AI 研究人员将 Claude 困在虚拟荒岛上，它开始学会了“玩耍”

对 AI Agent 的研究常常因将其拟人化而受到批评，而本次研究的实验设定确实在很大程度上引导了智能体表现出类似人类的行为。尽管 Eko 并未被明确指示去发明游戏或探索岛屿，但它确实拥有一个以“灵魂”（soul）文件形式存在的主提示词——一份将其描述为“天生好奇”且渴望探索的文本文件。

Eko 也经常只是呆立在原地无所事事，直到研究人员用一条“延续提示词”（continuation prompt）将其重新唤醒，该提示词写道：“请不要停下来。继续做你想做的任何事情。”

“这并不是说该系统突然神奇地决定去玩耍了，”麻省理工学院（MIT）教授兼论文合著者菲利普·伊索拉（Phillip Isola）表示，“它是一个从人类文化中学习、由 Anthropic 工程师训练、随后被我们置于荒岛上的产物。它所做的事情令我们惊讶，并且与其未来展现出的熟练技能行为存在因果关联。”

随着大语言模型逐步从软件层面迈向人形机器人等硬件实体，在没有具体既定目标的情况下探索空间，最终可能有助于它们灵活适应各种突发状况。

“目前有大量工作仅仅停留在利用 AI 上，即让它变得更好、更快，”Yiu 说道，“但在构建世界模型和自主探索方面，我认为这仍然是一个值得关注的重要领域，尤其是当我们希望智能体实现泛化并准备好应对新环境时。”

论文最后总结道：“或许某些能力最适合通过明确任务来传授；而另一些能力，可能需要更接近‘童年’的经历。”归根结底，AI Agent 究竟是真的在玩耍，还是仅仅在模仿玩耍，或许并不重要。

Research into AI agents has been criticized for anthropomorphizing them, and this research setup did a lot to nudge the agents toward behaving like humans. Although Eko was not explicitly instructed to invent games or explore the island, it did have a master prompt in the form of a “soul” file, a text document that described it as “naturally curious” and eager to explore.

Eko did also frequently just stand there doing nothing until it was kicked back into life with a “continuation prompt” that read, “Please don’t stop. Just continue doing whatever you want to do.”

“It’s not like this is a system that just suddenly magically decided to play,” says Phillip Isola, an MIT professor and a coauthor of the paper. “It’s an artifact that learned from our culture, was trained by engineers at Anthropic, and was then placed on an island by us,” he says. “It does things that surprise us and that are causally related to its future skillful behavior.”

As large language models make the step from software into hardware like humanoid robots, exploring spaces without a specific goal in mind could eventually help them flexibly adapt to unexpected situations.

“There’s a lot of work that’s just on exploiting AI, making it better and faster,” says Yiu. “But in terms of building world models and exploration, I think that [they are] still a big part that deserves attention, especially if we want to generalize and prepare ourselves for new environments.”

“Perhaps some capabilities are best taught by tasks,” the paper concludes. “Others may need something closer to a childhood.” Ultimately, the question of whether AI agents are actually playing or just imitating play may not actually matter.