast amount of resources to accomplish that. And then you start getting subject to the forces you discuss.
Great artwork too!