Writer and builder, working on AI safety.
I'm Eli Freedman. For a decade I've made hard ideas legible — essays for a broad public, and scholarship on medieval Arabic and Jewish philosophy at the Hebrew University. Alongside the writing, I build: AI-assisted apps, search engines, and translation pipelines, shipped end-to-end.
Now I'm pointing all of it at AI safety — communicating and researching the questions of epistemic agency, human autonomy, and alignment, and carrying them to the humanistic and religious publics that safety media overlooks. My work on tenth-century Islamic legal theory turns out to speak directly to alignment: licit stratagems (ḥiyal) as a millennium-old account of reward-hacking, and a medieval debate over whether the good is discoverable by reason as a precursor to value alignment.
About
I work at the seam between the tenth century and the present. My research, in the Arabic department of the Hebrew University of Jerusalem, is on Saadia Gaon, Qirqisānī, and the shared Arabic theological culture of Jews, Christians, and Muslims in the medieval Near East.
Alongside the scholarship I translate professionally — critical and parallel-text editions out of Hebrew, Judeo-Arabic and Arabic; localization between Portuguese and English — and write journalism that carries serious ideas to general readers.
And I build. Using modern AI tooling with a human-expert pass, I make web apps, search engines, and translation pipelines that put classical Jewish texts within reach.
I'm now bringing all three — writing, research, and building — to AI safety. I'm completing BlueDot's AGI Strategy course, and in conversation with the Hebrew University's AI Governance Lab and with Sinai and Synapses on religion and positive AI futures. Open to AI-safety communication and research collaborations.
Hebrew · Judeo-Arabic · Arabic · Português · English
Projects
Editions, translations, and software — shipped.
More references available on request, including academic-publisher projects.
Writing
Scholarship for the footnote-readers; journalism for everyone else. A few places to find it:
Working on AI safety, governance, or reaching new audiences?
Tell me what you're building — safety communication, governance research, or getting these ideas to new publics. I'd like to help.
freedmaneli@gmail.com