Skip to content

Old school

Maciej Cegłowski writes:

About two years ago, the Lisp programmer and dot-com millionaire Paul Graham wrote an essay entitled Hackers and Painters, in which he argues that his approach to computer programming is better described by analogies to the visual arts than by the phrase “computer science”.

When this essay came out, I was working as a computer programmer, and since I had also spent a few years as a full-time oil painter, everybody who read the article and knew me sent along the hyperlink. I didn’t particularly enjoy the essay . . . but it didn’t seem like anything worth getting worked up about. Just another programmer writing about what made him tick. . . .

But the emailed links continued, and over the next two years Paul Graham steadily ramped up his output while moving definitively away from subjects he had expertise in (like Lisp) to topics like education, essay writing, history, and of course painting. Sometime last year I noticed he had started making bank from an actual print book of collected essays, titled (of course) “Hackers and Painters”. I felt it was time for me to step up.

So let me say it simply – hackers are nothing like painters.

Cegłowski continues:

It’s surprisingly hard to pin Paul Graham down on the nature of the special bond he thinks hobbyist programmers and painters share . . . The closest he comes to a clear thesis statement is at the beginning “Hackers and Painters”:

[O]f all the different types of people I’ve known, hackers and painters are among the most alike. What hackers and painters have in common is that they’re both makers.

To which I’d add, what hackers and painters don’t have in common is everything else.

Ouch. Cegłowski continues:

The fatuousness of the parallel becomes obvious if you think for five seconds about what computer programmers and painters actually do.

– Computer programmers cause a machine to perform a sequence of transformations on electronically stored data.

– Painters apply colored goo to cloth using animal hairs tied to a stick.

It is true that both painters and programmers make things, just like a pastry chef makes a wedding cake, or a chicken makes an egg. But nothing about what they make, the purposes it serves, or how they go about doing it is in any way similar.

Start with purpose. With the exception of art software projects (which I don’t believe Graham has in mind here) all computer programs are designed to accomplish some kind of task. Even the most elegant of computer programs, in order to be considered a program, has to compile and run . . .

The only objective constraint a painter has is making sure the paint physically stays on the canvas . . .

Why does Graham bring up painting at all in his essay? Most obviously, because Graham likes to paint, and it’s natural for us to find connections between different things we like to do. But there’s more to it: also, as Cegłowski discusses, painting has a certain street-cred (he talks about it in terms of what can “get you laid,” but I think it’s more general than that). So if someone says that what he does is kinda like painting, I do think that part of this is an attempt to share in the social status that art has.

Cegłowski’s post is from 2005, and it’s “early blogging” in so many ways, from the length and tone, to the references to old-school internet gurus such as Paul Graham and Eric Raymond, to the occasional lapses in judgment. (In this particular example, I get off Cegłowski’s train when he goes on about Godel, Escher, Bach, a book that I positively hate, not so much for itself as for how overrated it was.)

Old-school blogging. Good stuff.


  1. D Kane says:

    > old-school internet gurus such as Paul Graham

    Isn’t Paul Graham still an internet guru? Y Combinator, the company he founded, and which he still plays a leading role in, continues to be (probably?) the single most important institution in start-ups and/or venture capital. One can plausibly claim that Raymond is much less important/influential today than he was a decade ago, when he was writing books like The Cathedral and the Bazaar. Graham is, if anything, even more important today (and 10 or 100 times richer) than he was then.

    • I don’t think Paul Graham has had any real intellectual contribution to public discourse, sure he wields a lot of money… But people aren’t talking about his ideas in this way. Same for Phil Greenspun for example

      • Rahul says:

        Don’t you wield a lot of influence, when you have a lot of money?

        Most people aspire to make a lot of money, so when they come across someone who has, they naturally assume he knows something they would like to know?

        • Well can you point to a bunch of people discussing Paul Graham’s ideas about society or computer programming or politics or that sort of thing between say 2013 and 2018? That’s what I mean about influencing public discourse. In the early 2000s Graham wrote a bunch of essays and people would respond to them with counter essays like this. I think these days Graham’s influence is to just allocate money to startups doing stuff. So if you consider “Dropbox” an intellectual achievement of society… Then sure, but to me that’s like calling “southern California Edison” a major intellectual triumph…

          • D Kane says:

            “Point to?” No. But if you don’t think that Y Combinator in general and Paul Graham’s essays in particular have been hugely influential in SV/SF start-up culture over the last decade, then I doubt your knowledge of this topic. Whether or not that influence is the same thing as being an “internet guru” is a matter of definitions.

        • Martha (Smith) says:

          “Most people aspire to make a lot of money”

          Is that really true? Is there evidence supporting that statement? (I’m not disputing it, just wondering, because it struck me a surprising.)

  2. >About two years ago

    When I read this I immediately said “wait that was about two DECADES ago. Then I noticed you were on a flashback trip…

    Heady days those early blogging times. People used to have stuff to say and not just Tweet about how much sucks.

  3. Typhon says:

    Be sure to read Cegłowski’s “A Rocket to Nowhere” and “A hundred Years of Turbulence” as well. He’s one of the best and most entertaining internet writers I know.

  4. Daniel Weissman says:

    I love Cegłowski’s blog! Looks like the link may be missing though? It’s if anyone wants to check out the rest of his stuff. And if anyone has recommendations for folks who like this blog, that one, and Shalizi’s and would like more in the same vein, I’d love to hear them.

  5. Jonathan says:

    Or it’s that some people do feel a certain closeness between parts of their own mind. The long-standing association of math to music is often brought up as though the mathematics of pitch and the division of music into some form of counted time somehow translates into JSBach, to which I’d say ‘have you ever head a vacuum cleaner, ‘cause that’s also noise – typically a B-flat – and it sounds like crap’. Or why do so many designed products produce minor intervals; you’d think that if math and music were so closely related, they might try to make it sound less sad, more major, more affirmative, especially since it’s their product and you’d think they’d want you to feel good being near it. But no … because there is no real connection between math and music, just an affinity that some people have in their own heads.

    I draw and paint. There are affinities to coding and similar work. 1: it requires arranging discrete pieces so they function as part of a whole. 2: it requires a lot of brush work just to get through a section. 3: efficiency matters a lot, a whole lot. See Picasso ‘rose period’ for an example of the last: he reduces emotions to gestures efficiently instead of burying them within the traditional painting form he was trained in (and which he never was very good at). But there are many parts of drawing and painting that have nothing much to do with coding, notably that you rarely code without knowing where you’re going but art is often made by trusting intuitions of line and shading, and it’s only after some time that you realize ‘oh, that’s what I did.’

    But to dismiss similarity because one uses a keyboard and the other a brush or knife (or fingers) is shockingly limited in imagination. Computer programming is an abstraction of what the machine actually does, so what I’m doing by moving my fingers appears on this screen which appears on a website that people access and all that is a creative structure built by people. One could say it’s a form of public art: you aren’t making money off this blog, though I suppose it may be professionally valuable to you. But I don’t go there because then you start wondering about the reality of reality and I may get an acid flashback. The idea that programming is art is embedded in the concept that a painting of a person is not the actual person but a rendition that conveys messages. If you can’t see the similarity, there’s something wrong with you.

    • Thank you for this comment. Admittedly (I’m quoting James Merrill with this word), your comparisons are more compelling than the one that Cegłowski quotes from Graham. Being “both makers” doesn’t cut it; “arranging discrete pieces so they function as part of a whole” does, in combination with your other points.

    • Keith O'Rourke says:

      Imagination is likely a key component for anyone excelling at anything but it may be for very different purposes.

      And “attempt to share in the social status that art has” is what I believe a lot of the push-back I got when trying to meaningfully interact with artists and qualitative researchers – that’s not really creative and your imagination is largely underdeveloped. ( For instance the my comment before this post was posted – )

      Sort of the same as when you ask a question/make a comment at a conference – in essence claiming to have an important idea and being smart and the presenter responds with that not really an important i and I doubt you are in fact smart.

      • I think imagination is one aspect of what differentiates people who find something like machine learning or pushbutton SAS type statistics attractive vs people who find Bayesian statistics and Stan attractive.

        The Bayesian framework gives you a means to evaluate what data means for the content of your model, and it does this regardless of what the model looks like (modulo computational difficulties). The creative content of a statistical analysis in Bayes is *the choice of likelihood/model* and the prior itself expresses opinions about what numerical values of parameters are within the realm of reasonable given what your model is intended to express about the world. A likelihood function Distance = Speed * Time is not by itself meaningful until you start to say things like “Speed of a real world vehicle of the type we’re discussing is somewhere between a few miles per hour and a few hundred miles per hour”. If you wind up with some other values fitting your data well, it’s because you have a bad model of reality… like perhaps there’s a fixed delay in starting the time meter that makes it seem like long distances are being traveled in nearly zero time.

        Machine learning / pushbutton SAS methods are more appealing to people who *have no interest in a creative mathematical modeling process*. And sometimes that’s fine. It’s ok not to care why in July you should mark down the price of bathrobes if you want to maximize revenue, or why people from Tennessee consistently spend more on transportation in October than people from Arkansas or whatever. Sometimes you just want a summary of facts in the form of a formula with some errors, so you don’t have to SQL select a billion rows of data and take an average.

        In many areas where statistics is applied, such as Psychology or Biology, the mathematical backgrounds of the scientists is such that they really aren’t creative model-builders, and they really do want a certainty factory. The fact that certainty factories don’t exist even though about 70 years of stats textbooks pretend they do is a real detriment to these sciences. It let people pretend that they didn’t need to do creative modeling and led to terrible scientific models like “power pose” or “stereotype threat” or whatever. Qualitative ideas with no real quantitative power to predict, yet sufficient p-value baloney to declare success. The same seems to be true of ideas like “cancer genes” the idea that some genes are more frequently expressed in cancers is not the same thing as having a mechanistic model of how cancer comes about.

        • Rahul says:

          Graham thinks hacking and painting go together.

          You think good imagination and Bayesian thinking goes together. Isn’t that somewhat similar?

          • I actually do think hacking and painting go together. But you have to understand what kind of hacking Graham is talking about. He’s the kind of guy who would create a whole language to describe problems of the type he wants to solve. A little like Picasso inventing cubism or the impressionists creating a new style. There used to be a special status that the word “hacking” had, meaning a kind of tinkering that connected otherwise disperate areas. For example the MIT tech model railroad club connected control systems and communications systems to model trains so they could achieve cool coordinated motions… There’s a lot of artistic open ended boundary pushing in that kind of activity.

            I disagree with Cegłowski in that while the actions of hackers and painters are both very disperate the motivations to create cool things, and the role of the aesthetics of the final product are very similar, at least when you understand the term hacker in the right cultural context.

            Interesting that you mention engineering and sailing, because a lot of aeronautical engineers enjoy sailing, the beauty of harnessing the wind to extract energy is very likely the attraction.

            • Also people creating something like openchartplotter or openstreetmap have much more hacker cred than say goodle maps. The point is they are creating ways for society to cooperate. They focus on creating public goods: knowledge, shared experiences, etc. That’s a lot like say early burning man or Banksy, or the like. It’s very different from “if we give away this commercial service we can glean information about people and sell it to the highest bidder”. That doesn’t drive creativity in the way that people using their own “hacks” does. When the consumer and the producer are the same person you wind up with hacking, or artistic painting… It’s rare for someone to get famous for designing cereal boxes as a commercial artist, or software for Gmail. But we do know say Steve Wozniak or Richard Stallman or Claude Shannon or Richard Hamming.

              • What you might be able to say is that Cegłowski conflates the term “hacker” and “coder”, whereas Graham’s “hacker” is specifically talking about people with a creative drive to create new technologies usually for aesthetic or creative or social reason. For example the guys who work on Blender, or POVRay or Signal or OpenWRT or OpenStreetMaps or OpenSIPS/FreeSWITCH or the R core language or the Julia language or the ANTLR grammar generator or whatever, those are hackers.

                There’s a big difference in motivation, aesthetics, purpose, etc between the hackers that work on Stan, and the army of hired programmers keeping SAS’s revenue stream coming in.

              • Anonymous says:

                Little known fact: SAS is short for SAtan’s Software.

            • Obviously I have some thoughts on this topic :-) actually I wanted to say my blog is hosted on my vanity domain: I registered because I thought that the action of contributing to public discourse on the internet was a lot like painting murals or chalk paintings on the street. The point is to have passers-by interact with the work, without attempting to control the venue (the internet itself). The comparison to something like Facebook or Google Plus or Twitter is that those are more like gallery curated walled gardens: to participate you agree to a terms of service, and apply for an account and the provider can choose to censor you and also the commenters… there are limits to the kinds of material you can put up, the gallery owners will promote certain things based on their commercial interests (Facebook’s “Most Relevant” or “recommended for you” or whatever)

              I personally think that is basically like Gertrude Stein’s salon for statistics and social science, and I value a lot the individual creative contributions of the regular contributors. I thought to call out a few by name here but there are too many.

    • Martha (Smith) says:

      Jonathan: Nice response. I had come up with your point 1, but appreciate the fuller discussion

  6. Rahul says:

    Often someone likes to do two disparate avocations, one of which is perceived by society as “cooler”, hipper, liberal, creative or some such and the other not so much. e.g. Hacking vs Painting.

    At that point I feel there’s a natural tendency to delude oneself that those two are somehow kindred. It’s some kind of imagined transference from the cooler hobby to the more mundane one.

    And hence the belief that e.g. there must be some shared attributes in me that make me a both a good engineer and a good sailor.

    • Andrew says:


      Which do you think society perceives as cooler, hipper, etc.: hacking or painting?

    • Jeff says:

      This isn’t exactly the same but this conversation reminds me of “Leaving the Atocha Station,” by Ben Lerner. The narrator, an American poet living in Madrid, confesses that he finds reading others’ poetry very difficult. He also has limited Spanish proficiency, but describes a transference when reading poetry in Spanish: “My inability to grasp or be grasped by the poem in Spanish so resembled my inability to grasp or be grasped by the poem in English that I felt, in this respect, like a native speaker.”

  7. Wayne says:

    I’ve heard — perhaps an urban legend — that IBM looked for what makes a good programmer and the one thing they found was musical skill. So would Ceglowski allow that a hacker is like a composer of music?

    Personally, I like Graham a lot and will buy his analogy between hacking — in the classical sense of elegant coding, not it’s modern sense of breaking into things — and art. Then again, my training is in computer science.

    Obligatory mathematician’s joke: Q: what disciplines do you know are not a science? A: any discipline that adds the word “science” to the end of the title.

  8. Dzhaughn says:

    It is so like painting; interior house painting. It’s 90% prep, you have to move everything out of the way, put it back, not make a mess, and ideally your work ends up unnoticed.

    If only we could find a way for painting to give thieves the ability to enter your house.

    • AllanC says:

      This reasoning works just as well for a plethora of other construction activities.For example:

      “It is so like hot-applied waterproofing on a parking garage roof deck. It’s 90% prep, you have to completely clean the area, apply the product, don’t make a mess, put back the overburden and ideally, your work goes unnoticed (read: no leaks)”

      Ditto for expansion joints.
      Ditto for unexposed concrete slabs/walls (but with deflection instead of leaks).
      Ditto for structural steel (but with deflection instead of leaks).

      I am unconvinced this is sufficient reasoning to select painting out of all the possible related construction activities! That said, I don’t really mind the comparison to painting. I just saw an opportunity to talk about waterproofing and I took it.

      BTW: If you subcontract your painting work and some members of the crew are thieves then you’ve solved your last problem.

      • As long as we’re using analogies, I’d say coding is like assembling Ikea furniture, and Hacking is like the guy in his garage putting together a custom dresser with dovetail joints and hand selected boards with interesting grain patterns and inlaid nautical emblems or whatever. Sure the end results are both functional pieces of furniture but one expresses a unique vision of what is important to the creator who is also the client.

        Designing the Ikea furniture in the first place is somewhere in between — Engineering, which uses similar skills but in the service of efficiency and mass production. It just happens to be the case that making perfect copies of computer products is trivial, so compromises for mass production are less of an issue.

        All these coding camps that are all the rage will result in people who can put together bits and bobs to make some Ikea furniture, but are not going to result in people who create custom languages for defining stateful ip firewalls or invent the PDF format or design the Opus sound codec.

  9. Ian Fellows says:

    If you are you referencing a blog post, is there a link to it?

    I’m a bit disappointed you had such a negative reaction to GEB. I remember there being a lot to like in there. In particular it got me thinking in different ways about proofs, axioms and self-reference. For undergrad me, it gave me a much clearer idea about how the problems brought up by Godel were “unfixable” in a way that other texts on the subject did not. Admittedly, I doubt it would have had anywhere close to the impact if I read it now, given I’m a working Statistician, but it was great for where I was at the time.

  10. ArtnScience says:

    My day job consists of a lot of statistical programming. I also enjoy pencil sketching and other media. I have often thought of my “code” as art…from what it does to how it is arranged. The process of coding feels just as creative as when I need to stop for a moment and think about how to represent the world through the use of a pencil.

  11. Gerry Quinn says:

    Nothing here to contradict my long-standing theory that the primary quality of Lisp as a language is making its programmers think they are geniuses.

    • Dzhaughn says:

      Cute, but if it were true we would all use Lisp.

      • In fact, it’s the other way around, relatively few people use Lisp, and the population that does is certainly enriched with people with higher SAT/GRE/Intelligence test scores compared to say Java programmers or C++ programmers which is mostly a group of people hired to do some commercial task.

        It’s an aesthetic thing IMHO, if you like things like group theory or nondeterministic pushdown automata or whatever, you’re also going to find Lisp appealing. It’s also relatively hard to motivate those guys by money. If you just want to write some code that scrapes a website for data to help you day-trade stocks or need to create weekly sales reports for your boss then you’re going to want something with a library that basically does your task for you and you aren’t going to care much about the aesthetics.

        There’s a reason Abelson and Sussman wrote SICP instead of “Code Camp: Learn Java in 4 easy weeks”

  12. SeanMatthews says:

    Bit late to this discussion, but an interesting reference point here is The Science of Art by Martin Kemp, which does not talk about programming, but does talk about the extensive, complex and mathematically sophisticated planning that went in to – in particular, but not only – classical works of art. The artists Kemp discusses did not approach the problem of making pictures as one of daubing paint on a canvas and seeing how it went.

    • ArtnScience says:

      Interesting point…reminds me of my recent efforts to faithfully draw birds. It took a bit of time to study the anatomy and physiology of my target and understand how muscle, bone, and feathers work together to form a shape. Birds in flight involved understanding a bit about the aerodynamics and nature of winged-flight.

Leave a Reply to Jonathan