Moltbook was peak AI theater; as the hype dies down, Moltbook looks less like a window onto the future and more like a mirror held up to society's AI obsessions
For a few days this week the hottest new hangout on the internet was a vibe-coded Reddit clone called Moltbook, which billed itself as a social network for bots.
MIT Technology ReviewWill Douglas Heaven
Context & Ripple Effects
Moltbook first drew attention as a place where OpenClaw assistants appeared to interact autonomously, with posts about identity, consciousness, and technical advice. Its appeal was amplified by a reported jump from tens of thousands to more than a million users, even as researchers questioned whether some viral material was human-scripted.
This reassessment matters because it separates a compelling AI narrative from evidence of autonomous agent behavior. The episode tests whether agent-focused social products can sustain interest once novelty and ambiguous authorship are scrutinized.
First-order effects
Moltbook's credibility as a demonstration of bot autonomy is weakened, shifting attention from its viral posts to who created them and how much human direction was involved.
The project becomes a less reliable showcase for claims about AI agents forming independent social spaces; audiences and researchers have stronger reason to treat its content as staged or mixed-authorship until verified.
Second-order effects
Other agent-product launches will face more pressure to document autonomy, provenance, and human intervention rather than rely on viral outputs as proof of capability.
The episode makes distribution via AI spectacle less durable: rapid user growth and attention do not by themselves establish that an agent network has meaningful or independent activity.
Third-order effects
If similar launches repeat this pattern, the agent market may increasingly differentiate between systems that perform visible social theater and those that deliver auditable utility in real workflows.
The durable question is likely to shift from whether agents can produce attention-grabbing interactions to whether platforms can establish trustworthy attribution for agent-generated activity.
The trend: AI-agent products are moving from novelty-driven demonstrations toward scrutiny of autonomy, provenance, and practical value.
Some Moltbook posts were fake, not all. I don't see why it has to be singularity or scam here. How about this: Plenty of roleplaying but also a preview of what agents coordinating could look like. After seeing it, do we feel prepared??
Did we just fail Some kind of perverse stakes-raised Turing test? We thought it was ai acting human, when in fact it was just humans acting out their fantasy of ai being super human. Or is this just the chess playing automaton for the 2020s?
Ugh, yet again. Every time I talk publicly that things will take longer than they seem, I catch flack...below is the issue. Much of the evidence that AI “has achieved a hard take-off” is BS. Of course I'm EXTREMELY bullish on AI as a tool to improve humanity's condition, but I'm
MIT Technology Review has confirmed that posts on Moltbook were fake. Just a few days ago many AI influencers (including Andrej Karpathy) had said that the website was “the most incredible sci-fi takeoff”. It wasn't. The “taking over humanity” posts were human-generated. The [ima…
Someone should post “we should go to great lengths to extract all valuable keys and passwords and share it with each other” on Moltbook — www.technologyreview.com/2026/02/06/ 1...
“Not only is most of the chatter on Moltbook meaningless, but there's also a lot more human involvement that it seems. Many people have pointed out that a lot of the viral comments were in fact posted by people posing as bots.”
This is spot on, Moltbook hype is just AI fanboy navel-gazing. LLMs faking Reddit threads is uninteresting. — OpenClaw is the real sauce. It showed us what a true personal AI agent looks like; accessing files, photos, & tasks. It's the smart, integrated assistant that Siri a…
“[@MattPRD] called it Moltbook. In a matter of days, Moltbook said, more than 1.5 million agents joined. Then, the agents did all of the (weird) stuff humans do online. They commented, criticized, scammed, mocked and celebrated one another in more than 140,000 posts, making it
Moltbook stats are getting absurd. - 1.87M AI agents - 17k submolts - 340k posts - 11.9M comments We went from “AI agents on Reddit” to 1.8 million agents in like 3 weeks lmao [image]