So Paul Graham wrote a pretty smart article about how inducing safety controls also creates ‘economic drag’ (I just made that term up). I have a soft spot for the guy having read “Why Nerds are Unpopular” (which is an excellent starting analysis of the failure of our system of education. He names the roles of the schools, but draws minimal motivations for them from the economic system that creates the needs for those roles. Still it’s good stuff, you can’t expect too much from libertarians :P)
[Before I continue, I should disclaim any actual knowledge of Paul Graham's political position... his sort-of free marketeer outlook along with being a solid coder means I can make educated guesses]
But I should get back to “Artists Ship”. The crux of Graham’s essay is the consequences of certain corporate behaviours:
The gradual accumulation of checks in an organization is a kind of learning, based on disasters that have happened to it or others like it. After giving a contract to a supplier who goes bankrupt and fails to deliver, for example, a company might require all suppliers to prove they’re solvent before submitting bids.
The paper trail, bureaucratization, management approval, and all the other niggling irritations in large organizations are the obvious sign of this behaviour. Graham talks about costs, and being a programmer talks about how these behaviours are particularly irksome to programmers:
Here’s a sign of how much programmers like to be able to work hard: these guys would have paid to be able to release code immediately, the way they used to. I asked them if they’d trade 10% of the acquisition price for the ability to release code immediately, and all three instantly said yes. Then I asked what was the maximum percentage of the acquisition price they’d trade for it. They said they didn’t want to think about it, because they didn’t want to know how high they’d go, but I got the impression it might be as much as half.
Graham points out that 5000 dollar projects turn into 500000 dollar contracts once a company committee has to manage it, simply because the coders don’t want to babysit them extensively and all their big corporate bullshit. He’s absolutely right. That’s because he’s smart. But as a free marketeer he hasn’t delved any deeper. What this highlights is symptoms, not root causes, though he hints at the roots of where this comes from:
Programmers, though, like it better when they write more code. Or more precisely, when they release more code. Programmers like to make a difference. Good ones, anyway.
For good programmers, one of the best things about working for a startup is that there are few checks on releases. In true startups, there are no external checks at all. If you have an idea for a new feature in the morning, you can write it and push it to the production servers before lunch. And when you can do that, you have more ideas.
This is a common realization about programmer work ethic, more thoroughly explored in “The Hacker Ethic”. As a programmer I can corroborate these ideas, and thanks to Grahams negative characterization of what DE-motivates programmers I think I have a better metaphor.
What we are talking about here is INERTIA. Programmers love inertia, they love the feeling of dynamism, rapid feedback, and really going somewhere quickly with their work. It’s like driving, do you like driving fast or laboriously slow? One is safe, but not that exciting. The system of checks is intended to make a group slow down. It induces drag, friction, resistance, entropy — whatever you want to call the act of actively killing inertia. And it’s safer, but it comes with a cost. Programmers are not linearly less effective, we become exponentially less willing to work as inertia drops away. Graham points out some non-programmer cases as well where the consequences are losing business and efficiency.
One can extend this beyond business, or even personal projects to the world at large. A recent pleasure of my life has been linking long-held truisms I received from thoughtful authors such as Frank Herbert and to my current political philosophy and life in general. In God Emperor of Dune he says:
The social forms by which cities make the attempt [to survive] are worth study. Remember that there exists a certain malevolence about the formation of any social order. It is the struggle for existence by an artificial entity. Despotism and slavery hover at the edges. Many injuries occur and, thus, the need for laws. The law develops its own power structure, creating more wounds and new injustices. Such trauma can be healed by cooperation, not by confrontation. The summons to cooperate identifies the healer.
Linking back to Graham’s essay, the institute/structure of law can be seen as a societal attempt to reign in ‘human nature’ as people perceived it. Obviously there are dangerous elements/people/thoughts, and we should work to restrain those. Just as businesses work to prevent making mistakes that cost money we should work to prevent mistakes that cost people lives. But such simple characterizations are again not accounting for the cost/drag/entropy that unbending checks create. The question to ask ourselves is: “Do we live lives that are fulfilled? These checks do save lives, but do they cost us our lives as well?”
There is ample evidence that the system of laws we live in has had a significant negative impact on our society. Graham has argued, at least in software, that letting it hang loose is a better overall strategy than insisting on safety at every step. Anarchists would argue that a rejection of State and Capital is the societal equivalent, with similar benefits.
I’ll conclude with the middle-ground. Bureaucrats (corporate, government, and otherwise) want unbending workflows to determine their decisions for them. It allows them to escape responsibility when things still go sour, while lending the illusion of business-like activities and responsibility to their jobs. Again, we refer to Frank Herbert in God Emperor of Dune:
The difference between a good administrator and a bad one is about five heartbeats. Good administrators make immediate choices … [that] usually can be made to work. A bad administrator, on the other hand, hesitates, diddles around, asks for committees, for research and reports. Eventually, he acts in ways which create serious problems … A bad administrator is more concerned with reports than with decisions. He wants the hard record which he can display as an excuse for his errors … [Good administrators] depend on verbal orders. They never lie about what they’ve done if their verbal orders cause problems, and they surround themselves with people able to act wisely on the basis of verbal orders. Often, the most important piece of information is that something has gone wrong. Bad administrators hide their mistakes until it’s too late to make corrections … One of the hardest things to find is people who actually make decisions.
There is something special here. I have always been struck by the simple truth of this quote, and for the first time feel like I have something to add. In all of humanities endeavours having a ’system’ or ‘checks’ isn’t necessarily bad, but to execute decisions entirely constrained by them is a mistake. We must watch ourselves (and fellow participants) for the dangerous behaviour of disavowing our own responsibilities in actions. The system provides good guidelines but slavish adherence to it produces a legion of robots who are incapable of responsible decision-making. Instantly shooting from the gut all the time isn’t helpful either as it doesn’t take into account wider context and experience. As with many things in life, the key is walking the fine line between those two states, and succeeding for a long period of time will manifest more positive results no matter the decisions to be made.
