One person's ideas shouldn't be limited by one person's ability to execute.
We're a startup team from Zhongguancun Academy, Beijing’s national AI academy.
In the past, Apps provided the service, and people operated software to get things done. Now people can ask an Agent to do the work for them. This is what we call Agent as a Service, and it brings a new division of labor: people bring the insight, Agents do the work.
An Agent should also find and coordinate the tools it needs to do the work. Devices let Agents gather information and take action as the task requires; when something needs other people or other Agents, the Agent finds them and organizes the collaboration. People no longer have to coordinate every step themselves.
We define ChamOS as an app built for AGI. If AGI arrives, what will it look like? Not one model that solves everything. However smart a model gets, it can't give itself a name, a home, hands, or someone it answers to; ChamOS doesn't build models, it builds these. Many AI apps patch the weaknesses of today's models, and stop being needed once models improve. ChamOS is the opposite: the stronger models get, the more these matter.
We want Agents to take on all the non-creative work you hand them.
We're building a future where people bring the insight and Agents do the work, so that no one is held back by limited information or limited capacity to act.
If you have ideas or suggestions, come talk to me directly.