Science, done in daylight
iChancellor’s AI is an independent research academy founded on a simple premise: the most consequential technologies of this century — artificial intelligence, embodied machines, and quantum computers — should be developed where everyone can see, verify, and benefit.
To advance intelligence — digital, physical, and quantum — as open science in service of humanity.
Mission, vision & objectives
To advance intelligence — digital, physical, and quantum — as open science in service of humanity: every model, dataset, and result published where anyone can verify it, reproduce it, and build on it.
A world where the most consequential technologies of this century are developed in daylight — where Action Models, intelligent machines, and quantum computers belong to the scientific commons, and a small academy in the open can set the standard the closed giants have to answer to.
- 01Ship the fully open Chancellor family of Action Models — Sep 9, Oct 10, and Nov 11, 2026 — with weights, data audits, code, and evals on day one.
- 02Launch the next generation — LAAM, WPIM, and GPM — on 11.11.26, extending action learning to whole activities and many bodies.
- 03Prove physical intelligence in the real world through RoboMan field trials in factories, farms, response teams, and labs.
- 04Reach open, reproducible quantum error-correction milestones on the QuArc program.
- 05Grow from 5+ to 100 partner universities by 2027, with free open access to our platforms.
- 06Study the human impact of every technology before it ships — through our social science program — and publish 100% of our research openly.
Four commitments
Radically open
Weights, data, schematics, failures. If we learned it, you can download it. Openness is our method, not our marketing.
Rigor before reach
We publish when results replicate, not when a launch window opens. Every claim ships with the evidence to check it.
Frontier for good
We choose problems by human impact — climate, health, scientific discovery — and measure ourselves by outcomes, not benchmarks alone.
One academy
AI researchers debug robots; quantum physicists train models. Our edge is the distance between our labs — measured in meters, not org charts.
The people steering it
Our leadership team spans machine learning, robotics, quantum physics, and research operations — and every one of them still publishes.
Bankatesh Choudhary
Chief Research Scientist
Dr. Rahul Choudhary
VP, Medical AI
Amit Roy
VP, Physical AI
Sumit Singh
VP, Quantum Computing
How we got here
- 2021
Founded
iChancellor's AI launches as an independent, non-profit research academy with an endowment dedicated to open frontier science.
- 2022
First open model family
Our first fully documented Action Learning Model datasets take shape, built for training-data audits — preparing a first for models at that scale.
- 2023
Physical AI lab opens
The embodied-intelligence program begins, pairing learned world models with real hardware from day one.
- 2024
QuArc program launches
We begin commissioning and concept work on a superconducting quantum testbed, and commit to open calibration data for every experiment.
- 2025
Atlas-H stands up
Our humanoid research platform takes its first untethered steps; the control stack is released as open hardware and software.
- 2026
Three frontiers converge
Chancellor-1 tops open reasoning benchmarks, QuArc-128 crosses the error-correction threshold, and Atlas-H enters season-long field trials.