I use Airgradient air quality sensors in a couple places in my home to keep my eye on indoor air quality. One of the reasons I chose them is their open approach to data sharing. I got a note from them this week asking my opinion on how they should handle a situation where another org they work with is refusing to share data in a similar way. They’ve detailed things in a blog post if you want to get the nitty gritty.

It got me thinking though that Airgradient got themselves into that situation because they took a maximalist approach to openness, and didn’t think through additional controls that might be needed if they cared about whether or not someone else didn’t “play nice” in a way that they implicitly were saying they expected them to. They just assumed that other folks wouldn’t take advantage of them. To be clear, I think Airgradient will navigate this well, and in this case it’s not a particularly sticky situation, it’s just the thing that got me thinking.

This kind of thing comes up frequently in Open Source. A particularly ugly version of this is playing out in the Wordpress community as I write this, and shows no signs of wrapping up anytime soon.

Taking a maximally open stance like this is appealing in its’ simplicity, but I would argue that kind of position is almost always wrong. To actually achieve the goals that a maximally transparent, open, or socially-beneficial position is trying to achieve controls need to be put in place to curb the actions of bad actors.

It’s why open social communities need conduct guidelines to keep hate speech at bay. It’s why capitalism needs regulations to keep abusers of all kinds under control. It’s why Open Source licenses continue to proliferate and get increasingly nuanced. The need for being explicit about things that it may have been safe to make assumptions about in the past has never been more clear.

It’s really a systems design problem at the end of the day, and looking around right now I can see the problems with the design of our social systems cast in particularly frightening relief. It’s clear that there were a whole lot of assumptions about how people in would behave underpinning the design of our most important social and governmental systems, and that capture at the scale that we’re experiencing simply wouldn’t happen.

As we’re trying to hold the existing systems together to keep people safe and healthy, it’s important to be thinking about the non-obvious design assumptions that are embedded in our systems, and how we can avoid those problems when we get the chance to build what comes next. Don’t fall into the trap of the toxic appeal of a simple and maximalist position. It will almost certainly not do what you hope.

Good luck out there.