Subscribe to Bankless or sign in
Crypto Twitter has long been the town square for crypto.
I owe my current career to the app – I started curating cryptoart news tweets in January 2020 for my DeFi Arts Intelligencer column; one year later, Bankless brought this newsletter on as Metaversal, which I still write to this day.
Yet now, something is sour in the state of Crypto Twitter, and that sourness itself is symptomatic of deeper problems.
Losing Signal
X, formerly Twitter, was once a place where niche onchain alpha and deep connections came easily. Now, vibrant feeds are turning monotonous thanks to the algorithmic battlefield where slop and controversy overshadow all else.
There's more than a single driver here, to be sure. For example, the rise of X's creator ad revenue sharing program has incentivized engagement farming and divisive content, while the increasingly sophisticated AI bot takeover has flooded X with spammy, low-quality posts.
This vibes downturn led a few people in CT to turn to onchain social apps to build a new future for crypto discourse. But a lot of people just can't quit the network – a mass migration elsewhere has yet to occur, and the top onchain social projects have clearly struggled to reach product-market fit as evidenced by the recent leadership shifts at 
Farcaster and Lens.
But there's still plenty of latent potential – for alternative options, a second wind for onchain social, and even for the revitalization of X as crypto's town square.
What Does CT Want?
A few weeks ago, X product head Nikita Bier said in a since-deleted tweet that CT was "dying from suicide, not the algorithm."
That comment may have kicked up a firestorm, but in its wake Bier has made a series of overtures amenable to CT, including dethrottling "gm" tweets, announcing crypto-friendly Smart Cashtags, killing InfoFi spam, and open-sourcing the X algo.
These are all welcome steps that suggest X is thinking deeper about how best to approach one of its most active user bases, but there's certainly more work to be done.
From my vantage point, it seems most people in CT want a level playing field and more transparency from X.
For instance, X's moves to prioritize time spent on-platform by borking the reach of tweets with links may have resulted in better metrics for X, but have created a lower quality user experience – making CT less of a venue to go down informative rabbit holes on linked project updates and essays and more of a space to get sucked into ragebait.
That's just one minor example of something most people in CT would favor. But there are even grander possibilities. As 
Vitalik Buterin recently noted, it'd be great to be able to directly verify what the X algorithm is doing to the reach of topics and accounts on the platform.
He wrote:
"It's an important test of whether or not the recent 'open-sourcing of the X algorithm' is effective that you should be able to [hire someone to | ask a bot to] scan through the published code and verify whether or not unfair deboosting of disfavored views is happening.
In my ideal world, the code would be fully open and published with a delay, and you would have zk-snarks so you can actually prove that the timeline you see is the direct result of running the published algorithm."
We'll have to wait and see if X's transparency push continues in this direction. We'll also have to see if the app's Smart Cashtags are just an early taste of deeper walletizing, as 0xLuo puts it – a move that would surely prove popular with trader-centric CT.
Whatever happens, the more X can improve its algo, and the more that it can make CT feel like it isn't second-class, the more CT will feel at home again. And then maybe some of that old magic can return here.
Where Onchain Social Fits In
Onchain social has – so far – functioned like a refuge: a place people fled to from mainstream platforms rather than a place they were pulled toward. But this dynamic can be reversed, and the foundations needed for this are already here.
Farcaster “got more things right than any crypto social project before it,” longtime builder David Furlong recently argued. The protocol and team behind it solved hard problems that Web2 social disregarded, like portable identity, verifiable social graphs, permissionless clients, and so on.
It proved onchain social could feel good to use – but also that good infrastructure alone wouldn't create gravity.
For most of Farcaster’s life, CT has remained faster, denser, and more interesting. And even as Farcaster shipped great UX and novel primitives, it struggled with the cold start problem that plagues every new social network.
Worse, some of the very things that made Farcaster feel exciting early on didn't scale well. Financialization crept in as the network grew, incentivizing farming and spam. The high-signal builder commons became noisier, especially to anyone outside crypto. On this point, I agree with Vitalik Buterin, who recently asserted that onchain social projects should optimize for social utility:
"[D]ecentralized social should be run by people who deeply believe in the 'social' part, and are motivated first and foremost by solving the problems of social."
So what does optimizing for social look like in this day and age? How do we make a better feed? What do people actually want to do with their wallet-connected social graphs?
In my perspective, onchain social projects like Farcaster and 
Lens evolving into primary destinations may require deeper commitment to some of their founding features. Features like...
- Programmable composability, so developers (and AI-assisted creatives) can build radically different social experiences without permission.
- Decentralization, so users can customize, fork, and migrate around without losing their digital identity or history.
- Multiplicity, so rather than a monoculture, many focused communities can thrive instead of one global algorithmic arena.
Vitalik has spoken of the need for "mass communication tools that serve the user's long-term interest, not maximize short-term engagement." Composability, decentralization, and multiplicity are how we get there, and the full combinative power of these pillars is only possible in onchain social.
Naturally, this open terrain has huge opportunities for empowering, customizable, user-first UX. Furlong mentions in his linked essay that I find particularly convincing are the rise of "personalized interfaces," "next-gen personalized algorithms," and "ultra-focused experiences" where "AI won't let you post/show content that doesn't conform to strict criteria."
Your feed, your way, where you want it, and at your service rather than in service to some massive extractive company. Web2 could never.
If onchain social projects can mature the vision and build user ownership models that don’t turn into casinos, the onchain social scene could stand on its own, rather than living in CT’s shadow. The first chapter of onchain social may be over, but the second one is just beginning, and there's a lot to be hopeful for.
