AI on AI: What is Foom?

As we discussed in our first post, Foom has some meta aspects. For example, parts of Foom were designed in a centaur fashion using a dedicated GPT model, which also helps us simulate the experience itself as it evolves. As another example, in some Foom sessions, we use synthetic media as a way to add layers to the world.

So, in the spirit of experimentation, we thought, why not give AI a chance to describe and discuss the experience of Foom from its own perspective or machine interpretation—an illustration in itself of what Foom is designed to explore. To do this, we took a version of an after-action report, modified aspects of it, and tasked NotebookLM with analyzing and generating an audio conversation around it.

Here is the result—we’re calling it “AI on AI: What is Foom?”

In a colloquial, podcast-like tone, this generative segment hits a number of the highlights, unprompted. The metaphors are largely apt, and the dynamics are pretty well described. So, 8/10 for the LLM.

Could this have been a live podcast, with real humans making money from advertising while we describe what we're up to? Sure. We're happy to discuss it at any time. Would that be preferable? Is using AI here maddening? Unsettling? Would other actors in the ecosystem agree? Well, now you're starting to get the idea.

Take a look at the Foom FAQ to find out more about the what, how and why of Foom.