you can just do stuff
arpa and xerox parc, agency, stoicism, a set of guidelines for ambitious endeavors
a friend of mine posted a funny tweet:
I love the image it conjures of this classroom, full of mildly confused kids, looking at each other, unsure, until someone decides that it's not some kind of joke and they actually can just do what they want. it also made me think of two other silly stories with similar themes of getting the mildest taste of empowerment without being trained or told to take it
one of my first managers when I first got into tech told me a story about a service outage he was party to earlier in his career. the servers were down, nothing was available. no amount of troubleshooting had successfully resolved the issue, or even offered any leads. this was a company with service-level agreements, so real customers and real money were on the line. him and his partner in the server room had no idea what to do
then, for some reason he couldn't explain, he realized it would be fixed with a hard reboot. but there was also the possibility that the services might not come back on at all
in his telling, there was hesitation, as if he was waiting on direction or permission to proceed. but he realized that, ultimately, he was responsible. it was his decision
so he did it, and it worked
in my early 20s, in an act of youthful folly, I decided to go to college. I was in english for awhile, then had a small existential crisis over the nature of meaning, and decided to switch to math. none of this ended up working out for me; eventually, I was kicked out for money reasons, and it took a couple years' remove until I learned how to think on my own again
one of my favorite classes, though, was introduction to mathematical proofs. most of my other classes felt rote, stifling, and none of this was helped by the fact that I never opened the books and expected to coast through tests based on the subset of lectures I actually attended
proofs was different because it was all fun little puzzles. nothing hard, just proving facts about odds and evens, induction, mostly a set of exercises to get a feel for systems of formal logic. this was years before I started programming and found, in that, a much better medium for amusing the autistic part of my brain with fiddly trifles
one day in class, the door was propped open, and someone, returning from the bathroom, closed it. this ended up posing a challenge at the end of the class, when we discovered the door had been propped open because it didn't have a handle
a crowd of students formed around the door, banging and yelling trying to get the attention of someone outside, while the professor reclined behind his desk with the satisfied smile of a man who's getting paid regardless of the outcome. I was a shy kid at the time, and interacted with virtually no one in school, so I silently made my way to the door and looked through the hole where the handle should go, without giving any indication what I was doing. as far as I could tell, the missing handle was nothing more than an ergonomic square peg for a square hole
so I want to the professor's desk and started rifling around in it while he regarded me with idle curiosity. I found a pair of scissors, shuffled back through the crowd, put the tip of one blade in the hole, and opened the door
I love these stories as tiny little microcosms of agency. none of them mean a whole lot. nothing done in them required any special talent or novel insight. none really affected the world in any meaningful way. just some kid learning by accident that things can happen because you decided to do them
When you grow up you tend to get told the world is the way it is and your life is just to live your life inside the world. Try not to bash into the walls too much. Try to have a nice family, have fun, save a little money.
That's a very limited life. Life can be much broader once you discover one simple fact: everything around you that you call life was made up by people that were no smarter than you. And you can change it, you can influence it, you can build your own things that other people can use.
Once you learn that, you'll never be the same again.
one reason I bother to write at all is because I like to encourage people to exhibit more agency. I love the image of a world full of competing, empowered individuals. I feel that one of the highest leverage things I can do is contribute, in my own small way, to the people around me feeling capable of doing more of what they want, not feeling constrained or restricted. I think freedom is something you take with force, and the people who want it should take it
not everyone wants it, and that's ok. I used to feel superior to the people who just want the nice family, some fun, a bit of money. but that feeling was a bit insecurity, and a bit envy. I hope they get what they want, I hope they find systems that will take care of them and protect them, I mean this with all affection
ambition is probably a disease. I don't know if there's an inborn fire, or if it's something that's forged into you by pain early in life. but it's probably something wrong, deficient. there's a hole inside you you're trying to fill up and nothing is ever good enough to fix it
but it's also essential that some people have it, to improve the lot of everyone. mitochondria were originally parasites. things don't have to be healthy to be useful
I recently read a wonderful position paper written by dominic cummings in 2018, titled "on the arpa/parc 'dream machine,' science funding, high performance, and uk national strategy." cummings has some interesting things to say about entrenched bureaucracy and coordination failure, along with a lovely vision of british technofuturism driven top-down by an uber-competent post-brexit government, which, in 2022, reads like a report from the twilight zone. but the first half of the paper is a concise set of takeaways that explain why and how the epynomous research labs actually accomplished the magic they did
I knew the fruits of arpa and parc, as do most of us I think. but I didn't know much about the people themselves and the way they worked together, beyond the most basic facts. I'd probably start reading the dream machine, just on the strength of this paper. but I'm writing this on an airplane that that doesn't have wifi. so if you'll indulge a summary of a summary...
arpa was set up in the wake of the sputnik shock, mostly working on weapons systems, but the powers that be requested some small fraction of the org's resources be put into computing. the head of the org recruited joseph licklider and granted him a $10m/yr budget, which he disbursed as freely as he could get away with to the best people he could find. licklider did this to advance his vision of the future of computing, which still has not yet been realized: a presciently transhuman idea of computers not as mere tools we use with our hands, but as physical extensions of our minds. the next stage in human evolution
as cummings articulates it, licklider cultivated a research community of brilliant scientists and engineers, provided them with generous funding and a grand vision, and let them work toward accomplishing it however they pleased. the result was englebart's mother of all demos, and the internet. one of licklider's successors, robert taylor, went on after arpa to lead the computer science division of xerox parc. and from parc came everything: the pc, the gui, the bitmap display, word processing, object-oriented programming, laser printing, and ethernet
cummings's goal in writing this is to lay out a programme that he hopes (hoped?) the uk could leverage to produce the kind of world-changing research that america did in the mid-20th century. but the advice is general-purpose and should be of interest to anyone putting together their own hermetic research groups, fringe engineering collectives, or technology-oriented cults. I've condensed his points down to seven:
vision: licklider laid out a singular idea of the future which kept everyone aligned without exerting any control over their day-to-day activities. various researchers at parc assert they did the best work of their lives there, even though many continued to work together in other companies and labs after. they can't really articulate why parc was so special without delving into mysticism, which I think makes the case even stronger. alan kay ascribed it to the vision, which he called "a magnetic field." "we were all kind of pointing at north." when asked what taylor did to keep everyone producing such quality work, butler lampson said "damned if I know. it's just magic"
intensity: every one of the key people at parc felt that it was their best work, but also their most strenuous. people worked around the clock, slept in the office, neglected personal relationships. they sacrificed. cummings cites richard hamming arguing that effort compounds, so putting in a bit more work and thought at the margin delivers massively outsize rewards on a long timescale
most people probably do not ever want to work like this. and no one can sustain it forever. but here and elsewhere, cummings essentially argues from the tail. n-sigma people putting in n-sigma effort to get n-sigma results
people: several points dovetail into one theme, that the goal of everything is to fund top talent, connect them with their peers of the same caliber, and get out of their way. researchers would be granted budgets to spend as they desired, and they were free to work on anything they pleased, in line with the overall vision. people didn't care about credit, so they freely collaborated. good ideas flourished and bad ideas withered, but ideas had time to develop or die naturally
"top talent" is key: great people want to work with great people, and one average performer can drag down the entire group. licklider and taylor were both famous for their instinctive ability to spot the best. researchers at parc were given power in the hiring process, and low performers were regularly eased out of the org. and aside from team dynamics, there is another argument for this from the tail: a single top performer can do more than entire teams of those who are merely good
lawlessness: an antonym for bureaucracy. related to no micromanagement or deadlines, bureaucracy of any kind is a cancer. quite literally, in that it becomes an organism of its own, sapping resources from its host, and growing to consume it until nothing is left. half of taylor's job was fighting constant rearguard action to keep xerox out of parc, and once the firewall was breached, parc died. licklider got away with things in his day which would put you in jail now, such as directly soliciting procurement proposals, and even helping write the grants that he intended to fund. all wildly illegal under anti-corruption statutes
the difficulty is, all these rules inevitably exacerbate the very problems they intend to solve. the rules become a system to game. people who just want to do the work get discouraged and leave, while people skilled in gaming systems show up and work them for all they can get. it comes back to the previous points: the solution is to find the people who will do the right thing on their own, because they are part of the team, believe in the vision, and are deeply devoted to doing the work
speed: the meeting which led to arpanet took 30 minutes. the originator of the memo that led to the alto speculated he could build it in three months. speed is motivating, dynamism is exciting, the best people want to move fast. and, continuing the theme of the synergies between these points, speed prevents the accumulation of process. a rolling stone gathers no moss
failure: in kay's words, it's baseball, not golf. your organization needs to be able to tolerate failure. perhaps even a majority of projects will fail. that's the nature of experimentalism, and organizational fear of failure can only hold back people who want to bet big, instead selecting for middling talents
purity: successful research is almost never done in pursuit of a business objective. otherwise, you wind up with timetables and micromanagement, even if they aren't explicit. people need to be free to pursue what is interesting and promising to them
xerox, texas instruments, and at&t, among others, all failed to commercialize the groundbreaking work done inside their walls, but not because they failed to motivate the researchers to focus on commercial ends. if they'd tried, that research might not have happened at all. the proper role for product, as it relates to pure research, is to sift through the material the researchers produce, looking for rough diamonds to polish and bring to market
a related problem is letting funders try to pick winners on the basis of merit, rather than purely investing in talent. demanding particular, knowable outcomes is fine for business. but for research, it constrains the time horizon and locks you into incrementalism, choking out speculative work
one thing that's interesting is many of these ideals have permeated startup culture. I could easily rephrase the point about great people to be "A-players want to work with A-players," speed and lawlessness as "move fast and break things," failure as "swing for the fences" and "maximize at-bats," and various other venture capital truisms. but the startup ecosystem—that is, the glut of companies produced by the post-gfc infinite money assembly line—has not produced outcomes comparable to arpa and parc
the two points that startups invariably fail on are "vision" and "purity"
vision is, of course, something they all aspire to, but few live up to it. there's an expectation that every hr software suite or on-demand pet food delivery service has to articulate why it's a groundbreaking attempt to reshape modern living. at least ai and crypto are trying to break major parts of society. marketplace apps and b2b saas are mostly about extracting rents
purity is the most glaring mismatch, and in this case, no one can really be faulted. venture capital exists to fund businesses, not research. if good research should come out of a venture-backed startup, I'm sure everyone will be very happy about it. but at the end of the day, venture doesn't want to be exposed to technology risk when market risk and founder risk are already so great
if academia was a reliable source of rough diamonds, there wouldn't even be a problem. the optimum of "high impact" and "good business" would be something like the polishing: letting research happen in a university setting, and funding its commercialization. we all wish there were a thousand genentechs, but in the end, there was just one
I do think it's true that these principles are often used by middling companies to extract excessive labor toward middling ends from people who don't know any better. but I also think it's true that some people crave the struggle, some even need it, and sometimes the ends are worth it. this is an argument from the tail as well
typical people shouldn't be forced into making extreme sacrifices just to survive, and no one should make them for typical companies. 99.9% of the time it's a bullshit deal. I probably wouldn't sleep on the floor at twitter. I'd give almost anything to sleep on the floor at parc
anyway, I'm not writing all this to propose solutions to science funding. stripe press republished a book by donald braben on the topic called scientific freedom: the elixir of civilization, and while I think his solutions would recreate the problem anew, his articulation of the problem is sufficient. I want to read hamming and bush and dream machine, and maybe write more on this topic when I have more developed thoughts. lots of people are thinking about this problem though, and I think they are better suited to tackle it than me
instead, I'm writing this to describe a simple blueprint of how one might have impact, as an individual looking for a community in which they can do real work
articulate a vision, or find someone else who has laid one out that you can believe in. find the best people and become close to them. challenge each other to keep up, be the best possible versions of yourselves. work hard, work fast, pour as much of yourself into the work as you are able to. disregard rules as much as you can without seriously jeopardizing your life or liberty. and don't worry about a practical outcome. just aim to do the biggest, most important thing you can
I don't put any faith in institutions and don't expect committees and policy papers to solve any of our problems. "process" is for value-protection, and it eventually eats the value it purports to guard. to create value, I believe in people doing what they want, especially what they're not supposed to do, attracting friends and allies to their cause, and inscribing their wills on the earth. I think our institutions have failed, and it falls to the unaffiliated to push us forward again, people acting toward their own ends
xerox and the department of defense were "process." licklider and taylor were men
one of the books most crucial to my development, and still one of the dearest to my heart, was franny and zooey. catcher never really spoke to me, maybe because I first read it in school, maybe because I found holden's style of naive immaturity grating. but franny's style of naive immaturity spoke to me very deeply
the title characters are the youngest pair of a line of siblings, all of whom were precocious as children, touted on radio as lovable savants, and then left to develop their own unique intellectual maladies as adults. the book is haunted by seymour, the oldest of the children, who went on a ballistic trajectory from academic genius to spiritual guru to dead by his own hand. the next child, buddy, never tried to live up to the example set by his older brother, and settled into reclusion and mediocrity. these two eldest of the siblings raised the two youngest on a diet of religion and mysticism, hoping to spare them the alienation of the ivory tower, but only ended up forcing on them alienation of a more twisted kind
the two stories focus on franny, in the midst of a nervous breakdown. her boyfriend lane and brother zooey both blame an esoteric religious text and her half-committed attempts at using the jesus prayer as a monk would, lane stupidly but zooey incisively. zooey savages her for practicing wrongly, and selfishly, trying to make a point to her about purity of motive. he stops when, mortified, he realizes all he's really done is emotionally break her. the story ends with a saccharine point about spiritual oneness and the omnipresence of the divine
having read salinger's other stories, particularly how he deals with the topic of enlightenment, the impression I've always got is that he does think franny is wrong. but I love franny, I empathized with her from the start, sympathize with her now, and I always felt she was right. maybe not correct in her approach, but justified. the things that she feels, I felt, and if religion is a way out of the alienation and the pain, then it's her right to use it however she needs
franny and zooey are both lonely and judgmental. they both feel superior to others, but frustrated that everyone focuses on the wrong things. they're better than other people by other people's standards, but also hate these standards, feel trapped by them, and have no way of explaining this to anyone but each other
the main difference between the two is zooey lets loose and terrorizes the people around him, while franny suppresses, does what she's supposed to, and ends up shattering from the stress. franny isn't really breaking down because of the jesus prayer. it's a salve, a distraction. a possibility of escape. franny is breaking down because she's torn between acting the way that everyone expects her to—the way she tries to force herself to behave, for the sake of appearances, for the sake of others—and actually letting out her true feelings
but underneath that, for both of them, is loneliness. "they made us into freaks," zooey asserts, about their brothers, but he isn't sure how much he really blames them. but something happened, whatever it was. inborn cleverness, public performance, spiritual training, something separated them from the mass of mankind. and now they're cut off, adrift. zooey lashes outward and franny crumples inward because they fundamentally cannot relate with other human beings
franny was a perfect totem for my own late teenage angst. I grew up used to the feeling of being ignored when I did good but punished when I did wrong. something flipped in adolescence and I went from silently suffering it to doing wrong as much as I possibly could. I wanted to show that I couldn't be controlled, that no one could actually hurt me, and, predictably, it only made things worse
I remember when I was in the process of failing out of high school. I was sent to a psychologist to figure out what was wrong with me, someone meant to coach me back onto the proper path. I had "so much potential" that I was "wasting," and that was the only thing that seemed to matter to people
it's so funny, in retrospect, how impossible it was for him and I to communicate. I legitimately tried to explain to him what was wrong. everything felt empty and hollow, life was pointless, society was meaningless, and no one else seemed to understand this. like there was no one on earth who could relate with this feeling of being totally cut off from everyone and everything. he responded by showing me ged practice tests and reassuring me that it wasn't too late to have a career
but aside from emotional reassurance that someone, somewhere felt crushed by the world like I did, even if she didn't exist, the other thing I got from franny was a reading list. I'd always been a stubbornly materialist nerd type, and while I loved literature, I couldn't see any substance in the world beyond the physical. it took me another decade after that to really start to sense it. but because of franny, I could at least learn. I read epictetus, I read sri ramakrishna, I read the gospels and the gita
epictetus really got to me. stoicism has, from its inception, always been a way to cope with an intolerable and inescapable reality. it's not a healthy stance, and I am never quite sure what to make of comfortable people in control of their lives who adopt it because they see it as a form of rugged masculinity. stoicism is for people who can't change their lot and must learn to live with the pain. aurelius couldn't quit being emperor, and epictetus couldn't quit being a slave
stoicism and the gita got me through a lot as I moved out on my own and had to figure out how to actually survive. I gave them up after I no longer needed them, but I still thought about them a lot. the thing that really bothered me about stoicism in particular, though, was its intense bias against action. maybe epictetus takes this further than others, I don't actually know, but at least to him, the only thing you can control is your own internal state. if you lose your riches, if you lose your freedom, if everyone you love is scoured from the earth, there's nothing you can do to stop it. the only thing you can do is accept it
it appealed when I was helpless, but once I felt like I had some control over my life, I grew to despise it. how can you be so weak, to let those things happen to you?
I knew he was right, in a sense. senseless things happen, things you can do nothing about. but how can you accept it? you have to fight, you have to win. if you try hard enough, you can defeat reality itself
stoicism is cope, but this attitude is cope too. scraping and scrambling for power to make up for the lack of it you once felt. the fear you have that it may still elude you, despite all your efforts. this feeling held immense sway over me much more recently, but I'm more equanimous about these things now
he's right, in a sense, that it's all fleeting, and ultimately you need to learn to live with loss. but I do think he makes an error, and his error is treating a continuous thing as if it were discrete
you can't control reality with perfect fidelity. but you can influence it. if you believe his essential point that the world is full of things you have power over, and things to which you must acquiesce, there is another answer than to passively accept your lot. it's to go out and find the longest lever you can
Thanks for reading alice maz! Subscribe for free to receive new posts and support my work.
I would have said this without qualification a few years ago, but I do wonder whether openai is in this league now. however, I'm not sure it invalidates my point, considering that openai started as a research org before turning into a startup
thanks for writing this- a good reminder at the right time.
I spent a good chunk of the last few years in academia, almost certainly too long. first it was fun, and there were always some good bits, but being surrounded by systems that are pointless- where everyone knows they're pointless, but does them anyway- rots your soul. it can become the way things are for a chunk of your brain, before you realize it.
it's odd- that sense of agency, that internal locus of control, is incredibly powerful, but also so fragile. things can erode it or damage it very directly. something you (or at least I) have to maintain, and cultivate, and fight entropy for.
Very cool and awesome post 👍