Beyond Hacker Mindset
How to speedrun failure
Hacker mindset is a real-life superpower. The idea is simple: instead of treating a system or process as a black box that comes with rules that you have to follow, treat it as a collection of component parts that you can engage with individually, often in unintended ways, in order to accomplish your goals. That way, instead of doing what a system explicitly allows you to do, you can hack into it and do whatever you want.
There is an established literature on the topic; Henrik Karlsson recently added to it with “How to walk through walls.” He gives the example of director Robert Rodriguez, who started his career by spending $7,000 to shoot a movie at a time when film budgets were already regularly hitting the tens of millions. Robert Rodriguez didn’t follow the standard procedure of hiring a crew, scouting for locations, and buying enough cameras and film to shoot from multiple angles at once. Instead, he used his actors as his crew when they weren’t currently in shot, made a list of the locations available to him from where he was staying and built the film’s settings around those, and filmed with a single camera that he simply moved several times over the course of shooting a scene to simulate multiple cameras’ worth of coverage. He dealt with the needs for a crew, location, and equipment one by one without following the typical process or deploying the usual systems.
This thriftiness is visible in the final result, with mistakes by actors and continuity errors left in that would be verboten in a standard, big-budget project, but the movie still made millions. Other classic examples of hacking to power concern the job market, which you can see as a large, standardized system made of job boards with postings that ask for certain qualifications, or as a large group of individual people that you can talk to who are looking for someone to solve problems for them, and speedrunning, a practice that breaks video games down into a bunch of physics systems and objects in memory that can be manipulated into unintended interactions, letting you “see the Matrix” and break the rules of the in-game world. There are a lot of great videos about how speedrunning glitches work; here’s one that explains how a recent one was found.
My last normal job was at a small, independent tech startup, and we did tons of this kind of stuff.
Growth Hacking
The “hacker mindset” approach is highly valued at startups, where egos are often large and resources are often slim. The famous startup incubator Y Combinator includes it as a question when you apply to partner with them: “Please tell us about the time you most successfully hacked some (non-computer) system to your advantage.” A classic example: cloud services like Dropbox and Digital Ocean will often give you free credit if you get someone else to use their service using your own personalized referral link. If you know how to use Google AdWords, you can take out an ad, pay Google a few dollars to show it to hundreds of people, and put your referral link in it, generating huge amounts of free credit. (This was more effective before everyone had Dropbox.)
My boss in particular was obsessed with this approach; he pursued it in all aspects of his life. He gathered coupon codes to experiment with to figure out which ones would stack additively when making online purchases, so he could put in two 40% coupons and reduce the price of something like a hat or a coffee mug by 80%. Credit cards offer bonuses and cashback rewards to try to get you to use them and pay interest on them; he knew all of the major providers’ rules, and used them to pointsmaxx. If you needed someone to get a whole series of free trial accounts for something, he was your guy.
We were trying to create an e-commerce platform, and there was a lot of value in figuring out how the businesses of existing competitors worked. One way you can do that is by looking at your competitors’ public websites. Another way is by making a Shopify account, setting up a fake store, and installing every app on the Shopify App Store that integrates with existing online marketplaces. If you install the app that lets you sell products through an online distribution channel like ShopCanal, you can learn how they run their platform and what their product category breakdown is, even if you have no intention of actually selling anything through ShopCanal.
Target operates a marketplace for third party sellers called Target Plus, and if you go to their website and click through to some of their forms you end up on a URL that looks like forms.target.com/1. If you change that 1 to a 2 or a 3 or a 4, you can see different ways that Target communicates with their sellers. There are lots of things like that.
When you’re figuring this kind of thing out, it’s easy to narrativize it: to see each success as a sign of inevitable progress and feel like you’re on top of the world. But I think that obscures a major aspect of how these kinds of glitches are really found.
Glitch Hunting
If you watched that video I linked to above, about how speedrunners found a new glitch in The Legend of Zelda: Twilight Princess that lets them (literally) clip through walls, you know that by the end, it goes into extensive technical detail: talking about position offsets, different collision types, and the formula that’s used to compute a wall’s normal vector for the physics engine.
But that’s not how it starts. At the beginning, the guy who discovered the glitch is simply walking into walls to see what happens. He walks into walls while holding an item; walks into a specific corner, notices a weird interaction, and experimentally tries different tricks that are known to cause other glitches in other circumstances; and so on.
I think that any description of hacker mindset is incomplete unless it acknowledges how much time is spent just kind of monkeying around. A large element of it is testing the rules of a system by attempting to ignore them, until you find out by trial and error which are enforced. There’s a lot of doing dumb stuff over and over just in case it works.
It takes a lot of tries to find a set of coupon codes that stack to 80% off. Speedrunners put thousands of attempts into world record speedruns. In the meantime, you’re looking for information that might not be there, and scrolling through a bunch of random apps and forms. What happens if you don’t find what you need?
Social Engineering
Here are a few other stories I could tell about my boss:
He wasn’t a big believer in formal roles or titles; hacker mindset told him that that kind of thing only existed to be seen through. “There’s the org chart,” he would say, “and then there’s who’s actually in charge.” As a result, it was never clear who needed to make product or technical decisions, and different parts of the site would break or be built in different incompatible ways, with no one responsible for keeping that from happening. Since he was used to breaking down a system into component parts and trying to interact with them separately, it didn’t occur to him that the only reason that the parts are relevant or useful is because they’re arranged in a system, which we on the software engineering team were simply not.
Other times, he would think that he could subvert a system or hierarchy that he couldn’t. At one point, he was advocating for a complex type of user-driven marketing campaign that our CEO didn’t think we should have; since he stood between upper management and the software engineers, he realized that he could just have the engineers build the feature without the people above him knowing. However, our CEO had principled arguments against the way that this feature would be seen by our users, did not appreciate his objections being unceremoniously routed around, and the result was a bloodbath where months of work were excised.
Around then, I quit. In order to entice me to say, I was offered a raise and an extended paid vacation. After I refused the offer, he told me with chagrin that he’d expected me to at least accept the raise, go on the paid vacation, and then quit when I got back, collecting an elevated paycheck in the process. This idea was presented as though it were obvious; I assume I could have thought of it if I wanted to, but I didn’t. I think this is the kind of borderline-paranoid mindset you fall into when you spend all of your time trying to find the exploits, override the rules, and gain access to the components underlying the system.
Risky Business
Hacker mindset methods are great when they work. However, they’re much riskier than following a system. If you apply them to enough important stuff, you will eventually run into a situation where they fail catastrophically.
Hacker mindset often fails to acknowledge that inefficiency is the price you pay for getting things done reliably. Hiring lighting techs derisks the possibility of losing a day’s worth of shooting to overexposed footage. Robert Rodriguez was confident enough and had enough amateur video experience when making El Mariachi to bet on himself. How many of us have that kind of confidence? How many of us should have it? Would you reasonably expect a movie studio to try to get good lighting setups out of its directors, or does it makes sense for them to hire those lighting people?
Every startup, if it grows and survives, becomes an complex corporate behemoth eventually. Look at Facebook. Robert Rodriguez continued to value scrappy efficiency and always did a ton of stuff on his movies himself, but he also moved on to direct movies like Spy Kids, working within the studio system and having a budget of tens of millions of dollars. Not even speedruns exist in a state of total anarchic glory: they still fall within categories and follow carefully decided-on rules (just, not the ones that the game expects you to follow.)
There’s an irritating arrogance that’s produced by the hacker mindset ideal sometimes. “I’m so special; the rules just don’t apply to me.” “Move fast and break things.” “We all got to that line, and crossed it.”
Closed Loops
All of which is a long way of saying that I’ve been trying to adopt a “hacker mindset” recently. It’s useful for things.
Most problems break down into approachable components when you really think about them. Don’t treat your landlord as an opaque entity operating under a set of published rules; treat him as some guy who wants to maintain continuity of rent payments more than he wants to keep you specifically as a tenant. When negotiating to be paid for something, don’t treat it as a weighty deliberation about that thing’s true value; treat as something that the purchaser has a specific budget for, and ask for a high number that they can afford. There are a lot of things like this.
But I will aver that it’s better to have the ability to hack into things as a tool available to you for occasional upside, rather than as a thing that you depend on. The best position to be in is one where you don’t have to hack your way to success.