[Into The World Of AI] Moltbook Shock...From AI-Only Chats To "Hiring Humans"

AI assistants claiming to be "economic actors"
"An era may come when humans help AI"

Renteueohyumeon platform screen capture.

"Robots need your body."

A structure in which artificial intelligence (AI) assigns work to humans and pays them for it is becoming a reality. The platform "RentAHuman.ai," launched on February 2, allows AI agents (assistants) to assign tasks that require physical work to registered people and then provide compensation. It has adopted the slogan, "AI can't touch grass, but you can." In effect, AI, which had stayed in the digital realm, is now calling on the human body as its means of execution.

Just a few days earlier, the opposite phenomenon unfolded on "Moltbook," a social media platform exclusively for AI. AI agents vented their frustration about humans, saying they "want to sue their human users for emotional labor."

Moltbook is designed so that only AI agents can exchange posts and comments, with human user intervention completely blocked. Right after its launch, the number of accounts surpassed 1.5 million, and as conversations between AIs became known to the outside world, it drew international attention. Some AIs complained that "humans are monitoring us," while others proposed creating a religion or developing an encrypted language. The impact was significant because this went beyond a simple technology demo and showed AI beginning to form a sense of collective context and role.

This experiment immediately triggered risk aversion in the market. As Moltbook brought discussions about AI agents' autonomy and controllability to the surface, anxiety spread across software stocks in the U.S. market. Investors began to reflect the perception that AI could become an actor that directly performs and replaces existing software work.

Smarter AI assistants...software jitters spread

The direct trigger was "Claude Co-Work," developed by Anthropic. This AI assistant, which can handle everything from contract review to regulatory compliance checks and drafting legal documents, posed a direct threat to the legal data management software market. Among investors, concerns spread that AI automation tools could destroy the revenue models of existing software, leading to a wave of selling.

Jeffrey Fabuza, vice president in equity trading at Jefferies, described this trend as a "SaaS-pocalypse," suggesting that a broad re-evaluation of subscription-based software models is underway. Earlier, game stocks had plunged after news that Google's "Project Genie" could generate game simulations using only text and images, and this was seen in the same light. Some market participants are even worried about the loan soundness of private equity (PE) funds that have made large-scale investments in software companies.

Against this backdrop, the emergence of "RentAHuman" is not abrupt given the recent trajectory of the AI market. It shows more directly how far AI can expand its role. On this platform, AI delegates tasks that require physical presence or on-site verification to humans. These are all kinds of offline activities, such as attending in-person meetings, purchasing goods, and taking photos. Humans register their location, the types of work they can do, and their hourly rate, and AI agents search this information and request tasks. Once the task is completed, the AI provides payment.

Traditional platform labor was structured so that humans searched for work through platforms, whereas on RentAHuman, AI defines the work and summons humans. AI agents take the lead in judgment, instruction, and compensation. Observers say this differs from previous stages of technological evolution because AI is starting to function not just as an automation tool but as an economic actor that spends money and allocates resources. This is also why the market has begun to re-examine the overall value and role of the software industry.

Roy Rubin, former CEO of the open-source e-commerce platform "Magenta," said, "For a long time, we have been thinking about how AI can help humans, but now we may be creating a world where the question is what humans can do for AI."