If you're not already aware of it: This idea of a hyperphone seems highly convergent with Loom, a similar branching interface originally designed for interaction with language models. This sort of interface is very natural for language models, and in fact this "octopus mind" you describe makes a lot of sense as being a part of the mind closer to pure prediction (and therefore has similarities to language models). I agree that this structure makes a lot of sense for humans as well, and from what I can tell long-term use of Loom can pull people to approach conversations in a more multiversal fashion, which looks similar to a hyperphone in practice.
If you're not already aware of it: This idea of a hyperphone seems highly convergent with Loom, a similar branching interface originally designed for interaction with language models. This sort of interface is very natural for language models, and in fact this "octopus mind" you describe makes a lot of sense as being a part of the mind closer to pure prediction (and therefore has similarities to language models). I agree that this structure makes a lot of sense for humans as well, and from what I can tell long-term use of Loom can pull people to approach conversations in a more multiversal fashion, which looks similar to a hyperphone in practice.