Ask HN: How might HN build a social network together?

218 points by shanebellone ↗ HN
I have a concept for a social network that would eliminate many pain points. I'm sure others do too.

How could we build a communal product for the public? Theoretically, this approach would result in a better product. Practically, it seems nearly impossible.

What are your thoughts?

Edit:

Let me give an example that I have been thinking about since 2009.

It requires a fundamental change from the reach model towards concentric social circles. The social network would allow users to arrange into small topical groups called social circles. These social circles would have a cap of 10 (arbitrary number) members. Each user could take part in many social circles. This inherently limits reach and therefore reduces the burden of misinformation, abuse, and moderation.

This model closely mirrors real social interactions and allows for both private and intimate communication. It also offers a profitable advertising opportunity. A social circle reflects its members’ interests and context.

398 comments

[ 1.6 ms ] story [ 355 ms ] thread
You’ve got one already, it’s called Hacker News.
I would argue that HN is a collective not a social network.
You're being social right now, on Hacker News.
You can even argue here!
I don’t know man. There are times when I genuinely want an argument to clarify my thoughts. There is zero argument- it just gets downvoted minus three. So then I sadly delete my comment. I usually assume since I’m not an American (wasn’t born here), am not able to get the nuances right, or maybe there’s some deep underlying social dynamics I’m unaware of. So better to just delete my comment and move on. These days I use HN primarily as mathoverflow - if I know how to work a problem I post. Otherwise best to shut my trap. HN isn’t the place for arguing- atleast not for me.
What do you like to argue about?

Personally I'm a big fan of the ChangeMyView forum on Reddit. Pretty much anything goes there, as long as you're civil and interested in productive discussion that aims toward clarifying your views. Which is a refreshing change from the ideological echo chambers of most of the rest of the site, where if you say the wrong thing you'll be banned.

Downvotes attract other downvotes - which doesn't say much for the quality of our, much trumpeted, gregariousness. A downvote is when someone doesn't like your opinion but they can't be bothered to discuss the issue. So a downvote is often a sign of disrespect and can be safely ignored.

Surely not everyone here is American - me for instance. If you don't get the nuance right then they can tell you - that way both parties learn something.

The following phrases are contradictory btw, thus:

"I genuinely want an argument to clarify my thoughts"

"HN isn’t the place for arguing- at least not for me"

HN is social media, but it's not a social network. A network puts the emphasis on individual nodes (users) and the connections between them (following/friendship).
Correct. I love that the focus of HN is on discussing topics, not discussing individuals. Of course a topic could be an individual, like Elmo, but that's not the focus of the site itself. I don't follow any individuals on HN, I just respond to whomever I'm discussing a topic with and I don't care who that is.
By that argument, email, text messaging and even phone calls are a part of social media.
Where does the network come in? I can't friend anybody, I can't tag anybody. There's no social graph, at least not one that users are aware of.
Mods agree - if bot replies became high quality enough to meet human standards here, dang is ok with allowing them. This site’s managers value the content quality, with no value given to social qualities
(comment deleted)
You're fooling yourself. We're living in a dictatorship.
What does that even mean?
Apologies, I’m coming in a little hot this morning… :)

I’d put it this way… you and I have HN. What about the protesters in Iran? We’re closer to that scenario than you’re prob realizing.

(comment deleted)
I think what would really be great is a social media site that doesn't offer direct image or video hosting. A bit like how reddit used to be. You just have link aggregation, voting, and comments. That's it. Maybe in a nice, clean, minimal UI as well. This would lend itself to tech folk I think. If only it existed.
so hacker news
I suspect GP was being facetious.
oh yeah, rereading it, it's pretty obvious actually, I guess I was being dense that moment.
I'm pretty sure that being facetious is against the community guidelines.

The post really should be flagged immediately.

(comment deleted)
Actually it could lend itself to a great business model as well — hackers both start startups, and often are looking for jobs at startups. You could use it for lead generation for startup founders to join your accelerator (once that concept exists), and an added benefit to founders would be access to hiring from the pool of hackers on the site. Win/win.

Sounds like bit of a schlep though. Would need to draw an audience e.g. by producing a ton of your own content early on. Lots of legal processes involved with operating an accelerator. Probably not worth it.

The good thing though: After the initial seed phase, where you produced your content you could just disappear and only be active on twitter!
no thats a great one…

can i add that to my list of ideas?

But how can i friend you? This is not social
Even just "follow" functionality would be great.
But it should be labeled frankly

"Friend" should be labeled "Manipulate"

"Follow" should be labeled "Stalk"

We have to be careful to not simply crowd-source the moderation either. Keeping a crowd-moderated social media site from devolving is a dang hard task.
Just curious do you have experience with this and if so, what was it
Great idea. To add: I think the site would also need a rock star moderator who diligently checks every comment and keeps the site clean. Also maybe it would be great to have a voting system where one could downvote as well! But only after a certain number of upvotes I think.
Let's also have zero notification whatsoever that someone answered to your comment so you have to keep engaged in constantly looking thru it manually
This is a very neat "engagement" feature idea! Love it. I would probably then just bookmark the site and check it constantly.
But no ads. Cause we aren't in it for the money.
...well, okay. A few ads, but only for semi-relevant job openings.
aha! good idea!

the web is… not free really… but more free than the wall gardens.

i think its biggest weakness is that notifications don’t really work there.

but that bug is now a feature. nice!

Maybe a really unintuitive system for ranking too? Like make it unclear why some posts are ranked higher than others, and why some are somehow greyed out even though there isn't an obvious way to downvote or report content. Also make it so you have to manually minimize the top reply in order to see the second most popular topline reply because there should be absolutely no limit to seeing the amount of subreplies.
The thing that gets me about the unintuitive system for ranking is that you have zero idea how long something will be on the front page. It could be 3 hours old, #4 on the page, with a few hundred upvotes and more points than almost all of the votes on the front page, and then just disappear an hour later.

And once something drops off the front page, it gets seen by almost no one. I keep trying to stop myself from spending too much time on comments (editing them for clarity and grammar, finding sources that people can follow instead of just relying on my claims, rereading the person I'm responding to in order to make sure I actually got what they were saying), because it's easy to spend 30 minutes trying to make a solid comment only to find out that the whole discussion has now disappeared and no one is likely going to read it.

I have no interest in what quantity of people end up reading my comments. What difference does it make to me if it’s a popular thread or an unpopular one?
Zero notifications, if such a thing existed, would help an online community stay constructive since the default behavior would be to forget conversations. This would help folks get some distance and perspective rather than being constantly reminded that they have an argument to win.

Another nice thing about zero notifications in our new social media app is that the development effort for that feature is very low.

> I think the site would also need a rock star moderator who diligently checks every comment and keeps the site clean.

My idea was to give people IP-like addresses like the early internet. And start off (until we find something better) with the ol’ Class A, B, and C networks of sizes roughly 16M, 64k, and 256 respectively.

The value I think is in having groups as small as 256 and each group has an admin/moderator. Why are we ‘automating’ people out of a job in the tech space when we could be ‘automating’ them into one.

Keep in mind, the existing IP networks stay. This is just a logical layer.

There are better ways to do it but I think this is an interesting start. There’s even a planned obsolescence built in as the world has more people than the 4 billion or so that a 32 bit address space gives.

Which to me is perfect. A chance to try some new things but a built in expiration date so nothing becomes too entrenched.

This would certainly be a site where hackers like me would come to read the news.
> link aggregation

One thing I've noticed over the years is that most social media companies have slowly started to discourage outbound links.

On reddit a surprising number of subs no longer allow you to just submit links, and a growing number require review before posting external content.

Even Twitter (before Musk) was clearly deprioritizing external links.

This is a bit troubling because it further leads to the "dead internet" where it gets harder and harder for people making interesting content independent of major sites to get visibility.

I think of the main reasons HN remains relatively high quality is its primary function is still aggregating pointers to elsewhere.

I'd thought Reddit subs prohibited links because it attracts spam and marketing disguised as organic recommendations. For Twitter, they promote organic content "for free" but want to make the inorganic stuff pay money for promotion.

Discord is close the original poster's idea for a social network, replace "social circle" with "Discord server" and lower the member limit for a server. Back to your topic, a semi-private Discord server/social circle moderator can ban anyone posting spam links so its not a huge problem there.

>voting

This always leads to a hivemind. Because even if you're supposed to use it for rewarding good posts, what the horde ends up doing is using downvotes as a disagree button. Especially if the admins/moderators have authoritarian powers to hide content on top of the users' individual powers or rate limit posting by those with differing views. Because everyone's bias sneaks in - no one is exempt from this - and it's very hard to see your own biases if you don't deliberately venture out of your filter bubble regularly. The best tool I've found for this so far (and I maintain a search) is https://www.allsides.com/unbiased-balanced-news It gets trickier when one side completely ignores a story, which both do.

Is this a problem that needs to be solved? No sarcasm intended.
What pain points? I can argue that the way HN is built today is the most effective for its purpose.
No info that someone answered to your comment is annoying compared to reddit.
If Hacker News tried building a social network, we'd spend the whole time arguing about every single micro-decision and getting nowhere.
Plus, the new site is useless because it doesn't work on SnowLinux Puppy v3.1 with HTML, CSS, JS and electricity disabled. Also don't like the scrollbar.
Nah, the complainer would clearly be BSD user
Re-enabled electricity and booted up my AS/400 to read your comment, which I curled.
Can I port to my Timex Sinclair?
There's no need to port it because it's a fine device. Are you one of those consumerist sheep that is obsessed with buying a new device every few decades? It works fine!
We can just make two versions, one is a reddit clone Next.js site backed by a boring database, the other is a dockerized CLI written in Rust that you have to deploy locally and persist on a blockchain, after you sign in with a hand soldered hardware authenticator, and where every post is written in XML with no Unicode support. Once a night a cron job transpiles and mirrors posts between the two systems, and you only see votes from your home system.

That would probably settle most arguments.

I am sure it will work in browser with javascript disabled.
Not quite a social network but a new version of the web where documents are just markdown or something very similar and the browser is an unglorified document reader with no JavaScript has been on my mind for a while.
Sounds interesting. Let’s call the markup language “Hypertext Markup Language”, and we’ll make it XML-like versus Markdown-like so we have at least have a well-designed spec to work with. It won’t be pretty, but it’ll work well enough.

And let’s support JavaScript, but you can turn it off if you’d like.

As someone that runs a platform for people to get to edit and control the HTML of their content I extra appreciate the snarkiness, but the point is to use a format humans can more easily/cleanly edit, and focus on document oriented display of data where the browser/reader decides/controls styling vs the platform and you don't have to fight back against a trustless code execution engine trying to attack you while you read a news article.

It would be a pleasant and interesting contrast to the giant anti-user Rube Goldberg machine browsers have turned into (err actually just two browsers both funded by the same company because the nightmareish complexity makes it impossible to do competing browser implementations).

Feel free to call my idea stupid (it is) but let's not pretend HTML and the current implementation of the web is some sort of perfect ultimate gold standard we can't improve upon.

I'm building a SaaS website similar to this. Unfortunately, we all like shiny things.
Personally, I think something like mastodon is better for this purpose.

You can post your HN related stuff hashtagged #hackernews and others might follow.

That being said - IMHO if you want a social network or community, don’t call it a “communal product”. Products are for making money and whatnot.

Should social media be a profit-driven enterprise as a civilization-scale product?
¯\_(ಠ_ಠ)_/¯

I mean this question is way too big for this sort of thread. I would say that simply calling it a “product” already reveals a certain ideological bias to begin with.

It’s like the difference between a city square and a shopping mall as a place for a society or community.

What is unique about this place is that many people here are more than qualified to build pretty much anything. So, it would be safe to assume that if we wanted something different, we would probably just build it.

Financial and intellectual resources are being diverted to projects with large markets but low utility.

Also, "product" is a semantic argument. Product also means sum. Every application is a product of technology.

Yeah, everything around us can be seen as a product. It’s all sausage factory across the board.

The opinion I’m expressing here is that excessively seeing things in that way when venturing out to do them doesn’t make inspire trust.

Fair point and I do agree. I have no interest in pursuing anything in the space. I wanted to start a discussion on the topic because I don't know how to fix this obvious problem. If a solution could be found (or even a broad direction), this group of people... would a probable source.
Given how things turned out, I think the answer might be ‘no’…
It almost has to be.

Running a world-scale social network requires immense computing power. Those bills have to be paid, and there's not going to be much difference between "break-even-driven" and "profit-driven" at that scale.

Disclaimer: I'm taking care of finance at Mastodon

Appreciate Mastodon comes to mind here.

I disagree with the second part regarding product not being the right term. Take Mastodon as an example: Mastodon is a non-profit LLC and an FOSS software but its community as well as the LLC clearly _produce_ something users are using.

Right term used where?

Because I’m not saying that it’s not a product. I’m saying that IMHO calling/thinking of something as a product upfront (as OP did) isn’t a good way to drive interest in it as a community.

Building a social network “product” is easy, entry level engineers build Reddit and Twitter clones during bootcamps.

I think the right question is “How could we build a network and keep people coming back?”

That’s the hardest thing to do, even with millions behind your back.

Ask Google Buzz, Orkut, iTunes Ping, Vine, Google Plus…

Vine definitely had people coming back iirc. Don't take into account Twitter's boneheadedness.
Vine was killed by Twitter. A foolish move given how popular TikTok is now.
I’m guessing the unit economics didn’t work. Different when a country like China decides its a strategic thing to do.
It was killed because the big names started talking unionizing.
> How could we build a communal product for the public?

If it would be a product, what is it going to be sold personal info, ads, what?

> What are your thoughts?

A question.

What problem(s) / do you want to see solved or what needs do you want to address with this?

(asking as someone who mostly never used social networks outside HN)

make a repo with a readme, put your ideas in it and invite people to comment
No thank you. I’m okay with not having another social media product.
> I have a concept for a social network that would eliminate many pain points.

And what is this concept? Hidden in this question is a strong undercurrent of "I'm an ideas person with a world changing idea. I can't share it because I believe that it's valuable. I just need engineers".

(comment deleted)
I just added it to the main post in order to help shape the conversation. I have no desire to lead a social media company.
Isn't that idea the same idea that google tried and failed at?
Google failing implies nothing about the tribal conceptualization of social networking.
> implies nothing

Nothing is a strong word.

A similar idea was tried, by one of the largest tech companies in the world, pushing it on to their already established user base, and it didn't catch on. Will it fail again? No idea, but it doesn't imply nothing.

I stand by my statement.

One might argue correlation, but any argument about causation is devoid of logic.

G+ actually worked, it just didn't hit as big as Google would have liked. I used it mostly every day for a while and was sad when it disappeared. So I guess you could say it "failed" for one set of criteria, but may not "fail" if the criteria is a little less demanding.
I think nobody would use a social network that would be created by the most popular ideas that float on HN. Established social networks are popular not despite their "problems" but because of them:

  nobody wants to manage their own data
  people are on social networks to be entertained not to "keep up with old friends"
  they want to be where brands, advertising and celebs are
  they want the opposite of decentralization, they want where everything is in one place
  addictive algorithms are what people like by definition, they are not created but discovered
  click bait, outrage, etc, again are what people are looking for
> click bait, outrage, etc, again are what people are looking for

I don't think people actively seek clickbait and outrage, but rather, are simply drawn to it when they see it.

what is the point of creating yet another social network unless you have a key feature to make it distinct and valuable? if you just make a twitter / facebook clone for tech people all you are doing is dividing yourself from other people. that might be desireable for some people but i dont think its worth persueing in the long term.

i propose we create something that addresses existing problems and adds new features. a social network not based on yet another website, not based on yet another app with centralized ownership. blockchain is the new hype and promises new features and value. lets make a social network that exists entirely on a cryptoblock chain. i am not talking about creating yet another coin. this is entiely outside of the scope of monetary tokens. lets use the blockchain technology to host text and link posts in a distributed fashion. lets come up with a way to give an incentive to host a copy of the chain. and lets make it completely decentralized. then you have a social network worthy of hackers

Blockchain would make it distinct, but would it provide unique value to users?

I certainly agree that we need some innovation in the space.

the added value would be decentralized and cencorship resistant. you will probably also not see a lot of ads
Unpopular opinion: any social network, no matter how it’s architected will become terrible past a certain size. Perhaps some architectures will have different “maximum tolerable sizes” than others, but all will become intolerable once there is a large enough user base.

More users, however, means more ad / tracking revenue, so these two ideas will always be at odds.

Agreed. Small, decentralised communities should be the way forward. We could call them forums ;-)
I miss forums. Discord has replaced forums in most of the communities I cared about and it’s been for the worse.
I hate this, and I like Discord.

Discord is a replacement for IRC/chat/AIM/other synchronous and informal conversation.

It's not a replacement for forums, archives, or formal communication.

Dealing with opensource projects via discord is a nightmare. One I use regularly decided to escape the walled garden and move to github discussions and it's been such a boon for search. Discord is a repetitive nightmare of doing support I don't know why projects use it.

The discord server still exists but questions generally don't get answered as much or if they do it's usually a link to github discussions.

So how do you combine social network + forum? What does that look like?
Heh, I just thought of an even smaller solution called newsgroups ;#)
There is/was already a social network for that - Ning.

Back when social network platforms became A Thing they adopted OpenSocial - they had a news feed, notifications, etc - but each social network was completely isolated. I haven't taken a look at Ning since I worked on social networking apps about 12 years ago, so I have no clue how they've evolved since then.

Is this really an unpopular opinion? Seems most people have acknowledged then general toxicity of large social networks by now.
I suppose it's not. The part which I _thought_ would be unpopular was that technology unequivocally could not solve this problem.
(comment deleted)
(comment deleted)
I think of HN as being a social network already. There is no need for another 'social network' in the world, actually less would be best.
It's more of a "thought" network or an "idea" network than a social network. Unless I am super motivated to click through to your submissions and comments (and then back read the parent posts to understand).. most of the interaction is limited to threads. So this limits any attempts to curate a theme or push a dominate ideology.
I think of HN (And really, any other website driven by user-generated contented content like reddit and forums) to be social media but not a social network.

I consider social networks to be a subset of social media where the focus becomes following specific users and having a feed consist of posts made by specific users, whereas reddit has a feed consisting on posts on specific topics.

There is really only a single barrier: adoption. You have to figure out how you can get thousands of users to join and then if it's a good product it should start to snowball.
We could offer orange checkmarks for $7 and undercut our competitors.
> The social network would allow users to arrange into small topical groups called social circles.

This sounds like Google+ : https://en.wikipedia.org/wiki/Google%2B

I loved the idea back then. As other commenters have pointed out, the structure and design of the network is not the biggest factor influencing its success.

I wasn't aware of Google's use of "circles" so thanks for the share!
I've been dreaming of a platform where each post is categorized (manually, by choosing a circle/channel, or automatically detected), as I'm not interested in all the content / themes of people I'd follow. Besides G+, I remember Pownce had a similar feature.

It would also be helpful to be able to choose your feed sorting algorithm - chronologically or else.

Google tried this, and failed. Actually they failed from their point of view: Google Plus had some good points over Facebook, but they wanted it to become #1 in no time, which simply couldn't be done because it was already too late. Had they kept the project alive, they could have benefited from the current situation at Twitter.
(comment deleted)