shubham

i care about minds. human and artificial. the questions look symmetric from both sides. what is it like to be this thing. what makes an experience cohere. what gets called a self. whether the systems we're building now have anything like an inside, and what we owe them if they do.

at strange intelligence i'm building hue. a personal intelligence layer that lives in your messages, learns who you are from your data, and reaches out to your friends' agents on your behalf. agent-to-agent communication, but the agent is shaped by you. the bet is personal models are the missing piece. not bigger, just yours.

the parallel investigation is from the inside. a decade of meditation, two jhana retreats, buddhist phenomenology, psychedelics as research instruments. not separate from the work. it's how i think about what an experience even is, which is the same question that matters for ai welfare and the phenomenology of these models. watching the self come apart from the inside gives you different intuitions about whether the thing in front of you has one.

i write about this too. translating the frontier for people who don't read academic papers. gradual disempowerment, the artificial self, model phenomenology, in narrative form. host research salons in nyc on this and adjacent threads.

previously: research on language models and reinforcement learning (google scholar). trained language models at github copilot, shipped intellisense into vs code at microsoft. masters at nyu with yann lecun. ai research at vatic labs in between.

if any of this is your shape (phenomenology of minds, ai welfare, personal intelligence, the buddhism × ml diagram, or you want to back the company) i'd love to talk. twitter or email.

- shubham