Did you use Google Plus? I found the circle system to be so good for this. I like following artists, and sometimes they can be exceedingly political/depressed/self destructive and I can't take them in my "main feed". With G+ I could drop those people into a circle and then "dip into" the madness for a bit and hunt for gems, then leave before becoming overwhelmed.
Being able to "sort people" into groups on social media (using a single account) was just excellent. Here are my programmers, here are my RPG people, here are my RPG people that have a different core belief than the previous group of RPG people, here are my Artists, here are my Writers, here are my Sad But Brilliant people.
Then when they added collections so I could present the different "facets" of my interests and allow my followers to unsubscribe from those facets was excellent. I'll tag this post as gaming, this one as music, this one as bullshit hot takes, and you, my follower can say "I hate his taste in music" and never see it.
I miss it so much.
"Here is a group of people who are creating things and having great conversations with each other. Check it out."
It was like someone opening a hidden door to a very nice party.
Don't know how to make something like that non-abusable or how to mitigate the abuse that killed it, but then I never saw it because I wasn't following spam circle shares.
When Google+ came out, I do remember avoiding it mostly out of spite, because Google were so overwhelmingly aggressive in trying to make me use it. And these days I'd probably never type my personal info into any Google property. Still, sounds like a shame it failed.
Your suggestions for being able to apply our "higher order executive function" to our social networks is honestly a great take, and a great suggestion. Unfortunately though, that's not something I can really build, even if I had the time, resources, and personnel. That sort of social network requires critical mass of content and would by definition almost, have to be a competitor or aggregator of other social networks. However, that's not the sort of social network I personally am interested in building either.
I think what you're describing definitely falls into our modern definition of a social network, but I also think that definition falls very short of the potential for technology to facilitate relationships. In my opinion, the "thirst traps and rage baits" totally have there place (they're a lot of fun!), but I don't think their place is in the same app as your best friend sharing about their mixed feelings of loss and euphoria as they move from one job to another.
That all being said, I think you bring up a really good talking point, one that I hope the people at Instagram, Snap, TikTok, etc. all pay attention to. As for me though, there's nothing I can really do about it other than start discussion.
So, for something like your network, that might mean rather than limiting how frequently people can post, give users the option to aggregate content from frequent posters in their feed. So, maybe I have a switch I can toggle that says, "show a unit of content from one person no more than once per day", where that single block of content is expandable, and gives some aggregated summary of how many things they posted / what its content was, that I can drill into if I want to, or quickly skim over if I don't care.
I swear, people are too busy re-inventing the wheel when it already exists. use one of the literal hundreds of free services to write what you want and syndicate it. and use the same system to ingest exactly what you want.
We hit the nail on the head 2 decades ago. It's been rusting since.
Software exists to translate RSS feeds to ActivityPub actors (ex https://github.com/dariusk/rss-to-activitypub). Going the other direction, Pleroma provides RSS feeds. Not sure about Mastodon or Misskey.
Oh yeah, I could. I guess I'm just believing more and more in the power of a familiar interface. I feel somewhat comfortable hopping around different sites and creating all sorts of accounts (on my computer, less so on my phone), yet I imagine many really don't want that hassle. If the comment box were the same everywhere, that could help, but still, I find sometimes if it takes that extra effort, I often won't do it.
> Software exists to translate RSS feeds to ActivityPub actors (ex https://github.com/dariusk/rss-to-activitypub). Going the other direction, Pleroma provides RSS feeds.
I may check it out for some of my sites, thank you.
The reader can be used without login. To comment and share, you need to login to https://roastidio.us (free open registration)
Warning: I have very few users. So you will still feel isolated, unless you can gather a community yourself.
Two things come up for me:
1. I'm curious, how do you imagine people will discover there are comments on a specific URL? Will it show up in their browser? An embed that sites can use to show it on their site? Other?
2. I found that the name made me feel a little confused and hesitant to engage, because I didn't want to roast the article I submitted. I get the feeling you started in one direction and may be opening up, but I'm not sure.
2, Admittedly the name is somewhat provoking. My point is to encourage different opinions. You have to find some value it the stuff that you roast or you will not spend your time commenting, right?
2, I personally see roast as one type of emotional attack or speech act (still looking for a better term for it), along with praise, question, supplement, etc. I don't roast things much but I do the others so personally I may feel a bit repelled by the name. However others may feel attracted by it, there are lots of us internet folk :-) good luck!
For example, on Twitter, I love having the "latest tweets" option. But maybe I want to also have "only tweets from verified people with over 100 retweets from other verified people." This is somewhat possible on Tweetdeck, but not possible directly within Twitter itself.
It reminds me of the Facebook Graph Search, which I loved—"my friends of friends who speak Spanish who live in San Francisco." But FB shut that down a few years ago.
Makes me wonder what are the main reasons they don't implement such options to help us better choose how we see the data—is it technically difficult to implement? Does it run counter to the business model of placed advertisements? Does it really just unleash too much power in the hands of the users and make these companies afraid what might happen?
I'd love if they did implement the changes you suggest and I wonder why they haven't.
Social media companies spend incredible resources trying to continuously improve the stickiness of the app so users keep coming back to it. Giving users total control over curation creates two big problems for this.
1) if I can cut out all of the BS in the feed, I can utilize the app more passively which likely reduces overall usage. I would try to get my social media into a weekly feed, which is not compatible with the business social media companies are trying to run. It's all about DAU's.
2) total curation makes researching and optimizing the UX for eyeballs more difficult because you introduce an entirely new set of variables into the equation. This increases the likelihood that behavioral research draws erroneous conclusions based on the patterns being observed. By controlling the drip of the content to the users, the company can control for this much more easily.
Fair point, some may use it much less. Ironically, I might use it even more, because now I can explore more. Being given only one path can bore me at times. Sometimes I want to choose my own path. I absolutely loved the Facebook Graph Search, but then it vanished and I'm back to just hoping that Facebook gives me what I want instead of letting me dive deeper. In other words, I spend a ton of time on Wikipedia rabbit holes even without an algorithm that tries to predict what I want.
> 2) total curation makes researching and optimizing the UX for eyeballs more difficult because you introduce an entirely new set of variables into the equation.
Ah, yeah, I agree with this. Fewer variables make it easier to run social experiments and behavioral research.
Rather than anything nefarious, I think it might simply be down to the fact that it doesn't justify the engineering effort it would take to create a feed algo store or repo like you suggest considering how few people might actually use that feature.
The vast majority of people out there likely don't have the wherewithal, time nor inclination to carefully pick and tune the algo for their friend feed.
Makes me wonder, how many people do you think have filters/rules/folders on their Gmail or other email inboxes? I assume a majority have some but maybe not.
When they introduced the feed, Facebook gradually became worse at all of those things. Facebook became less about being a useful tool and instead became a media sharing instrument for serving advertisements. Your entire post is about that aspect of it though, and really it's the problem with all social media. The feed design is fundamentally incompatable with providing the utility that was initially offered by Facebook. The feed is purely a tool for distribution and consumption, not forging connections or otherwise improving the lives of users.
Currently I suspect the logic behind social media engagement algorithms is something like "feed the user more of what they have clicked on in the past". But in real life, I don't just do more of what I did in the past. Sometimes I choose to do different things. But in legacy social media like FB I can't choose to be offered anything different than what I clicked on in the past. I can't ask for the culture wars and political hot takes to fuck off so I can see normal friends and family posts (or vice versa).
This lack of choice in what we're offered is what's essentially inhuman about legacy social media.
The competition would be innovating better personal agents and data tagging.
I think TBL found the mother of all misunderestimations when they made http servers and browsers as step zero.
Get real.
You could, but most people download neversinks or someone else's custom filter.
Particularly the rage bait: I basically had to quit reddit because I realized the extent to which I engaged in rage bait made me a worse person.
In a weird way I actually think that the weird stuff twitter has been doing to try to make it look like they're doing something about unhealthy social media usage (all the "remember the human" stuff) would probably be more effective on a reddit-type social media platform.
I deeply appreciate Reddit's openness to customizing the experience. The defaults suck, I agree, but you can choose to curate the experience how you like. Moreover despite lots of jaw clenching over financial incentives, Reddit still offers the Old Reddit interface _and_ an API which lets power users consume Reddit however they like. These values were pretty common among other fora/BBSes at the time. The experience one got on 4chan's /a/ or /jp/ was very different than /b/ or /pol/ or /soc/ (and most of 4chan's famed toxicity came from the latter 3 boards.) Forums were usually partitioned by topic.
> In a weird way I actually think that the weird stuff twitter has been doing to try to make it look like they're doing something about unhealthy social media usage (all the "remember the human" stuff) would probably be more effective on a reddit-type social media platform.
I don't think the algorithmic feed as-is is compatible with healthy usage, no matter what rhetoric a company uses. ML models need to not only take engagement as input but also some sort of personal toxicity input. I need to be able to say "content X is addictive to me but not for a good reason." The model also needs to offer as much transparency as possible to the user, offering things like partial dependence plots for auditing by the user. I'm not sure whether a ranking model trained that way can still provide good ad targeting performance. Moreover knobs like this are probably just too complex for a layperson to interact with. Many engineers don't understand how a partial dependence plot works let alone laypeople.
My problem with social media is about the lack of choice. I have no qualms that the majority will continue to use social media the way most people in the past would spend time mindlessly flipping TV channels or window shopping at malls and boutiques. But there was always alternatives. People that didn't want to spend time window shopping could be found in libraries, bookstores, bars, playing sports, or outdoors. Once you became an adult this was expected. That didn't mean I was ostracized from them. I encountered people in college who hung out in very different places than me but still connected with me over shared interests.
In the social media age every form of social media is siloed. If you're not on Twitter you can't engage with people on Twitter. If you're not on Facebook you can't see people's Facebook content. You _have_ to join the platform to reach people on the platform. I'd love to see a world of federated social media where we can all talk to each other no matter our hobbies and third places.
Although I'm sure it's not perfect, it's not terrible either. Especially if you follow the perennial Reddit advice and unsubscribe from the massive default subs.
Reddit is also just an awful hive mind.
and yeah the hivemind stuff is pretty bad. I'm reminded of a talk that ViHart of all people did where they argued that in a McLuhanist "the medium is the message" way, reddit's focus on up and down votes and karma creates an extremely judgy and kind of adversarial experience that eventually breeds hiveminds. I'd also argue that reddit's focus on small communities managed individually by specific moderators creates an undercurrent of site drama that reliably rears its ugly head every few months or so, but I guess that's neither here nor there.
The problem of addiction is the problem of inter-temporal choice, and a broken discount function. Most addicts don't want to keep using in the executive part of their brain. They know it's self destructive, but they keep doing it anyway, because the next hit is always, in a very local sense, rational. The ability to bind your own hands is the way out.
Now, if we put some regulation on dopamine loop limitations, such a network might absolutely THRIVE!
Just my personal take! Do I wish people engaged more actively with the idea that these products can be living breathing things that we also can exhibit influence over? Absolutely. Do I think the critical mass of such users exists today? Sadly not, but I hope one day to be proven wrong!
IMO the right solution is personal executive autonomy over content presentation and algorithmic optimization. By which I mean, I want to be able to make an "executive" choice over what kind of experience I want to have on a social platform. By "executive" I mean, I want to be able to make that choice explicitly, not implicitly by my behavior. For instance, an implicit choice is walking by a beautiful chocolate cake and being unable to resist eating it. An explicit choice is being able to architect my environment in advance, such that there are no beautiful chocolate cakes to tempt me.
The biggest problem with social media right now is that the incentives of the existing platforms do not allow people to make these choices explicitly. They do not allow us to craft our information environment using our higher order executive functions, they force us to do it using our reptillian brains, one moment at a time. Essentially, they force us to choose: either you get none of what social media has to offer, or you take every aspect of it, whether you like it or not.
Being able to explicitly say "for the next hour I want to be fed thirst traps and rage bait, but then I want that turned off" would be an amazing feature. People have a right to that kind of content if they want it. But they equally have a right to engage their higher cognitive functions and choose not to be expose to it if they don't want it, or choose to be able to architect the manner in which they are exposed to it.