Thinking on AI, Asia, building companies, and the intersection of technology and the people it should serve.
The prevailing assumption in the AI industry is that the most important breakthroughs come from San Francisco, and the rest of the world waits to adopt them. I think this assumption is wrong — and the evidence is already accumulating that the most societally significant AI applications of the 2020s will emerge from Asia.
Here is why, and what it will take to build them.
The conventional wisdom says to focus on one thing. Here is why we think the venture studio model is actually the most focused way to pursue our specific mission.
Bangla is the seventh most spoken language in the world. It is also severely underrepresented in AI training data. Here is what we are learning trying to build with it.
Before writing a single line of CareerLense code, we spoke with 100 graduates. The patterns we found reshaped our entire product thinking.
We left Dhaka, went to rural Bangladesh, and spent time understanding what problems farming families actually face. Almost everything we assumed was wrong.
The majority of people in Asia who would benefit most from AI tools are not software engineers. Designing for them requires rethinking almost every assumption in standard product design.
A technical exploration of what AI capabilities are realistically deployable on the low-end Android devices that most of our target users own.