Keep on trying to build better social media

More Human #1: Erik Hoel, Jonathan Haidt, and the dream of a better social medium

Hello! Welcome to the first in a new series from The Good Teacher. In this series of occasional posts, I’ll connect the themes of typical Good Teacher dispatches with reflections from others on Substack on topics important to our world.

My goals are simple: to promote and engage with the work of good writers and interesting thinkers—and learn something in the process. Today’s essay, the first in the More Human series, considers everyone’s favorite topic: what’s wrong with social media.

---

Edward Hopper, Nighthawks, 1942. Courtesy of the Art Institute of Chicago.

What’s the problem?

Research by Jean Twenge, Jon Haidt, and others shows us that social media is pretty much perfectly crafted to mess with us, and mess us up. (Twenge and Haidt have an overview of all research in this field available here.)

If your experience is anything like my own, you really don’t need research to teach you this. We know something’s off.

There are business reasons for this. Cory Doctorow describes the dynamics and economics behind this process, which he terms “enshittification”:

Here is how platforms die: first, they are good to their users; then they abuse their users to make things better for their business customers; finally, they abuse those business customers to claw back all the value for themselves. Then, they die.

The process has been documented extensively, and it has happened at pretty much every major internet platform—Doctorow focuses on Facebook, Twitter, Amazon, and TikTok.

But there are cultural reasons for the enshittification, too. And due to these socio-cultural reasons, Erik Hoel, a neuropsychologist and writer with a fascinating Substack, The Intrinsic Perspective, suggests that it’s just not possible to make social media better:

Stop trying to make a "good" social media site

A summary and a few quotes follow, but please go read the whole thing. It’s not long.

Hoel, like Doctorow, documents a similar process that takes hold of each and every social media site, tracing the same progression everywhere from the earliest sites to the latest wannabe Twitter replacements:

For a new social media website, going from “omg it’s so great we’re inviting another 5,000 people!” to “we will beat you with hammers” takes about two weeks.

Hoel argues that it’s simply impossible to avoid this outcome, because

what you are trying to outrun is human nature. No design of social media can get rid of what I like to call the “semantic nadir,” which is what you’ll inevitably experience if your tweet ever goes viral, wherein eventually someone will take your tweet in literally the worst possible way.

No matter what we try, Hoel argues, we keep on failing, and we’ll always keep on failing, because the fundamental problem is that social media appeals to bad, but deeply seeded instincts we have to develop in-groups that we ingratiate ourselves to via virtue signaling, which we zealously protect against outsiders through denunciation. Hoel suggests:

Twitter and Facebook, by creating an all-to-all connected web structured and assisted by viral algorithms, has beaten Dunbar’s number and resurrected the ability to run culture and society mostly through gossip. Mark Zuckerberg, originally just trying to create FaceMash, a website for ranking the hotness of Harvard girls, in his second attempt with Facebook accidentally summoned back our first, and perhaps most consistently oppressive, form of government.

You can’t escape constraints on who to engage with—social media just replaces civil society’s mediating institutions (neighborhoods, religious communities, local government) with new gossip-mechanisms that reward the loudest, most divisive voices.

(A recent guest post on Haidt’s Substack, written by a former Facebook employee, documents in detail how Artificial Intelligence is likely going to make things even worse. I can’t wait!)

---

Architecting a better social medium

I’m not so sure Hoel’s right. I think it might be possible to create a healthier social medium. To do so, it would be necessary to work against these bad but deep rooted instincts by providing different incentives.

Here’s my proposal, framed in terms of seven key problems with social media and how to fix those problems:

I. Fighting with evil strangers about politics

Problem: We tend to behave ourselves when we’re around people we care about. But social media seems great at encouraging people to be nasty to each other, and give us all many opportunities to do so. Trolls take over, and lure others into trollishness.

Solution: I agree with Hoel that Facebook & Twitter’s combination of gossip / rule-by-the-cool-kids, when combined with their all-to-all connection structures and viral algorithms, is deeply oppressive. Haidt and Twenge have extensively documented how social media is especially damaging to vulnerable teenagers.

So start by limiting each user’s social graph to 150 people.

A good social medium should have as its foundation people you already know IRL—your friends & family. A good social medium should also deliberately on-board new people, especially young people, by providing a supportive, limited environment as a starting point.

In short, social media should stop trying to replace society, and supplement it instead. Assume small and local by default, but keep channels of exploration and discoverability open and allow users to tap into those as much or as little as they want.

II. Conspiracies and disinformation everywhere!

Problem: One person’s “disinformation” is another person’s truth. One person’s conspiracy theory is another person’s basic understanding of reality. Both predate the internet, but they really seem to flourish on social media, where you can find a group of people who will support whatever crazy theory you want to propose—and another group eager to denounce you as the antichrist.

Solution: Set up a tiered feed that deprioritizes national and global news and prioritizes local issues that are both intrinsically important and verifiable.

• Friends/Family – users’ home feed. Once you’re caught up with posts in that feed, you can move sequentially to higher tiers (or, as I like to think of it, moving down into progressively lower levels of Dante’s Inferno):

• Local – your local feed would give you information about schools, local government, civic and cultural organizations, local businesses, etc. Because you see all content you subscribe to, local orgs would know their posts would appear in all followers’ feeds—and wouldn’t have to worry about cultivating support from users who weren’t part of their local target market.

• Cultural – where you could follow writers, sports, music, and other culture that you care about.

• Regional – the regional feed would be similar to the local feed, but at roughly the state level. Follow your state politicians, your big-city newspaper (if it’s not your local newspaper), and other regional interests.

• National/global – national and global news would be put in its place, relatively far away from users’ default and most easily accessible feeds. Still available, but deprioritized. The NYT shouldn’t be tasks with being the arbiter of truth for the universe (however much it might like that role).

A social medium set up around tiers like this would inherently limit the ability of information to “go viral,” because users would see less of it. Things you read in lower circles of hell lower tiers would stay there, even if you shared them, and your friends would only see them when they descended to that tier. Or perhaps only if they followed those same outlets already.

And there’d be less performative virtue signaling by people and institutions if they weren’t worried about constantly performing on a global scale.

Would there still be disinformation? Yes, of course—there always is. But its reach and impact would be limited.

III. Kardashians FTW

Problem: Content that is rich in “virality” isn’t necessarily rich in truth. Bullshit often goes viral. Good stuff often struggles to do so.

Solution: A benefit of the tiered-feeds approach: it would impossible to “go viral.” You couldn’t rely on simply being a provocative asshole to build an audience. You’d be more accountable to—and have more incentive to engage respectfully with—friends and family, even when they disagreed with you.

Frankly, I think most people who are “successful” on current social media would hate this aspect of a tiered-feed social medium. Which is exactly why we need it.

Instead of virality, a healthy social medium would build in opportunities to build your audience organically. Want to write? Or create art or music? Just start doing so. With the support of your friends & family, you could level-up to the local or cultural tiers, and keep going.

IV. Who walled off my garden?

Problem: Every social medium has eventually become a “walled garden” that seeks to keep you on its on site, so it can serve you its own ads. Walled gardens limit the power of the Internet, attempting to make their own sites the only sliver of the Internet that users ever experience.

Solution: A health social medium should support the “open web”—it shouldn’t try to keep you away from the Internet, but rather be a portal through which you can explore the parts of the Internet most relevant to you.

You should be able to pull in and link out to any other site you want—unlike all current social sites, which constantly try to prevent you from leaving.

Similarly, if you’re building a public-facing business or organization page, blog, etc., you shouldn’t have to build separate sites for each social medium. Build one here, and it will be accessible to everyone—whether or not they’re a member of this particular network.

V. Social media addicts anonymous

Problem: It’s ridiculous to even have to type this, but a good social medium should not profit from addicting its users.

Solution: Build in, allow, and even encourage time limits per tier.

A healthy social media should allow you to set your own time limits and restrictions on each different tier of content. You might want to limit your national news intake to 30 min a week—go right ahead. You might want to see all your posts from friends, family, and your local community, but limit your time on other tiers to 10 min a day—great!

Social media shouldn’t be paid according to how much time you waste, but rather on how much time you spend achieving your aims in life—keeping in touch, learning, building a business, experiencing culture that you love.

(This decision has implications for the business model, as well. See point VII.)

VI. Pandæmonium

Problem: As Umberto Eco said, “Too many things together produces noise, and noise isn’t a tool for knowledge.”

Solution: Look, many of our minds are already so addled and confused that we’d hardly know where to start organizing the onslaught of information we face daily.

But current social media orgs have abandoned the very idea of letting their users define what they see. My Facebook feed is polluted by stupid “recommendations.” Twitter constantly pushes their “For you” algorithmically recommended feed.

Yet we all already have way more sources of information than we can possibly handle. I’m not super active on social media, but I have Facebook, Twitter, and LinkedIn accounts, this Substack newsletter, Micro.blog, RSS feed subscriptions, and subscriptions to exactly 189,727,465,932 email newsletters.

If a social network embraced the open web, users could choose to funnel most of their content through the network, and filter it all through the same hierarchical structure, so they could engage with what they wanted, when they wanted it, and fit it meaningfully into the structure of their daily lives.

This tool would be as much a web browser as a network—an app or even a social operating system more than one more website to check.

VII. Laughing all the way to the bank

Problem: Social media get paid when you stay on their site, so it’s in their interest to try to addict you, feed you content they think will enrage you into engagement, and force feed you new content even when you haven’t asked for it.

Related problem: The loudest, most caustic, divisive, and controversial voices have an inherent advantage relative to the calmer, more reasonable, more charitable ones.

Solution: The only way to build a social medium that won’t just start healthy but can stay healthy—the only way to avoid the “enshittification” of the medium—is to ensure that it has a sustainable business model from the start.

A health social medium would allow users to access the friends and family feed for free. But using it at other levels would require a minimal monthly fee. And for companies that set up pages for regional, national, or global reach, there would be reasonable but higher fees per tier.

People and businesses seeking to build audiences at higher levels—for instance, a successful small business at the local level that was looking to expand its reach to the state level—would be able to pay for access to the higher account tier. Perhaps they could also pay to appear on the “discover” page for related, successful pages at that tier.

In short, you pay more depending on the size of the audience you’re reaching. The network thrives when you thrive.

---

So, who’s in?

Hoel concludes by comparing social media to cable news:

Eventually, all media dictatorships end. Some people of my generation, the millennials, still watch cable news, but no one from my generation would say that they fully trust cable news the way that older generations do. We just don’t take it that seriously. […] I hear that Gen Z spends a lot of time on semi-private forums like Discord, preferring that over the huge, everything-all-at-once web of connected accounts that is social media. Maybe that’s the beginning of a shift, and so we can eventually, slowly, recapture a bit of what was lost. But first people have to stop building new social media sites, and go build something else.

Good! But semi-private Discords have their own problems, too. And not just that low level secret agents use them to leak classified docs. They can lead to their own echo-chambers, tacitly training people to engage only with those who think like them, with the same interests, beliefs, and assumptions about the world.

We need a system that strikes a balance. Social media that can allow you to escape the at times stifling confines of local, gossip-fueled society with interest-based communities through the scale of the Internet; and that can equally temper the echo-chambers of special interests with a foundation in our friends, family, and community.

It may not be paradise, but at least it wouldn’t be hell. So, with that exciting pitch, who’s ready to get to work?

---

Thoughts?

I’d be glad to hear your thoughts on this proposal.

• Do you think there’s any hope for social media? Or do you agree with Hoel that we should stop trying and deal with what we’ve got, or opt out entirely?

• Would filtering information through a tiered, hierarchical feed strike a good balance between local, gossip-drive community and mass Internet all-is-equal-to-all chaos? What problems or conflicts would this approach create?