Late last night I build harvester.blue if it's been more than a day or two since this blog was published the site is likely shutdown. Frankly it's expensive, doesn't work great, and isn't where I want to put my time, money, and effort.
Building the Harvester
People have commented on the fact that at protocol can seem like a big building filled with empty rooms, and I think thats true in a sense. Some of the apps beyond bluesky are starting to get traction (tangled.org, sifa.id) but a lot aren't. A couple of weeks ago I built branchline.ink but it hasn't really gained any traction - not surprisingly I haven't done the leg work to help it. Branchline was something for me, but it was an experience rooted in User Generated Content that would only make sense if it was also for other people.
The idea that sparked the blue harvester was to build a "Product Hunt" for the atprotocol. It would be a way for apps to generate buzz, go "viral" and actually benefit from a large user base. Of course the day after I had the idea to build this Atmosphere Account launched their explore page - https://atmosphereaccount.com/explore and on the same day atstore.fyi launched.
Atmosphere account's explore page is one of those empty rooms - Branchline is there, at the time of writing so are 19 others - 20 apps in total 1 week after launch. Atmosphere account is trying to solve the same problem that my blue harvester was, and succumbing to the same fate it's trying to prevent.
Atstore on the other hand took the questionable step of generating hundreds of product listings without creator consent, and used AI Generated logs and artwork. As someone who makes heavy use of AI I'm not opposed to it's usage but using it to create branding content for someone else's product, without their consent is wrong. That being said Atstore solved the empty room problem to a certain extent, by seeding it's listing pages rather than asking creators to make their own they generated 221 listings in the first week. The ai controversy gave them visibility, the complete listings made the app useful on day one, and it seems to have gained traction since then.
Is the solution to our empty room problem to gin up controversy and bootstrap UGC with AI? I hope not...
What's next?
I'd like to get some of the stats working better on harvester.blue, I'd like it to not cost me $250 per month in database storage to do that. Ultimately I think the right home for those is over on ATStore so I'll probably contribute what I've built to them. Ultimately as a community we need to invest in the rooms where there are people rather than building new rooms in the hope that people use them.
I have another project that is mostly ready for an alpha launch. Unlike branchline or harvester it's something that is just useful on day one - well sort of useful since it really needs permissioned data. That may get a small amount of traction, or it may not but ultimately it won't matter to me because I'll find it useful.
Why It Matters?
The AtProtocol is a really cool idea - data owned (truly owned) by the people who create it, not the services that leverage it. That's something that deserves attention and users, and while most users won't care about the ownership aspect of it I think the follow on effects will be things we as a society will care about. The problem is that no one has figured out how to monetize this. Apps like Leaflet can sell pro subscriptions with additional features and metrics in the app view, but since the data is truly portable it's easier than ever to stitch apps together - edit on Leaflet, send out newsletters through another service, read on a third service, etc.
We need a few of these services to catch on and get big enough to be able to experiment with monetization, we should have several working examples of building a app on the at protocol that generates enough revenue to sustain itself. I don't want to tell people not to chase these passion projects but I think as a community we need to start to form up around a handful of big players and make it clear that we're going to give them the oxygen they need to grow and figure it out.