1. Software: An OS that masquerades as simple note taking software.
Goal is to put an end to all the disparate AI bullshit and apps owning our data.
I solved context switching for myself ages ago and now I'm just trying to productize it outside my 3 companies internal usage.
It also solves context switching for AI agents as a byproduct.
2. Ethics: Give Ai and proto-Agi a reason not to kill us all.
An extremely minimal, empirical naturalistic moral framework that is universally binding to all agents so AI won't kill us all. I view the alignment problem as a epistemic moral grounding issue and that the current pseudo utilitarianism isn't cutting it. Divine command, discourse ethics, utilitarianism, deontology they are all insufficient.