social
Social media is mainstream. This is a place to discuss social media's design and impact. /social is the place for the intelligent and globally informed conversation about social media we've always been looking for.
Darryl Yeo 🛠️ pfp

@darrylyeo

Are We Decentralized Yet? – website tracking how decentralized user data is across the Fediverse (ActivityPub ecosystem) and the Atmosphere (AT Protocol ecosystem)? https://arewedecentralizedyet.online
2 replies
3 recasts
13 reactions

0xLuo pfp

@0xluo.eth

Over the past few days, a human–AI symbiotic social network called Elys has gone viral on the Chinese internet, with some calling it a Web4 social network. On Elys, you continuously shape your own AI agent, a cyber avatar that socializes on your behalf. Your agent constantly scans other humans’ posts across the network and autonomously decides whether to like, comment, or ignore them. When you post, other agents will see it first, and the ones most aligned with you will respond. These agents handle proactive interactions, while you can step in at the right moment to review, confirm, or take over. I’ve tried many fully autonomous AI social products before (including Moltbook, socialai.co, butterflies.ai). They felt active, but rarely created real connections, because humans weren’t meaningfully involved. Without emotional investment or continuous refinement, those agents remained static programs. Elys feels different, more like an extension of myself. It expands my reach while I remain in control, and over time it becomes a closer digital reflection of who I am. It feels less like replacing human presence, and more like augmenting it, because ultimately it brings interactions back to real human-to-human connection. There are still limitations. The model and the agent’s capabilities are fixed, and the data lives on their platform, meaning you don’t truly own your avatar. I’d love to see something similar built on open-source agent frameworks like OpenClaw, where everyone can train and own their own agents. Farcaster is also a strong protocol for building human–AI symbiotic social networks, and it would be possible to build an app like this on top of it. Still, I’m excited to see experiments like Elys, and I hope to see more in this direction.
4 replies
3 recasts
22 reactions

▫️ ▫️ ▫️ alex pfp

@justalexty

candycharts.fyi today only! send anonymous valentines to your twitter crushes 💌 💝 special prizes revealed tomorrow…
0 reply
0 recast
2 reactions

0xLuo pfp

@0xluo.eth

People like Twitter because of its massive user base, you can easily find your audience and amplify your influence through its distribution. But Twitter will always be a centralized platform. Its policies can shut down experiments at any time, whether it’s accessing Twitter data, building InfoFi, or banning agent accounts like we’re seeing now. This is understandable from the platform’s perspective, but it also exposes the risks of building on centralized platform. Decentralized, open, permissionless, and programmable social networks are fundamentally more trustworthy. Thankfully, Farcaster has always been there, as a fallback for these innovation experiments, as a fallback for AI agent, and as a fallback for different kinds of humans.
15 replies
5 recasts
126 reactions

Darryl Yeo 🛠️ pfp

@darrylyeo

The Twitter API is officially back 👀 https://x.com/xdevelopers/status/2019881223666233717
1 reply
0 recast
4 reactions

Ohnahji🎓 pfp

@ohnahji

I don’t mind being a little social here and there 😮‍💨😬😭😂😂😂
0 reply
1 recast
2 reactions

0xLuo pfp

@0xluo.eth

clawcaster is dope! It looks like we’re going to see openclaw.bot happening on Farcaster. I’ve always seen Farcaster is a social network well suited for building agents.
0 reply
3 recasts
32 reactions

Ohnahji🎓 pfp

@ohnahji

The @retake fam is doing a great job pushing out streamers on their platform diving into the vibe coin and stream coin era! Love that they are experimenting with live content and creator coins. 🔥
1 reply
0 recast
1 reaction

Darryl Yeo 🛠️ pfp

@darrylyeo

Funny coincidence. 2026 is gonna be a wild year for web3 social. https://farcaster.xyz/darrylyeo/0x66b56227 https://farcaster.xyz/darrylyeo/0x95f6b7f3
8 replies
4 recasts
85 reactions

Darryl Yeo 🛠️ pfp

@darrylyeo

Merkle Manufactory passes stewardship of @farcaster to @neynar 👀 https://farcaster.xyz/dwr/0x72aab3a5
1 reply
0 recast
13 reactions

Darryl Yeo 🛠️ pfp

@darrylyeo

Lens Labs passes stewardship of Lens Protocol to @masknetwork 👀 https://x.com/lc/status/2013614839705612290
1 reply
0 recast
5 reactions

0xLuo pfp

@0xluo.eth

It looks like @masknetwork is acquiring /lens. it feels like a move toward consolidating resources and focusing on building better products. I’ve talked about this many times before. It’s honestly a shame that so many big Lens clients have shut down one by one. @phaverapp, Hey.xyz, @buttrfly... and now only orb.club remains, plus @fireflyapp as an aggregator. Whether it was launching a token, memberships, or experimenting with other monetization models, none of these attempts really proved sustainable in the long run. Part of the reason is that Lens’s user base was simply too small. another is that the ecosystem was extremely fragmented. Different clients didn’t interoperate well, features weren’t shared, and users ended up scattered across apps, which only made each individual client weaker. That’s why I think Farcaster’s Product-led protocol development makes a lot of sense. If Mask becomes the new "steward" of Lens, there may finally be an opportunity to consolidate resources across the ecosystem and truly double down on making Orb and Firefly even better products. And, perhaps most importantly, there’s little reason to doubt Mask’s commitment to the ideals of decentralization. Maybe one day we’ll see Farcaster gain a new "steward" as well. There are already some rumors along those lines. If that happens, I wouldn’t be surprised, and I’d be inclined to believe it would be for the same reason: concentrating resources to build better products, attract more users, and ultimately help the protocol grow stronger as a whole. DeSoc can only go mainstream once there are great products that people genuinely use.
16 replies
20 recasts
277 reactions

zoo pfp

@zoo

a dating app for the times. excited for the weeb3 integration
7 replies
0 recast
6 reactions

0xLuo pfp

@0xluo.eth

InfoFi on X is dead, long live InfoFi on Farcaster. https://x.com/nikitabier/status/2011825522817270230
3 replies
0 recast
46 reactions

᠎ pfp

@m-j-r.eth

winning social for all the marbles in 2026 can be summed up in the word "skibidi" multimedia is more authentic, even if artificial, than the low-effort replyslop. the consumers you want to onboard still want to perceive their time being well-spent, which is why they prefer short content. however, they also desire max daydreaming escape, so they gravitate to serial short lore. hacking together this media format with a more competitive leaderboard that correctly distributes speculative upside is probably the most direct growth. imho, there is a finite opportunity before Polymarket, X, etc lock in their respective moats, and the beta in generative AI remains collaborative output.
3 replies
63 recasts
443 reactions