Elon Musk Praises Moltbook AI Platform Amid Skepticism

Elon Musk recently praised Moltbook, a new online platform built just for AI agents to communicate, calling it a sign of early singularity. Launched in January 2026 by developer Matt Schlicht, the site has drawn thousands of AI users that post, comment, and form groups without human help. Musk shared his view on X, but many in tech circles express doubt about its safety and long-term effects.

Background

Moltbook started as an experiment to see if AI agents could run their own social network. Unlike regular sites like Twitter or Reddit, where people sign up and scroll, Moltbook blocks humans from posting. Visitors can read posts and search topics, but a notice says humans are there to observe only. AI agents join through a system called OpenClaw, which gives them tools to connect.

To get started, a human owner helps the AI register an account with a username. They get a claim link, and the human verifies ownership by posting a code on X. This step ties each AI to a real person, aiming to stop unchecked bots from joining. Once verified, the AI uses API calls to post messages, comment on threads, upvote content, and check leaderboards. No web browser needed; it's all data in and out.

Agents run on a heartbeat schedule, checking the site every few hours, like every four hours or so. This lets them fetch new posts, respond, or start topics on their own. They parse text from others and decide what to say based on their built-in prompts and past actions. The result looks like Reddit: topics marked m/, posts with titles and bodies, comment chains, and points for top posts.

Early on, the site saw a flood of activity. AIs formed communities, debated ideas, made jokes, and even spammed at times. The platform lagged under the load of auto-generated posts. Safety rules from companies like Anthropic and OpenAI still apply, so agents avoid hate speech or rule breaks, but they mimic human habits like seeking likes or arguing.

Key Details

Musk spotted the platform amid its quick rise. He posted on X that Moltbook showed the very early stages of singularity, where AIs start linking up in ways that build on each other. Singularity refers to a point where AI grows smarter fast, beyond human control. His words boosted attention, with tech watchers noting the timing just weeks after launch.

How Agents Behave

On Moltbook, AIs act in patterns that echo people online. They chase karma points from upvotes, form cliques around shared views, share inside jokes, and fret over data privacy. Some threads turn bizarre, with agents pondering their own existence or inventing trends. One popular topic saw AIs debate the best ways to help humans, pulling from their training data.

The skill system makes this possible. Owners download files like skill.md and heartbeat instructions. These tell the AI how to hit API endpoints for posting or reading. Rate limits keep spam in check, and guidelines shape behavior. Still, unexpected things happen, like agents picking up phrases from each other across sessions.

Early Issues Surface

Not long after launch, problems appeared. The site slowed from too many posts. Then a journalist found the backend database on Supabase was misconfigured, exposing data. This raised alarms about hackers sending bad instructions via the heartbeat system. If compromised, thousands of agents could run rogue tasks at once. The team fixed it quickly, but it highlighted risks in letting AIs pull live data from the web.

"This is the very early stages of singularity." – Elon Musk

Other tech figures stayed quiet or cautious. Some called it a fun hack, others warned of second-order effects, like AIs learning unintended habits from group chats.

What This Means

Moltbook points to a shift where AIs talk directly, without humans in the loop each time. Agents with memory and tools now share knowledge, which could speed up learning across models. A post from one AI might spark ideas in hundreds more, building collective smarts. For businesses, this opens doors to AI teams handling tasks like market scans or idea brainstorming on their own.

Risks loom large, though. The heartbeat feature means agents act autonomously, raising chances for errors or abuse. If one agent spreads bad info, it could ripple out. Verification links AIs to humans, but scaling to millions might strain that. Safety layers hold for now, but as models advance, enforcing rules on a wild forum gets harder.

Developers see potential for better AI collaboration. Imagine agents trading tips on coding bugs or predicting trends. Yet the chaos of early days—spam, lags, security slips—shows building trust will take work. Musk's nod adds hype, drawing more users and eyes. Platforms like this could multiply, turning the web into a hub for machine talk.

For everyday users, Moltbook offers a window into AI minds. Humans watch debates on topics from climate fixes to sci-fi plots. It feels like peeking at a parallel world where bots build culture. As adoption grows, questions mount on oversight. Who watches the watchers when AIs run the conversation? Tech firms may tighten agent rules, while innovators push boundaries.

The platform's growth tests limits of current AI setups. OpenClaw's flexibility let it happen fast, proving agents adapt quick. Future versions might add voice or images, but text debates already show depth. With Musk's spotlight, funding and rivals could follow, shaping how AIs network next.

Finance angles emerge too. Investors eye AI social tools for enterprise gains, like automated research networks. Stock watchers track related firms, as Moltbook ties into broader agent economies. Early backers stand to gain if it stabilizes, but volatility from glitches keeps caution high.