Mills Baker


Dude, I just launched 360 photos and showed my parents and they were both like "I don't know what that is but it seems like you worked really hard on it so congratulations"


My mother like... It might actually kill my mother via boredom / disinterest if I tried to tell her what I do

My dad loves it, but my mom is just not really... Actually, I'll put it this way: to my mom, this is all just "business shit," and no business shit is different from any other; me saying "Wow, I think my proposal to balance sharing against engagement in feed is going to work and people will share more" is the exact same as some dude saying "Man, we've got Bradfield signed to this heavy equipment contract that is so favorable to our market position" or some shit

My fiance is the same TBH; what she loves about me isn't what I find interesting about me / my life. And sometimes it hurts my feelings a lot, like no joke a lot. But what can you do? People can't control what interests them, not well, I don't think!

But... doesn't it bother you a bit that we dedicate so much of our time towards something that means nothing to them?

I don't think it does anymore, because honestly: does it mean anything? Or actually I guess I'd say this: I think you're one of the best people I've met and I have insane enthusiasm and respect for you, and if you had quit Facebook and not launched 360 not one iota of that would be different. If you had never worked at FB and I'd met you and you were like a dog-catcher, I bet it would be the same.

I'm super in love with this author Simone Weil, and I love this weirdo quote of hers: "What is sacred in a human being is the impersonal in him… Our personality is the part of us which belongs to error and sin." (She was Jewish but liked to use the concept of "sin" to mean like "immoral mistakes" or "insufficiently rational thought")

So what do you take into account to have "insane enthusiasm and respect" towards someone? Most people would look at someone else's accomplishments, no?

No, accomplishments are nothing!

You might start there, but honestly: do you?! Have you ever?

What are my accomplishments?!

Friendship values something vastly more real and important that accomplishment, which is always contingent on the messy and unfair and irrational world

Lock a great person in a room, prevent her from accomplishing anything; will she not be great? Somehow what we love, what we detect and honor in friendship, is always greater than anything in the world, any status or achievement I think.

And so like, our parents don't necessarily care about our tastes, opinions, politics, morality, jobs, clothes, choices, etc. etc. But they do love us more accurately and more deeply than (e.g.) sycophants at conferences who love our apps.

This reminds me of a great Kundera observation: if a woman loves you for being smart, handsome, successful, deep, courageous, it means nothing; any woman would love those things! Who cares about being loved for obvious goods?! What matters is when a woman loves you despite the fact that you're stupid, ugly, failed, shallow, cowardly, etc.

So, judging someone by their accomplishments is only an artifact on the surface? What happens when you dig deeper?

I'm not sure; it's mysterious for sure!

But I can tell you this: I don't think either you or I likes the other for anything we've done.

I like lots of what you've done, but if you'd not done it I'd like you the same.

Haha, we haven't done much --let's be honest

Like, how do you explain someone what you do?

I'm still wrapping my head around how to explain someone what I do. Sometimes —like when going through customs — I just say IT.


I mostly tell people: "In theory I'm a social systems designer, but I spend most of my time chatting, writing, arguing, and failing."

Well, it depends a lot on whom I'm talking to; for the sake of concision with folks not in this industry, I usually just say "I design how software works," and the level of detail increases if (1) they're in the industry or interested in it and (2) they seem to want more detail!

If they want detail, I might say something like this: "I'm a social systems designer, which means that I try to foster emergent ends from social software products; in general this means using standard or commodified —"common"— interface patterns and conventions, so I do little visual, UI, or IxD invention, and much more analysis of what to build, why, how users experience software (and social systems), how to align company or product goals with user goals, how to know what people really want, how to structure development so that we're testing at the right points in the right ways, and so on."

If they reallllllly want detail, I usually try some analogies: "What would you do if you wanted more people to vote than vote today? You could make it illegal not to vote, but of course in the free markets we lack that option; you could try incentives, but that could morph into bribery / kickbacks; you could try making voting easier, or increasing the clarity of the effects of voting; etc."

Ok. Let's unpack that. How do you know what people really want?

I think the best way is to look for the freest choices people presently make and try to understand, in more or less classical liberal / individualistic terms, why they make those choices; so the first step is understanding, but not in terms of asking (since I feel like most anything that can be said on a survey is either obvious or false)

As a general rule: I am interested in taking the behaviors that have evolved in societies and communities and translating them into software environments; ideally we can make iterative improvements to how those behaviors work, for the individual and for the society, but we should be extremely cautious in attempting to do so, as even trying to achieve this is prescriptive as hell and probably misses the majority of motivations and experiences users care about

That's fascinating. Would you still call yourself a "designer"?

Do you even have an opinion on Instagram's new logo?

Hahahahah I do not. I mean: I have a subjective opinion that's approximately as "knowledge-based" and meaningful as my taste in sweets or sonatas or shirts, which is to say: not very.

Haha I like your shirts

Thank you —I do too!— but if I'm really honest, they aren't reflective of any system of thought or even a system of aesthetic notions; they are truthfully the result of my absorption of ambient zeitgeist trends and norms, and mostly seem to me proof that I am in a slightly cowardly way willing to adopt whatever totems my tribe rewards.

So you touched on something that is very near to my heart —being "prescriptive" with how you "improve" societal norms when you emulate them online

Yeah, I'm obsessed with that question myself; I would bet we're both obsessed for the same reason: this is the fundamental political question, the fundamental human social question.

Whenever I'm really pressed on "social systems design," I cite governmental constitutions as a great example of it

If you want a society in which people don't murder one another —let's say— do you put "thou shalt not murder" in the constitution? What principle could be more important than that? And yet (e.g.) the US constitution is more or less silent on the question of murder! How can that be?

It's because even something as profound as that question —and seemingly as simple— is not up to framers of social systems; in reality, it's up to the users of the system, and if some users feel that (a) shooting a burglar fleeing your house is murder while (b) shooting a burglar entering your house is not, while others differ, letting these constituencies adjudicate the matter and comparing how they differ over time seems the least prescriptive solution —and therefore the likeliest to lead to the right answer over time.

In the same way, classic "moderation" problems are never as simple as people think; if it were merely a matter of spelling out "bad behaviors" and then enforcing policies, the entire world would run with legal codes the size of matchbooks and nothing would be hard.

Instead: one person's harassment is another's freedom-fighting; one person's victimhood is to another actually an abusive framing of a disagreement; etc. It's very easy, incidentally, to be accused of moral relativism when you espouse the view, roughly as I do, that 90% of online bad acts are the result of norm collisions. But I am not a relativist; I just don't think I have any greater *political* rights than any other person.

(I still believe I know what's right and wrong and I still agitate for my beliefs)

Let me ask you this, is a designer always at service of what users want?

What could be more important than designing systems that allow communities to develop, clearly express, and compare the results of their norms and beliefs?

It's virtually like inventing the fuckin scientific method! If we do it well, which we have not to date I don't think (unless you consider "the Internet" a "designed software product," which I do not).

Some people say "users don't know what they want"

Some governments say "people don't know what they want"

And people *want* Trump

It's an empirical fact

Dave Eggers wrote a nice piece on attending a rally of Trump's, a rally at which he saw POC and interacted with scores of people who seemed ordinary, nice, and completely indifferent to Trump's policies per se; like the supports of Sanders who claim they'll vote for Trump, none of these people really cares about (or probably even understands) the policies in question and what their effects would be. Instead: they dislike and mistrust "the system," which for the American right and left alike means "entrenched government interests" + "corporations (for the left: all; for the right: media and entertainment)" + "lobbyists" and they want a candidate who says he'll smash it all and seems capable of it.

Sure. But at what cost, right?

That's the tragedy

Not their motivations but their sensitivities

The reason Trump is a bafflement to people like us is: we think voters are electing policy platforms, which has never been the case. Not only do they not read about or understand governance at that level —frankly, neither do I— but they know from experience that who gets elected president seems curiously irrelevant; electing Obama seemed to have literally no effect on foreign policy, as an example (at least from the point of view of the pacifist). For these people, "the system is broken" and only "outsider candidates" are resonant.

And as designers, we should ask ourselves: what's broken here? Human nature, or something about the design of this system?

I'll share my opinion: it is the latter. I like humans and trust them. I believe voters now have plenty of reasons to feel disempowered, abused, mistreated by elites who don't respect their values. I believe both the left and the right have every reason to feel that way!

I think my question is a little different

And it relates to the concept or the elite versus the masses

It's very popular to defend the masses

Well, it depends on which masses and which cocktail party you're at!

But in both politics and design, is there a group of people that actually know more/better than we can ask for?


There is not *one* group, nor is there a stable group within any area

When there is, the power of control corrupts them anyway

The way it should work is as follows:

(1) politicians, rationally incentivized by election processes, address problems (looming, urgent, consttiuent-suggested, etc.) by proposing and enacting policies that they create with the people they deem likeliest to achieve the sought-after results, such that they'll not be voted out for failing (note that we tend to do this with war: there is broad agreement among Ds and Rs that generals should formulate war plans)

(2) constituents evaluate the effects of those policies on their lives; they do this via direct experience and via the input of cultural authorities —experts, even elites— whose opinions they'e evaluated and find trustworthy

(3) constituents remove governments that do not produce the results they want

Now: what's not working in the US?

From the point of view of someone who doesn't like Trump: politicians are not being rationally incentivized; modern campaigning, which takes place in a way unimaginable to the designers of this system, somehow breaks the process in which people remove bad governments? We can talk about 1000s of reasons why, but broadly this would be one big issue

A second: constituents are not "evaluating" in a way we think is reasonable. There are two issues here: (1) we dislike their values and think they are "objectively" wrong, which I'll return to in a moment; (2) we think they make analytical or evidentiary mistakes and listen to the wrong authorities; we think they're irrational

I think the second is extremely interesting and a richer example of what we're discussing. To start with —and forgive how insanely pretentious this all may be, but I love it!!!— what do we mean when we say values are objectively wrong? In general: we think that such values produce less happiness, less social success, worse societies; to say they're "worse" and "less successful" here means that given free choice over time, people will prefer society A to society B and will elect to live in it, making B worse

There are scores of indices: migration; standard of living; mortality rates; and more.

And we'd say: "Well, I expect my nation, in which we elect Sanders [or whomever], to have a better future than one which elects Trump, better in ways that only a fringe few would willingly elect to ignore."

People talk about the South this way: "We should have let them secede; they'd be a collapsing shithole right now."

And we know that if not this generation, then the next would say "fuck this; poverty sucks; oppression sucks; I'm out."

For some reason, people always hate unhappiness and privation!

I think you're talking specifically about "politicians" where I'm referring more broadly to "people with influence" and their responsibility (or lack thereof) to *serve* the masses as opposed to "guiding" them. I don't think either approach is "objective" but is there a world where we resign our subjectivity to one that is more "valid"? I mean we do it on different contexts; with our parents, our bosses, etc

Well I'm getting there!

lol I'm listening / reading

So: the problem here is that we have a mechanism that could reveal the "right" policies —the future— but decisions must be made now. And you can't persuade X, and X can't persuade you; and your values and authorities are different, even what you think life should be like! And yet: you share a single nation.

What makes all policy question harder in the US than elsewhere, by the way, is this diversity; when people talk about how swell Norway is I just can't believe it. Homogeneity is never difficult! This is a design fact of life; if all the elements share many similarities, design is pretty straightforward!

So how do we let a single people manage a nation in this way? Well, the original solution was: let many experiments run. Let Alabama find out what happens when they deny gays the right to marry, and their young people desert the state, and scores of talents like Tim Cook move to California and create $100B in value and 30,000 jobs, etc.

Meanwhile, let Colorado find out what happens when you make marijuana legal, and California what happens when you raise the minimum wage.

So —and to be clear, I don't think ours is working very well— federalism is an attempt to deal with the fact that no knowledge can be verified to be "objective" and no controllers can be taken to be "correct" or even "stably moral" at a design level

Instead of being prescriptive, set up a system that supports 13 individual community experiments, bound together for the purposes they share (defense), but free to fail (or succeed) and thereby persuade others (in either way) and foster progress

And this is a cool idea! I don't think we need to get into how it's broken —I think it ought to include some provision for federally-paid for relocation between states to make sure states suffer for being shitty— but that's a social system design.

In a world in which there can be no authority and no objectivity that persuades all —and "all" are your peers, "all" have the right to self-determination to the greatest extent possible; they will die, after all, and we can't take their right to life as they see it from them more than we must for the common good— solutions which maximize agency should be designed also to show which choices lead to the best outcomes; that's another thing: every piece of legislation should have to predict its effect and probably expire if that effect isn't achieved (unless re-upped by the legislators, etc.)

And anyway: the best way to persuade people is to show them that your design works and theirs doesn't. Testing is the only way, especially for things like "social policies," whose many effects are really unpredictable. It's not just that I think "no one can be said to be an authority"; it's that I myself, once I investigate an issue very thoroughly, am often amazed to discover that no one really knows what we ought to do except those who haven't thought very seriously about it

All easy solutions exist; all straightforward and obvious choices have been made!

I agree with you. It's in fact very easy to agree with that since it appeals to the populous. But it does seem like perhaps not the most efficient way to land at the best solution yet you're framing it as the only way.

It's also giving "the people" the power to decide what the best solution is. Are they (always) equipped to make that call?

Well: who gets to decide that? When and about which people? And what do we say to those "not equipped"? The biggest question of all, of course: *why do they not know they're not equipped?*

It's important to remember that people use others and authorities all the time, often for the most crucial parts of life; the most rabid Trump supporters use doctors, take their cars to mechanics, don't grow their own food, ride on airplanes, etc.

Anytime we wonder: "Why don't people just let the experts do this?!" we should also wonder "Why isn't their expertise apparent?!"

I would argue that it's because in general, governance is so fucking hard and so incapable of pure successes and so constrained by the multiplicity of the electorate and its demands (and by the checks of the system itself) that it's not at all clear to most people "who governs better"); no one is out there buying flip phones, but somehow they don't see the same superiority of value in e.g. Clinton

Ugh... high school never ends, huh?

My problem with that framing is that it puts the onus of proving expertise on the expert

It leaves no room for humility and it breeds boasting about one's expertise to be respected

it assumes no trust as it pits people against each other in a never-ending battle of proving their worth to one another

Oh well yo

It doesn't assume those things!

Whether people trust one another, authorities, knowledge from so-and-so or whatever source, is entirely down to cultural, community, and individual values, and that "mix" is not really controllable except by influence (again: influence!)

The question is whether we should coerce one another; we can only really do so in instances of *extremely* high confidence about our rightness, which we usually only claim to have about issues that are like at +95% agreement (e.g. no premeditated murder of innocents lol)

But humility is a value; trust and openness are values. How do we make values spread? To use our industry terms, in my view that's a content question and not a design question.

Indeed: what I like about design is this meta-relationship to the culture scrum; I like trying to foster a better cultural scrum —e.g. one in which more people find more of the information they seek to understand their world— but I don't actually like being involved in it

So for example: I think art is more important than technology for the creation and influence of values, far more.

I don't mean gallery art, obviously; I mean all of the representations in which the culture works itself out, sees itself, experiences and expresses itself. Video games, novels, movies, songs, and these days: essays, essays, essays, and more essays.

Essays are a'boomin'?

I like working for Facebook in part because people read more now than they ever have, and of a wider array of sources than they ever have. That to me is a very likely good way to positively influence humanity's development; I could be wrong about everything, but getting more people exposed to more information from more sources than before will almost certainly not be bad

Information ≠ Knowledge

No, information is not knowledge!

But it's the same principle as before: some information is, some isn't knowledge; the knowledge will "serve" its user better over time (in living, making choices, being happy, understanding life, etc.), or the citizens who have more knowledge will serve their communities better and the communities will outperform, such that over time people do tend to improve their knowledge, especially when it serves them to do so.

In any event, if a % of information is knowledge, one way to increase knowledge would be to increase the intake of information generally. And we return to the main theme: the question of "what is information vs. knowledge" comes back to values and other matters we cannot coerce one another about

Again, I generally agree but I wonder if "more information" just adds noise and cynicism to the way we consume and think about the world

I think it definitely does too

It's a huge bummer to me

Some possibilities among others to consider there: just as we hope that increasing the quantity of information will increase the exposure to knowledge, there may also be a % of information that is "bad" or whatever, or too much information may be psychologically problematic, or whatever, and the total effect will be (a) so bad it offsets the benefits or (b) bad, but not bad enough to offset benefits; or we could adapt to this, too, and e.g. a "norm" could evolve that it's "not cool" or "not polite" to write about your political opinions online, etc. Or: it could just be our perception, the result of seeing more reality, etc.

My point here being just this: if it's a problem, someone will make money solving it, or get elected solving it, or start a new country solving it

Isn't that putting too much faith in the status quo? What about revolutions?!

Someone can lead a revolution solving it!


So, you're saying: the world will run its course in the direction the majority will take it and your individual contribution to it is only as impactful as what the majority recognizes as impact?

Well, no! There are ways to have huge individual impact!!

But let's be reasonable, it's true that it's hard to have huge impact. Highly influential works of art or culture or scholarship can change a country or even a civilization's trajectory through opinions; designing a system which increases the rate of information flow among 1.6B —or rather, designing some aspect of that system— might have a similar effect.

It's not easy! But there are other ways: you can be a scientist and discover a foundational physical principle (lol), or cure cancer; or you can invent something awesome that accomplishes X, Y, Z, maybe reduces infant morality by .X% or whatever, etc. etc.

I think there are a lot of ways; you might be the person to start a sit-in that becomes a seminal moment in a nation's history. I think it's pretty hard to *intend* to do something of that nature and do it; I think chance has a lot to do with it.

I'm not sure I'd put "sharing more cat selfies" along with "curing cancer"

Well but dude you know this stuff is! It's "more cat selfies get grandmother online; her sister in the old country gets online to share and see them too, sees posts from niece about how she thinks of feminism," etc. etc. Or a billion permutations, 90% of which bring only local effects —but happier people are a happier world!— but some of which are huge. How many children will use Facebook first, maybe via, and will learn to X, Y, Z, raise their family's standard of living, etc. etc. I mean, you know I think our jobs are jokes or w/e but what's consequential is unpredictable and diffuse, and downstream from what we can target

And to my previous point, because the world moves at the pace of its majority —do you think it's inevitable for this happen?

No, it's not inevitable. Humanity could absolutely make a local error that destroys everything; while I'm not concerned that climate change is this error, I think it's inevitable that humanity will engage in a very serious nuclear war (I can't imagine anyone would disagree). A severe enough such war could kill us all.

And a million other things could wipe us out, too. I want us to make as much progress as fast as possible, and I think as long as we're around to try to achieve better lives, we will progress, but we could fuck it up horribly and even fatally too. I hope we don't!

Why does it matter if you do it or someone else?

It matters to me for purely personal reasons, for affective reasons: I have never done anything (would this really be something?!); I have had no effect on the world, and for some arbitrary reason that seems "fun" and "good"; vainly, I desire to make my own contribution, something that will reinforce my immortal sense of specialness (within my much more robust insecurity and self-loathing) maybe?! Who knows, but it's not a powerful drive honestly. I'd like it; maybe I'd love it. But I don't think I'll mind it not happening at all, either.

But in terms of my beliefs about humanity: me being at Facebook is about as important as Art Stevens working in the plant that made the first piano George Gershwin played on, which is to say: indeterminately important, but probably not crucial.

The train of influence there would run through *all* of music; maybe if Gershwin doesn't popularize some of the sounds he does, music evolves differently, stays "white" among teens, and a vital channel of inner-city expression, one which has made the black experience absolutely vital to a huge majority of young Americans, is lost, etc.?

I wonder if Art Stevens had any ambitions. Did he wonder the things we wonder? Or did he just care about the craft and nothing else?

I think everyone wonders about these things, or nearly everyone, at different times in their lives; you and I wonder about them all the time, but probably hella neglect other shit lol


it's been a while since I called my mom

Read by 2,014 others

Mills Baker


Mills is a Product Designer at Facebook who likes to write about literature, design, art, technology, love, philosophy, memory, history, and other nonsense.

About This Site

Talk Turkey is an attempt to have authentic conversations in public. Gabriel Valdivia interviews people over instant message and posts the conversations insitu on this site. Charlie Deets writes the code.

This website is not affiliated with, maintained, authorized, endorsed or sponsored by any real messaging service.