Research

Two questions worth answering well.

Rule-Before-Train AI Safety, and a Privacy-First Mobile OS. Each is backed by a working prototype or published feasibility study. Each surfaces what we're finding and what we want to achieve.

active

Rule-Before-Train AI Safety

Can rule-based scaffolding give us safer AI than post-hoc alignment?

Modern AI safety is largely retrofitted — we train models on enormous corpora and then apply RLHF, content filters, and prompt engineering to constrain behaviour after the fact. Rule-Before-Train inverts that: define explicit behavioural rules first, ship them as a working rule-based agent, and let machine-learning components extend that scaffold rather than replace it.

3 findings 3 goals
active

Privacy-First Mobile OS

Can a small team reach a daily-driver alternative to Android in 30–36 months?

OSphreX is a research initiative asking what a mobile OS would look like if it were designed around the premise that *the device you bought is the device you own* — and built by a small, principled team rather than a billion-dollar consortium. The feasibility study is published in full; the implementation is the open question.

4 findings 4 goals

Two questions worth answering well.

Rule-Before-Train AI Safety

Privacy-First Mobile OS

Cookies & analytics