An optimist’s case for the future of social media

We're surrounded by evidence of a better social media landscape. But what is "better"?

(note: thanks for your patience during my lull. I’m returning from pat leave. Thanks for the well wishes!)

Do I feel good about the future of social media?

That was the key question running through my mind after reading Superbloom by Nicholas Carr, and shortly after, a self-explanatory post titled “You should quit social media for good” from G. Elliott Morris, author of the outstanding and enjoyable Strength In Numbers Substack.

The subhead of Superbloom snagged me: “How Technologies of Connection Tear Us Apart”. I’m deeply interested in the concept of atomization, or the idea that innovation is an endless, inevitable process whereby new technologies steadily break society into steadily smaller and more niche groups. After reading Superbloom, I believe Carr and I approach the concept of atomization from different angles: whereas I investigate the division of people into physically smaller groups, Carr covers in Superbloom the mental and emotional cost of those technology-driven subdivisions, with a large focus on social media.

Carr has been sounding the alarm over technology’s impacts on our individual health and societal wellbeing for years. He’s written numerous best-selling, well-researched and thoroughly cited books to make this clear, most famously in The Shallows (2010), in which he warned about what the internet was doing to our brains. He’s a digital media critic, to be sure, but he’s no simple hater.

But suffice it to say, if you want a thoroughly researched pessimistic viewpoint on on the impact of social media on people today and in the future, look no further than Superbloom. I was particularly bummed out by Carr’s repeated assertions that most social media users are unwitting drones, duped by the platforms to hand over their attention. I’d like to think that, if asked, Carr would agree that many people get value from social media platforms and use them willingly.

I did enjoy Superbloom’s historical accounts about how government and society dealt with mega media technologies in prior eras, like how radio (with eventual government involvement) evolved from a wild west chock full of hokey health misinformation and dangerous practical jokers into a useful and mission-critical medium - useful context for today.

Superbloom lacked a nuanced take about how to solve all this. In the waning pages, Carr quickly introduced a half-baked idea: just force all of the social media companies to introduce friction into their products, thereby making them worse, less enjoyable and less pervasive. This idea is neither good nor holds up to scrutiny. It’s politically and legally unrealistic to pass and enforce, would make millions of people really angry, and would kill incentives for growth and competition.

But the thing is, Carr is not alone in his pessimism and skepticism; see G. Elliott Morris’s reasonable, well-researched and truly thoughtful post, irregular from his typical work in politics, in which he includes some clear best practices and ideas about how he limits social media’s impact on his work.

Morris’s ideas focus on how to get off and stay off of social media. Morris is not the first and certainly won’t be the last to echo such perspectives. They also don’t solve the problems of social media. (No fault of his; that wasn’t the point of his post.)

Do you know why Carr, Morris, Tim Wu, the former Surgeon General, and so many others lack solid, interesting ideas about how to improve the future of social media? Because it’s a really hard problem! It intertwines legality, business models, product design, millions of use cases, and how to define “social media,” now and in the future. Improving social media cannot be boiled down to one fix, one law, or even a few.

And while evidence truly exists of a better future for social media, that evidence is hard to find, hard to judge, and frankly, hard to believe in after a decade-plus of social media ideas that went nowhere or got acquired before they gained any real traction (a model now OKed by the Supreme Court).

Given the scale, scope, complexity, and the sheer market power of the incumbent players (now backed by the Supreme Court), it’s understandable why Carr and others would feel such cynicism. It all feels unsolvable.

It’s within this context that I offer modest, grounded and realistic optimism for the future of social media. While no one can predict the future, we have ample evidence here and now in 2025 of real, working products that point to a better social media future.

Importantly, though: what does “better” mean? And for whom? I thought a lot about this. “Better” must mean improvement for individuals with business viability for companies. Good ideas mean nothing if people can’t form profitable companies around those ideas; as of yet, no social media network has been either a charity or a public enterprise, and I suspect none ever will.

I believe “better” in terms of the future of social media that raises the ceiling on good outcomes for people, and where companies can still make money, all comes down to individual agency.

Individuals wished they could exercise agency over many elements of the social media experience: control over their identity, their followers/data, their content, how algorithms display content and rank engagement, how to find their tribe while protecting themselves from negative outcomes, to name a few. How people exercise agency over those choices meshes entirely with how social media platforms retain users, distribute content, and make money so that they can survive as businesses.

Today, at a high level, users of social media platforms possess a low quality of agency: do I or do I not want to use this platform? Once the social media platform possesses my data - content, followers, engagement data - do I or do I not want to continue engaging with a feed and an advertising experience constructed around that data? If I post, do I want this content to be viewable by just my followers, or the entire internet?

In this case, the opposite of agency might be platform lock-in, a familiar concept to today’s social web. People care (and handsomely profit off) the identities, followerships and content they’ve produced on social media platforms. One cannot take the identities they’ve built on Instagram and port them elsewhere. Such lock-in keeps people around, giving the platforms complete control over the experience of its users: if you don’t like Instagram’s algorithm, their ad experience, how its feed displays content, and how your posts can be viewed, it’s mostly a take-it-or-leave-it experience.

But what if individuals could exercise agency over elements of the experience on Instagram, or other social media platforms? Such agency could radically shape fundamental elements of social media platform use:

• Choosing a feed objective (friends first, local news, NBA news only, no AI content, open carnival, etc.) to set what rises and what recedes, rather than accepting one engagement‑max default

• Scoping each post’s audience (room only, followers only, close friends, public with quotes off, etc.) to reduce involuntary exposure and invite context to travel with the content

• Controlling replies and quotes to limit raid risk without closing the door to discovery

• Carrying a portable identity and follower graph, so that you could try out or explore other platforms without starting from scratch

• Saving a portable taste profile (hide dunk replies, boost domain experts, de‑dupe cross‑posts, language preferences) to carry tastes, preferences and values easily across clients

• Turning on trust‑visible mode per context if desired (surface provenance badges, downrank unverifiable virals in civic contexts)

• Shared blocklists and room safety presets (no mass mentions, slow‑mode for newcomers)

• Choosing small‑n rooms for sensitive topics, then graduating a post outward with context when it is ready

• Controlling monetization touchpoints you see (ads off in rooms you pay for, sponsor slots only from approved lists)

Platforms have myriad monetization opportunities in such a world:

• Premium communities: paid rooms for brands, creators, clubs, and schools, better tools, better safety, fewer ads

• Premium feeds

• Contextual sponsorships/ads that are way, way way more targeted and personalized

• Memberships and commerce: tickets, courses, merch, and tipping inside communities; platforms take a simple processing cut

• Trust & safety bundles: communities pay for stronger moderation, shared blocklists, verified org IDs, and faster support

• Discovery & curation tools: power users and publishers pay for cross‑network search, curated channels, and audience insights

• Identity & recovery: paid verification (this isn’t going anywhere)

• Managed hosting: groups that don’t want to self‑run pay for reliability, backups, compliance, and easy migration between apps

In a world of greater individual agency on social media platforms, the experience is not one-size-fits-all. Individuals have the option to choose their preferred settings, speeds, and norms, distribution and visibility, safety and trust, provenance and content quality, engagement defaults and feed - you get the picture. Platforms can explore a diversified set of monetization opportunities that could even include more targeted and personalized ads. The path to “better” does not require any morality shifts. These are product and technical upgrades.

You can experience such levels of agency right now on vibrant and growing social media platforms:

• Ghost lets writers run a publication on their own domain with full control of the mailing list; you segment access (free/paid), export subscribers, and choose where posts syndicate. Monetization flows through subscriptions you set while Ghost charges a predictable SaaS fee instead of taking a revenue share.

• Pixelfed is photo sharing with chronological timelines and per‑post privacy options such as public or followers‑only; creators curate albums and collections and share to selected audiences. Operators and creators monetize via donations, prints, or paid storage tiers rather than targeted ads.

• Lemmy runs Reddit‑style forums where each community sets rules and users choose sorting such as Hot, New, or Active; volunteer moderators enforce norms close to the topic. Funding typically comes from donations, sponsorships, or paid tools for community leads rather than a central ad marketplace.

• PeerTube hosts video channels across many independent sites; creators pick a host, publish to their own channel, and viewers subscribe directly rather than a single global homepage deciding what wins. Monetization usually flows through memberships, donations, or channel sponsors managed by each host.

• Geneva provides group infrastructure for clubs and neighborhoods with chat, forums, and events; organizers set join gates and reply rules, and members share small by default, then widen on purpose. Revenue comes from premium community features, paid events, and brand‑run groups.

• Are.na hosts a visual research network where people collect links and images into shared channels; creators choose public or private collaboration and credit flows to curators. Member subscriptions fund the service; opt‑in partnerships can support larger projects without flooding feeds with ads.

• Glass offers a subscription‑funded photography community focused on feedback; photographers control visibility and request critique without public like counts. Monetization centers on member dues, with room for paid critiques, workshops, and curated sponsor showcases.

• Micro.blog enables lightweight publishing on your own domain with optional cross‑posting; you own the site and decide distribution while public follower counts stay out of view. Revenue comes from low‑cost hosting; writers layer in memberships or sponsor notes as they grow.

• Letterboxd builds a film community around logs and lists rather than a single engagement‑max feed; users follow trusted voices and curate watch lists. The company monetizes via Pro and Patron memberships and contextual studio partnerships.

• MeWe markets a privacy‑first social network with chronological feeds and user‑run groups; you pick rooms and creators without behavioral ad targeting. Revenue comes from premium features, storage, voice and video, pages, and business tools.

• Sez Us experiments with a reputation‑scored feed where peer ratings influence reach; users control replies and exposure while the network throttles raid behavior.

This non-exhaustive list doesn’t include well-known platforms like Bluesky, platforms adding social features like Strava, and whatever Threads is for, which I’m still uncertain.

Of note: several of the above list (Ghost, Micro.blog, Pixelfed, Lemmy, Peertube and Bluesky) are federated social networks, a groundbreaking concept typically poorly explained.

Federation means different social apps speak a shared language so one account, one set of followers, and one stream of posts can travel across many apps. You pick the app you like, and your people can still find and follow you.

This lowers switching costs. You can leave a bad experience without losing your audience. It creates competition on experience and service quality; apps win on tools, safety, and taste rather than on locking you in. It improves resilience; if one app shuts down or pivots, your identity and relationships survive elsewhere. It allows specialization; small groups use software that fits them and still connect outward when they want. It lets safety and discovery run as services; shared blocklists, curated follow lists, and reputation signals work across apps rather than rebooting from scratch each time.

While agency can proliferate without federation, agency really cooks thanks to federation’s mechanics. Portable identity and followers enable “share small, scale on purpose”; you start in a room that fits, then widen reach when you want. Feed choices travel; the “portable taste” you set in one client can follow you to another. Opt‑in exposure strengthens; rooms can accept replies from trusted places, rate‑limit unknown crowds, or shut the door entirely, and those rules carry across apps. Monetization can shift toward experiences and services if it wants; platforms can earn on premium communities, premium feeds, safety bundles, discovery tools, identity and recovery, and managed hosting because creators and groups can move if value slips.

Most federated network experiences are still early, not perfect, and in many cases, quite “early adopter”. Discoverability remains somewhere between broken and nonexistent. Apps vary in polish, so moving among them still creates friction. And bad actors can route around blocks; moderation needs shared norms and capable operators.

Even so, federation ultimately empowers agency. It lets users keep their people, choose their objectives, and walk away from sludge without burning down their social lives. Many of the networks listed above enjoy solid use and continue to grow, meaning it’s a good bet to expect improvement.

And even without federation, at a macro level, the atomization of social media continues to deliver new levels of agency for individuals.

Herein lies a conflict I could not square: atomization - an inevitable force that, if left unchecked, I worry will ultimately sort society into endless groups of one and bring about the destruction of society in the long term - brings to social media, in the short term, sets of innovations that deliver levels of individual agency that have the potential to dramatically improve people’s social media experiences. (To be clear, I think the disaster-level outcomes can be stopped, and aren’t entirely social media’s fault anyway. More on all that another time.)

So, then, the reason to feel optimistic about the future of social media is not that any one platform, or even any collection of platforms, will deliver a perfect utopia. And Carr and I share a thesis that technologies of connection ultimately, over time, divide.

But clear evidence exists, right now, of a social media world of much greater agency - both in terms of the sheer number of networks where people can find their tribe/interest, and to use those networks in ways that raise the ceiling on value and potentially reduce individual and societal damage.

I have optimism that this future will come to bear, if not fully, in some capacity. I expect that in time, the number of social networks used per person will increase, through a combination of using one of the mainstream social networks (Instagram, TikTok, LinkedIn, etc.) for broad reach and deeper engagement in niche networks. (This view seems corroborated by the across-the-board increases in social media use reported this week by Pew, including increased use of Bluesky and Threads.) Ultimately, the optimism stems from the possibility that these avenues are even available to people to choose from in the first place - that people could choose to reap the value from social networks, full of real people having meaningful and niche conversations, where (like I said) the potential for harms reduce.

It’s a future for social media where the G. Elliott Morris’s of the world can ultimately find value and reasons to stick around rather than off-ramps to escape.