It was near the end of April when I interviewed the creator of Bubbles.town, Benjamin Behnke, and I feel as though I've been using the platform for enough time to give a more long-term review of it now. Here are some of my thoughts, and if you disagree with any please drop a comment!

The Good

I want to start with the positive, which is that I've discovered and continue to find new blogs on a wide variety of interesting topics thanks to Bubbles. Blogs about family genealogy, wonderful illustrators, musicians, souping, Star Wars action figures, Google Maps, science fiction writers, bullet journals. I could go on and on, and that's just from this week.

There are so many topics I don't have any knowledge of whatsoever, and it makes me so happy that there are people who have dedicated blogs to these topics! Bubbles allows me to discover personal sites that I would not have encountered otherwise. It reminds me of visiting my public library and browsing the more obscure aisles of the Dewey Decimal system and being astounded that so many books on each particular topic exist.

In this sense, Bubbles has been a fantastic success story for the IndieWeb, and demonstrates how it is a net positive rather than a net negative, in my opinion.

That said, I do still certainly have my issues with the aggregator, particularly as it becomes more popular and used as a primary source for bloggers and IndieWeb participants. So, onto my qualms.

The Pace

I don't know if I visit the aggregator too often (two or three times a day) but I notice that the top posts change at a rather glacial pace. It reminds me of how, back in 2018, Reddit changed their default sorting from "Hot" to "Best" and the ability to use the site for breaking news near totally collapsed.

And I think this is especially puzzling since it's Blaugust, surely people are blogging like crazy right now, right? I'm not sure if this simply has to do with how Bubbles calculates its algorithm: Score = (1 + Votes × 25) / (EffectiveAge + 2) ^ 1.4. Now, this is indeed an algorithm, but it is not an impossible black box algorithm that is unknowable to even the engineers who designed it, the way corporate social media has. It's just plain math.

The Popularity Contest

One thing I don't like is the language used by the site, the about page specifically says that "you vote on the posts worth reading. The good stuff floats to the top. Everything else sinks over time." And this couldn't be further from the case in my experience. Posts on page 10 of the aggregator with zero upvotes are typically as interesting, or more interesting, than whatever is on the front page.

Also, something new I just noticed: there's now a classics page which shows off the all-time most upvoted posts and websites. As of writing, I'm in third place, with 340 upvotes from 156 voters. Zoo-Wee Mama! From a statistical curiosity perspective, I think it's neat to have a page like this, but for someone who already thinks Bubbles is a popularity contest, this just amplifies that argument.

I'm a writer that is deeply uninterested in how many visits my site gets or upvotes my posts get on Bubbles. But I know that's not the case for everyone. Some people get anxious or worried about what others will think, and Bubbles gives a rather obvious and problematic framing to that. As soon as you become focused on trying to cater to an audience instead of writing what YOU want to write, the entire process will quickly become unfun and burdensome. So if metrics are something you care about, I'd recommend not putting your site on Bubbles, or ask for it to be removed.

The Categorization

Bubbles has categories, and the way they work is rather neat. From the FAQ:

Categories are decided by a small statistical classifier (Naive Bayes) that ships with the binary and runs on the Bubbles server, with no external service involved.

This is far from perfect, and after using the aggregator for some time I would recommend adding more categories to help with filtering. Specifically, I'd love the ability to filter out genAI posts rather than all tech posts in general, since another criticism I have is that man, there are a lot of people writing about genAI on Bubbles. That doesn't have anything to do with Bubbles specifically, that's just a byproduct of people writing about genAI a lot in general.

A "meta" category would also be great, to be able to filter out posts that themselves are about Bubbles.town (like this one!).

The Community

Another issue Bubbles has is community participation. A place like Bubbles can only ever be as good as the willingness people have to use it. I'm a culprit here, as I do a bad job at upvoting posts I enjoy (this is a case everywhere I go, I am so forgetful of giving a heart on posts on Tumblr, Mastodon, Pillowfort, anywhere that has that ability). An aggregator only works if people interact and participate with it. Likewise, because the categories are defined automatically, they can be wrong (like my post on "Buying Them a Coffee" being categorized as food) and if you click on a category on a post, you can easily submit which one is more accurate.

Commenting culture is also incredibly sparse on Bubbles. I understand it's a little awkward and clunky, since you need to sign in and post comments via Mastodon, so I'm hoping people are instead just commenting directly on people's sites, linking to posts they like, or directly emailing bloggers.

I also will admit I have some beef with seeing certain bloggers on Bubbles, such as Joan/JA Westenberg who has over 14,000 subscribers on the Nazi-complacent platform Substack, and who was previously heavily involved in the Web3/NFT scene before deleting everything regarding that. Sure, Bubbles should be for small, independent creators who aren't just looking for another revenue source. And maybe I'm hypocritical for that criticism given my own stats, but I like to think at least I actually deeply care about the IndieWeb and haven't just decided to hop to the next most viable market.

The Heart of It

Overall, Bubbles.town is a weird project. It is definitely a "ask for forgiveness instead of beg for permission" kind of project. As in, if someone told me they had the idea of starting an IndieWeb aggregator using genAI and collect thousands of personal sites and blogs without asking anyone, I would implore them not to do that and think it was a terrible idea.

But Benjamin did this, and I would be dishonest if I said I didn't appreciate Bubbles.town. He has always been quick to respond and makes it easy to remove your blog.

If you view Bubbles as just one of many ways to interface with the independent web, along side directories, alternative search engines, webrings , blogrolls, human curation and also use other aggregators like nuchronic or scour.ing, then it's a great tool.

And if you're someone that's totally uninterested in aggregators and think they're problematic, that's also really understandable! There's no "right" way to IndieWeb or share yourself, and I'll keep trying to find new ways to discover people and sites and blogs that don't rely on automated aggregators. The discovery is sometimes the best part.

And, likewise, you are somebody that has a personal blog that you're active on, particularly if it's of a niche interest, I recommend you submit your blog to be added to Bubbles. Adventure is out there!