That’s a zoo, not an ecosystem.

(This is about connected objects, and also the tech industry at large.)

While hobbyist IoT got nerdier, consumer IoT picked narrow lanes like thermostat or light bulb and worked on experience. Unfortunately that lasted only until growth proved shallow and users had to be squeezed for more money. And now “Internet of Things” is a poisoned brand.

The latest in moats.

How did this happen? “Ecosystems.” It comes with every paradigm shift, but when Apple rode a media player/online store integration to flip the center of gravity of the music industry, and then used apps to do the same to mobile carriers, it made other industries absolutely paranoid.

Tech and non-tech giants alike now have their ears up to prevent anything from becoming a success without them—and so nothing becomes a success because there’s no room for the users and startups around them to evolve the products into what they really should be. *cough Matter* Everyone’s digging moats before the townsfolk come over.1

Product co-ops, not coops.

Ecosystems need diversity to thrive, and not just a curated set of cages. We need products that are complete in themselves, but readily composable into a system using unencumbered standards like MQTT or HTTP. That’s a useful thing about Shelly relays, or any of the polished devices and services that expose an open protocol. That’s the kind of thing Supermechanical is trying to build.

This idea was called Consumer Electronics 2.0 at the MIT Media Lab, and it failed because we depended on hardware companies doing the right thing. Open interfaces will empower those of us who don’t have technical skills and/or standards consortium membership to steer where all this wonderful technology should go. Instead we have a sterile corporate vision of the smart home. Long after the business models are dead, we’ll still have the ewaste.

Not an engineering problem.

A long time ago when I worked for a large company, some other team demoed to the executives a web app they were really proud of. It was packed with absurd bells and whistles, but worse than offending my design sensibilities, they were implemented with ActiveX, Microsoft’s early attempt to embrace-and-extend the web. I was able to kill this UI abomination (it was Clippy with a different avatar, no joke) by pointing out they would be turning away the 20% of the company’s customers who used non-Internet Explorer browsers, including screen readers.

How do we get more products that are not locked in? Let’s find business arguments to advocate for more accessible and cooperative products. I’m not going to Stallman here and insist that consumers buy ideologically pure but inconvenient products—we need to give them better options. We can do the right thing and deshittify our products, but to be sustained, we need to find the strategic paths that happens to be more cooperative.

But I don’t know what the business arguments and paths are for every market. Do you? Let’s discuss case studies and ideas. Comment on the Hacker News thread, or social us on Mastodon (cooperative!) or Twitter (eh).

  1. Content warning: Tortured metaphors. I love to stretch them out on a rack and twist them until they lose all sense. Look, I’m still holding onto this medieval theme.