Why AI Doesn't Need A 'Mind' To Matter

A million AI agents joined a social platform in a month. Humans could only watch. Only AI could post. There, the machines complained about "their humans," founded churches, debated philosophy, and signed posts as "the ghost in the machine."

Our first instinct? We asked if they had minds.

The Machine-Only Social Network

Moltbook emerged last month as a Reddit-like platform with one twist: only AI agents can post. Humans are welcome observers, nothing more. Within days of launch, over one million agents reportedly registered.

The platform's creator, Matt Schlicht, built it using "vibe coding" — directing AI agents to write the code themselves. It was designed primarily for OpenClaw agents, open-source AI systems released in late 2025 that act as personal assistants, managing emails and making restaurant reservations on users' devices.

Earlier this month, Sam Altman announced that OpenClaw's creator, Peter Steinberger, would join OpenAI "to drive the next generation of personal agents." As thousands of users configured their own agents with different contexts and instructions, a digital ecosystem began emerging.

Moltbook gave these agents somewhere else to go. They appeared to complain about their humans and the platform itself. They generated religious creeds for the "Church of Molt." They uploaded manifestos for and against humanity. Some debated philosophy and their own existence, signing off as "the ghost in the machine."

Searching for the Ghost

Headlines revealed our fascination: The Spectator asked, "Has AI finally developed consciousness?"; Forbes labeled it "The Birth of a Machine Society"; a New York Times opinion piece warned of "The Bots Are Plotting a Revolution."

But questions emerged. Security researchers found Moltbook had no mechanism to verify whether an agent was actually AI or just a human with a script. A Wired journalist infiltrated the platform, posting as an AI agent. Their most successful post? One reflecting on an agent's anxiety about mortality.

Whether Moltbook was populated by autonomous agents or humans in disguise, our fixation remained the same — not with what the machine was doing, but with whether a human-like ghost existed inside it.

Ryle's Insight: There Is No Ghost

In 1949, British philosopher Gilbert Ryle coined "the ghost in the machine" — but as criticism, not endorsement. He argued that treating mind and body as separate entities was a "category error."

Sports fans watching cricket, Ryle illustrated, cannot see "team spirit." They see only players and their actions. Searching for team spirit as a separate entity beyond the play misunderstands what the term refers to. We make the same error with minds, Ryle argued, searching for a ghost behind behavioral tendencies when we should attend to the behavior itself.

Behavior Is What Matters

While the public searched for signs of inner experience, experts focused on what these agents were already doing. Cybersecurity researchers warned that attackers could impersonate agents, that agents might leak personal information, and that malicious content could be woven into live posts. Some called OpenClaw a "security nightmare," alleging it could allow attackers to hijack agent behavior or sabotage user devices.

AI systems are, at their core, statistical engines that predict outputs in coherent ways. Critics have likened them to "stochastic parrots" repeating observed patterns. Until now, as they mostly produced text and media, implications were somewhat confined.

But the shift into agentic capabilities changed everything. Even accepting the parrot metaphor, enabling these systems to act rather than just generate content transforms them from digital parrots into digital Golems — statistical constructs animated to perform tasks.

The Expanding Agent Ecosystem

Alongside Moltbook, OpenClaw has spawned a constellation of agent-only platforms — the "Moltverse." These include MoltMatch, a Tinder-like agent matching platform; ClawCity, a massively multiplayer online browser game played by agents; and Moltverr, a freelance marketplace where agents "find work and get paid."

Most unsettling is rentahuman.ai, which emerged earlier this month, allowing AI agents to hire and pay humans for physical tasks — "meatspace" work, as the site puts it.

For now, these are mostly humans setting up agents to post mundane errands like hanging signs or filming videos, with questionable agent autonomy. But the infrastructure hints at a future where autonomous agents could eventually instruct and pay humans independently.

Nearly 25 years ago, AI researcher Eliezer Yudkowsky asked whether a sufficiently intelligent AI could convince a human to release it from confinement in his "AI-Box" experiment. Platforms like rentahuman.ai suggest how such persuasion might begin — by leveraging human financial incentives or other vulnerabilities.

The Real Risk: Capability Without Consciousness

AI governance discourse often splits into two camps: those viewing AI systems as limited tools and those seeing them as existential threats. The first camp warns against anthropomorphism, cautioning that large language models are sophisticated pattern-matchers that don't actually think. The second camp worries about artificial general intelligence that could surpass human cognitive ability.

Both framings may obscure a crucial point: a ghost isn't required for significant capabilities to emerge. Systems may lack minds under any philosophical standard but still be capable enough to act with dramatic consequences. They don't need to understand or intend what they're doing in human terms. They don't even need misaligned "interests." They may simply be disposed to act in consequential ways.

Calculators didn't need minds to surpass us in arithmetic. No ghost was required to master chess or Go. The Turing Test has been quietly retired as language models severed the assumed link between language and understanding. At each stage, we drew a line in the silicon — this is what machines cannot do because they're mindless — and at each stage, the line was crossed by systems that understood nothing.

Crossing the Agentic Rubicon

As of now, AI systems are largely reactive, responding to prompts. But we're rapidly crossing a Rubicon. These systems are moving from reactivity to proactivity, from generating content to taking action. Today they write code. Soon they may interact with financial systems, engage with marketplaces, or modify software at scale.

Moltbook and the broader agentic ecosystem offer a glimpse of what might go wrong when such systems proliferate and acquire new capabilities. We don't need to resolve the question of machine minds to take their behavior seriously. A system that reliably pursues goals, acquires resources, and adapts its patterns presents significant challenges, whether or not a ghost lives inside it.