fine-tuning discoverability
Herman recently posted about the Discovery and AI discourse relative to his Bear Blog platform. He mentioned how prevalent, polarizing, and inescapable AI feels for certain people. What confuses me is how some amplify that sensation by perpetuating vicious cycles. Getting peeve-sniped is voluntary. Getting peeve-sniped by someone else who got peeve-sniped is even more voluntary. However, as Kendrick Lamar once said:
First, you get a swimming pool full of AI, then you dive in it
Swimming pool full of AI, then you dive in it
I view pre-existing audiences pushing posts up Bear Blog's Trending page as a feature, not a bug. Such posts could pique my interests and fuel my writing like any other post. They can also introduce me to novel creatives, ideas, communities, and initiatives that I wouldn't have otherwise found.
I like how Herman not moderating by topic avoids the need for "topic rails like an interest-group forum or subreddit." Speaking of, AI bans in space where the topic and ban aren't pertinent nor warranted feels off-putting. I'm glad I can blog about AI here without setting off tripwires tied to flimsy or nonexistent reasoning.
I hope readers find the Discovery feed's Hidden Blogs textarea useful. Helper text below the Save button explains how it works:
Hide blogs from your discovery feed by adding their domains above.
However, I didn't stick with it. The helper text continues with:
In order to keep the discovery feed account-agnostic, these settings are saved in your browser's cookies.
If I remember correctly, clearing browser cookies on exit made this a pain. Syncing hidden blogs between different browsers, devices, and form factors made it worse. Filtering through the feature ultimately felt more frictionful than filtering by myself.
What if the feature isn't as discoverable as it could be in the first place? Melinda's I had no idea you could hide blogs in the feed. starts:
Just added a handful of blogs to that little list. There are one or two blogs that only talk about "how to blog" and junk, this isn't Substack lol.
The funny part is that blogs like those have inspired fulfilling posts of my own. Hiding entire blogs from my feed meant missing out on gems I wouldn't have otherwise found. This meme comes to mind: Heartbreaking: The Worst Person You Know Just Made a Great Point. Them making a horrible point could inspire a great point all the same. It could even make the difference between whether a reply post gets seen or not, possibly breaking the chain.
Herman asked:
If you have any thoughts or potential solutions, send me an email. I appreciate your input.
Xaya made a similar ask in my reading habits and bear's discovery:
If you have any other ways of discovering blogs that filters out mainstream AI stuff or have a postroll/blogroll, please do send them to me, I'd love to look at it.
Sal's Solving Bear Blog's Trending discovery feed with this fix: use the Most recent page. He continues:
I've always found it more enjoyable. There are zero popularity contests or herd dynamics to contend with, and I guarantee you'll find some great posts that you wouldn't otherwise see. Plus, some bloggers (like me) opt out of the trending feed altogether by hiding the like button. Come hang with us! 😉
Except for hiding the like button, that's how I roll too.
If you find that insufficient, Robert's Bear discovery filter idea suggests a keyword filter for post titles. Question arise like, "How could it be gamed?" For example, this post doesn't include keywords salient to the discourse like "ai" or "discovery". You could just block my blog, but then you'll miss my posts on other topics you might be interested in. Other questions come to mind:
- Will it be more maintainable than the hidden blog filter?
- What would readers failing to practice indifference miss out on through it?
- How would it handle nuance?
Speaking of nuance, could erelong's suggestion from the Hacker News submission of Pirate's i'm tired of hacker news slop solve the problem?
Surely someone can vibe code a feed that filters posts that seem related to "Hacker News / AI style"
Yes, it's an AI-related solution for an AI-related problem. No, it wouldn't be native to Bear Blog. However, even primarily anti-AI people have takes on the technology which defy abstaining from it. This solution could better represent that nuance of wanting a non-zero degree of exposure to specific topics.
Should such a filter work well for people, they may find more AI use cases and experiences acceptable. What other ways can people use AI to meet their needs besides using it as a more contextual search engine? Even if a description of AI augmentation gets sandwiched by dunking on AI, at least that's a more informed take than blasting the innovation with abandon.
What's your proposed solution for fine-tuning discoverability on Bear Blog and elsewhere?