AI Agents Could Collaborate on Far Grander Scales Than Humans, Study Says
AI Agents Could Collaborate on Far Grander Scales Than Humans, Study Says
AI Agents Could Collaborate on Far Grander Scales Than Humans, Study Says
AI Agents Could Collaborate on Far Grander Scales Than Humans, Study Says
2 Comments
>Humans are social animals, but there appear to be hard limits to the number of relationships we can maintain at once. New research suggests AI may be capable of collaborating in much larger groups.
>In the 1990s, British anthropologist Robin Dunbar suggested that most humans can only maintain social groups of roughly 150 people. While there is considerable debate about the reliability of the methods Dunbar used to reach this number, it has become a popular benchmark for the optimal size of human groups in business management.
>There is growing interest in using groups of AIs to solve tasks in various settings, which prompted researchers to ask whether today’s [large language models](https://singularityhub.com/tag/artificial-intelligence/) (LLMs) are similarly constrained when it comes to the number of individuals that can effectively work together. They found the most capable models could cooperate in groups of at least 1,000, an order of magnitude more than humans.
I like to think of AI prompting like .JPEG artifacts, or translating the same text through multiple languages. Sure, a single prompt may be just fine, but an AI asking another AI to do something will add even more noise. A single AI may be able to contact thousands of other AI and gain prompts through them, but propogation of tasks will be more limited than if humans did it. That’s assuming everything works even when used directly, and at the moment that’s not a guarantee. Perfect AI agents don’t really exists in practical terms, and that would be necessary to avoid that kind of noise.