Re: Who Likes the Likers

CircaDian:

> [...] I think a low-barrier-to-entry feedback mechanism is important for Gemini, then briefly mention why I think “thanks” is the right action and “like” is not.

>

> What I mean by “thanks” vs “likes” is that “thanks” are a private message to the post author, while “likes” are public and might be used for filtering/discovering content.

Thank you Morgan for the suggestion. This is certainly food for thought.

The gamification of social media and the resulting negative downstream effects are something we all should be keenly aware of. The larger a network becomes, the more the forces multiply and this becomes bigger and bigger of a problem.

When it comes to Bubble/GS.org, scale-wise there's plenty of time to consider appropriate improvements. I'm thinking the classic Like mechanism isn't too bad in this context, but making it a step more private — and perhaps psychologically healthy — is to hide the names of the likers from everyone except the post author.

Discovery algorithms are a topic unto themselves, but I don't think these become relevant until the scale is so large that the feed of posts is too busy to keep track of. I will keep thinking on this, though.

🕹️ skyjake [sysop]

2023-05-15 · 1 year ago · 👍 gritty

2 Comments ↓

☕️ Morgan · 2023-05-15 at 10:09:

Yes, a good point about scale.

I once spent the best part of a year building a system that as designed could not possibly scale to handle a large rate of requests; and that turned out to have been the right call when in the end it only had to handle a few hundreds of requests per day. It was even considered a success--because every request was very high value.

So, "wait and see" sounds good to me. Thanks.

🐵 cquenelle · 2023-05-15 at 14:35:

I suppose a site could have a per-user list of public labels and another list of private labels. They could start with a site-configured default list. That way you can control what kind of simple feedback you’re interested in. A subtle implication here is that the site doesn’t know which ones are positive and which ones are negative. So you’d have to aggregate based on which posts got ‘reactions’ without implying whether it was good reactions or bad reactions.