Dan Björkegren

In my class on AI institutions, the first assignment was to fight over what the assignments should be. That resulted in chaos--which was the lesson. AI will disrupt many parts of society. My class asks: what do we build next? 🧵 danbjork.substack.com/p/the-instit...

I did make all my students build demos—and they were great. I was impressed by how even nontechnical students made use of AI tools to build creative projects. The accompanying blog post shares a few! We were also grateful for a visit from @sebk.bsky.social

In my class on AI institutions, the first assignment was to fight over what the assignments should be. That resulted in chaos--which was the lesson. AI will disrupt many parts of society. My class asks: what do we build next? 🧵 danbjork.substack.com/p/the-instit...

AI makes traditional policy school assignments trivial, so I make students agree what to create. This year, technical students suggested software demos. Opposed: all who had never coded. That debate helped us understand how people feel when a new technology upends their lives.

Do markets believe that AI will transform the economy? Interest rates offer a clue. New working paper: I find that long-term bond yields appear to shift in opposite directions after releases of open vs. closed AI models. (1/3)

Isaiah Andrews and Maryam Farboodi documented the yield decline around mostly proprietary AI model releases in a thought provoking paper. Following closely, I find evidence that markets react differently to open weight model releases–possibly in the opposite direction. (2/3)

New working paper: Could AI leapfrog the web? Only 37% of sub-Saharan Africans use the internet. Cost is the #1 constraint. In a study with 469 teachers in Sierra Leone, we find AI works better than the web--and is 87% cheaper. arxiv.org/abs/2502.12397 Let's dive in... 🧵(1/X)

This is together with the excellent folks at Fab Inc who created the chatbot and are coordinating work at AI-for-Education.org. Also @educaid.bsky.social in Sierra Leone

New working paper: Could AI leapfrog the web? Only 37% of sub-Saharan Africans use the internet. Cost is the #1 constraint. In a study with 469 teachers in Sierra Leone, we find AI works better than the web--and is 87% cheaper. arxiv.org/abs/2502.12397 Let's dive in... 🧵(1/X)

AI can reformat knowledge to work better on small screens and costly, intermittent connections. Can it help catch up remote, low-income communities to the revolution of information?

New working paper: Could AI leapfrog the web? Only 37% of sub-Saharan Africans use the internet. Cost is the #1 constraint. In a study with 469 teachers in Sierra Leone, we find AI works better than the web--and is 87% cheaper. arxiv.org/abs/2502.12397 Let's dive in... 🧵(1/X)

Because querying an AI is 3,107x more data efficient, AI is already 87% cheaper than loading a web page for our teachers in Sierra Leone.

New working paper: Could AI leapfrog the web? Only 37% of sub-Saharan Africans use the internet. Cost is the #1 constraint. In a study with 469 teachers in Sierra Leone, we find AI works better than the web--and is 87% cheaper. arxiv.org/abs/2502.12397 Let's dive in... 🧵(1/X)

What do teachers use AI for? Mostly for facts and conceptual questions.

New working paper: Could AI leapfrog the web? Only 37% of sub-Saharan Africans use the internet. Cost is the #1 constraint. In a study with 469 teachers in Sierra Leone, we find AI works better than the web--and is 87% cheaper. arxiv.org/abs/2502.12397 Let's dive in... 🧵(1/X)

Love AI and economics? The NBER Digital and AI spring meeting will be at Stanford the day before Valentine's. Two weeks left to send us your best work! ❤️ www.nber.org/conferences/...

New technologies allow--and may require--new ways of organizing society. How might education, economies, and governance be rethought for an AI era? That is the topic of my new course AI Institutions at Columbia SIPA.

We'll be wrestling with the work of the future, who should train machines, and what values AI should encode. We'll explore whether decentralized minibus networks in Africa can tell us about planning cities for self driving cars, algorithms that bridge rather than divide, and ways to govern.

New technologies allow--and may require--new ways of organizing society. How might education, economies, and governance be rethought for an AI era? That is the topic of my new course AI Institutions at Columbia SIPA.

Much discussion about AI swings into hype or science fiction. It's easy to become unmoored. But most careful empirical studies tell us about marginal changes or the past. How can we understand the futures that are possible?