Bryan Cantrill has a great line from a talk where he cautions against anthropomorphizing1 Oracle CEO Larry Ellison. I think it’s also a mistake to treat any tech company like a person. Here’s the full quote:
\n\n\nYou need to think of Larry Ellison the way you think of a lawnmower. You don’t anthropomorphize your lawnmower, the lawnmower just mows the lawn, you stick your hand in there and it’ll chop it off, the end. You don’t think ‘oh, the lawnmower hates me’ — lawnmower doesn’t give a shit about you, lawnmower can’t hate you. Don’t anthropomorphize the lawnmower. Don’t fall into that trap about Oracle.
\n
Here are some examples of anthropomorphizing companies:
\nAll of these are mistakes. Your relationship to the tech company that employs you is not like a relationship with a person. It’s more like your relationship with your car2. If you drive poorly or fail to change your oil, your car may stop working or even kill you, but it doesn’t love or hate you. It has no sense of fairness. It’s simply a dangerous, useful machine: useful enough to be worth getting into despite the risks. Anthropomorphizing it makes it harder to avoid those risks3.
\nWhy is it so hard to avoid this trap? First, humans anthropomorphize everything by default. Second, whenever you’re interacting with the company, you’re interacting with a person: your manager, your colleague, someone in HR, and so on. Third, when you’re in the power of some entity, it’s psychologically comforting to think of it like another human. This is the same principle that caused early humans to attribute lightning to the power of Zeus, because Zeus (unlike electrical imbalances in storm clouds) can be placated with prayer and sacrifice.
\nRelationships with machines are very different from relationships with people. The biggest difference is that machines do not respect effort. If you put a lot of work into doing a favor for a friend, your friend will appreciate it even if the favor itself is not a big deal. Conversely, if you help somebody out at trivial cost to yourself, that help will be relatively unappreciated. But a car doesn’t care how much effort you put into maintaining it. If you spend hours replacing a car’s engine, but the brake line is damaged, the car won’t remain functional out of respect for your effort on the engine: instead, the brakes will fail, killing you.
\nTech companies work like this. Everyone who’s worked at one for long enough knows that it’s very hard to predict what work will be high-impact. The project you slaved over for months will often get a “nice, now what’s next” reaction, while the tool you built in fifteen minutes (or the question you answered in a Slack channel) will sometimes get wild praise and recognition. This can drive engineers a little mad, because they’re anthropomorphizing their tech company, and a person who behaved like this would be dangerously unstable. But if you treat your company like a machine, this shouldn’t be surprising at all.
\nIt’s hard to predict high-impact work, but it’s not impossible. The best way to operate at a big tech company — as I’ve said again and again — is to try and develop this skill, and to spend most of your time looking for the low-effort high-reward wins. It’s hard to do that if you’re secretly imagining your company as a person.
\nA natural objection here is to point out that tech companies (unlike cars) are composed of people who might indeed owe you, hate you, love you, and so on. According to this view, when people say “my company hates me”, they aren’t making the mistake of anthropomorphizing a corporation, they’re using shorthand to say “my boss and skip-level hate me”. I think this is still a dangerously sloppy way to think.
\nIn my experience, engineers who think they’re hated are often not. Managers in large tech companies do hold grudges, but (particularly the higher up you get) are tolerably cold-blooded. You don’t get to senior director or VP by needlessly throwing away useful tools. I can recall a few cases4 where engineers told me they were hated, but I later found out5 that managers considered them dangerous or unreliable: loose cannons whose usefulness didn’t outweigh their tendency to go off the rails and cause damage. From the perspective of the engineer being forced out of the company, it feels like you’re being run out of town by people who hate you. But if you want to avoid being run out of town, it’s useful to recognize when you’re being treated as a defective tool.
\nIt’s also useful to keep in mind the difference between “my manager hates me” and “my company hates me”, because managers are often temporary. You can sometimes just wait out a bad relationship with your manager. You’ll have a slightly uphill battle with the next one — since the handoff won’t be positive — but it’s certainly possible to get back on track. If you can figure out why your manager hates you, you might also be able to fix it (or at least mitigate6 it).
\nEngineers who’ve learned not to anthropomorphize tech companies sometimes go too far in the other direction, by treating companies as a computer program. This is the wrong kind of machine: tech companies aren’t people, but they aren’t programs either. They operate on incentives and illegible interpersonal relationships, not on consistent logical steps.
\nThe classic example of this is the engineer who reads the company handbook cover-to-cover and makes sure they follow all the rules. When they’re looking to be promoted, they take the role guidelines and make sure they’re following them point-by-point. They take what they’re told very literally: when the company says “X is our top priority”, they internalize that.
\nThe problem with this is that handbooks, guidelines and internal communications are the legible superstructure of a massive illegible web of dependencies that make up the actual operations of any company. That superstructure will not cover the real criteria for promotions, or the rules that are designed to be broken. The “top priority” of the superstructure can rapidly shift, leaving literal-minded engineers confused and disheartened.
\nIt’s understandable to feel ill-treated by a company, or to harbor resentment — particularly since there are actual humans at the company who are involved in that treatment of you. But companies themselves don’t hold grudges, even as they force out people who they’re incentivized to force out7. It can even be a mistake to anthropomorphize very senior managers, whose work personas are more identified with the machine the higher up you go8. I suppose at the top you have Larry Ellison, as Bryan Cantrill says, who is better modeled as a lawnmower than as a human being.
\nNo relationship with an individual person at your company will save you from being careless with the machine: it will chew you up and spit you out. You need to learn how to drive the car. This is scary, because it means that sometimes you will just get punished for no reason. But it’s also good news, because it means that you can get disproportionate rewards for relatively little effort, so long as you make the right decisions.
\nTreating as a human.
\n↩\nI like this example better than Cantrill’s lawnmower because cars are more central to our lives, like jobs are, and are also far more dangerous.
\n↩\nFor instance, if you’re prone to thinking “my car loves me, it wouldn’t spin out when I take this hard turn on gravel”, you are more likely to crash.
\n↩\nThough certainly not all. Grudges do exist.
\n↩\nOf course, this is my interpretation of the situation, and could be wrong.
\n↩\nThere are a range of techniques here, ranging from “do all your paperwork defensively” to “find some out-of-the-way project” to “ingratiate yourself with your skip-level”.
\n↩\nFor instance, people who make their managers look bad, or who publicly point out the inevitable inaccuracies in legible reports, or who campaign against the current big push the company is making.
\n↩\nI say “work personas” because these people often have a rich personal life that is largely firewalled off from their work selves. For more on how managers come to identify with the rules of the organization, I recommend Robert Jackall’s Moral Mazes, a book that has informed much of my thinking and that I will some day write a long-form review of.
\n↩\nIn 1973, Tom Lehrer famously sang that “selling out is easy to do”. That may have been true in the seventies, but it’s not true today: selling out requires both technical skill and a careful sense of how large organizations work in practice. One of the goals of this blog is to teach people how to do it.
If you want to live with uncompromised integrity, you don’t need anyone to tell you how: simply always do exactly and only what you want to do. You will be repeatedly punished for it — your managers will dislike you, you will lose jobs and lose out on the chance of getting hired, your financial situation will be less stable, and so on — but that’s just part of the deal. Like deadlifting four plates, it’s not easy, but it is straightforward.
Continue reading...