Education will be industrialized.
Every child is born knowing nothing about the Earth or the universe, and every one of them has to be taught. We still do it by hand: a teacher at a blackboard, a professor in front of a camera.
For hundreds of years, each new technology has let humanity industrialize something it used to do by hand. Education had to wait for AI. Now it can all be manufactured: lectures, exercises, answers to every question, lessons shaped to what each student hasn’t understood. Academa is building that.
Teaching runs on media.
That is why classrooms have blackboards and online courses are videos. To teach calculus you need axes, curves, tangent lines, and equations that change as you work on them. AI knows calculus, but it has no blackboard.
So we are turning the blackboard into software.
We are inventing a language that can describe every figure in every STEM textbook, animate it, and make it interactive. AI writes in it, and our graphics engine draws the result live, in the student’s browser.
curve = plot(x²)
P = point(on=curve, x=1.40)
tangent(curve, at=P)
say("Here the slope is 2.80.")A handmade video is the same for everyone. A lecture made of software can be different for each student: in any language, with any question answered on the spot.
Courses are manufactured. Classes are live.
A teacher looks spontaneous but walks in with a plan. So we manufacture the plan in advance: that is the course. An AI tutor teaches it live. Students interrupt, ask anything, play with the board, and the lecture carries on from where it stopped.
“Why is the slope 2x?”
“Let me try a steeper curve.”
“In Turkish, please.”
See what lectures will look like, and interrupt one: preview.academa.ai
Calculus first, then every course on Earth.
Calculus comes out in November. Then physics and the rest of undergraduate STEM, in every language, on one platform: like Coursera, except that we make everything on it. We are building a plant, not a course, so each one comes faster than the last.
We are academics who build.
Technology alone will never teach anyone. It takes people who understand a subject deeply and love explaining it. With cameras, one of us could make one course in a month. With this, we will teach every subject on Earth, the way we would teach it ourselves.
Sina Atalay (CEO) put a PhD at Princeton on leave for this. Before that: simulation software at CERN; RenderCV, an open-source tool used worldwide; and dynamics lectures on YouTube, made for the love of it, for which the thank-you messages still come.
Abdullah Geduk (CTO) is doing a PhD at Penn State, simulating turbulence on supercomputers, and has kept teaching throughout: fluid mechanics at the university, and students one on one.
We have been friends since our first year at Boğaziçi University, and we have always taught each other what we learned, from electromagnetism to general relativity. We are not here to test an idea. This is what we will do with our lives.