
https://xitter.cf/t.co/Gh2TwqEjCz Econ prof @OxfordEconDept, econometrics, inequality, machine learning, politics. @maxkasy@mastodon.social
Oxford(UK), Vienna (AT)
Joined January 2014
- Tweets2.7K
- Following833
- Followers9.9K
- Likes1.9K
Pinned Tweet
💥 📚 Next step on the book journey:📚💥
"The Means of Prediction - How AI Really Works (and Who Benefits)"
is now in the UChicago Press catalog, and available for pre-order online!!!
press.uchicago.edu/ucp/books…
Maximilian Kasy retweeted
Interview zu KI in der Süddeutschen Zeitung, auch in der heutigen Printausgabe: sueddeutsche.de/kultur/maxim…
Interview zu KI in der Süddeutschen Zeitung, auch in der heutigen Printausgabe: sueddeutsche.de/kultur/maxim…
Maximilian Kasy retweeted
Reinforcement Learning: Theory and Algorithms
rltheorybook.github.io/rlthe…
(Theory behind RL techniques used in RLHF)
Theoretical Perspectives on Knowledge Distillation: A Review
wires.onlinelibrary.wiley.co…
(Introduction to model distillation)
5/5
Maximilian Kasy retweeted
Reinforcement Learning from Human Feedback
rlhfbook.com/
(Engineering aspects of fine-tuning models to objectives from human feedback or verifiable rewards)
4/5
Maximilian Kasy retweeted
The Emerging Market for Intelligence: How Firms Buy and Sell AI
aeaweb.org/articles?id=10.12…
(Documents competitiveness of open weight models)
3/5
Maximilian Kasy retweeted
This creates opportunities for local / democratic control:
- Take open-weight foundation models
- Fine tune for local / democratically chosen objectives using RL.
Here some useful reading recs:
2/5
Maximilian Kasy retweeted
Thread: Key developments in AI / LLMs
1. reinforcement learning (RL, from human feedback or verifiable rewards),
2. open weight models; some based on distillation of commercial models, not far behind commercial alternatives.
1/5
Thread: Key developments in AI / LLMs
1. reinforcement learning (RL, from human feedback or verifiable rewards),
2. open weight models; some based on distillation of commercial models, not far behind commercial alternatives.
1/5
Reinforcement Learning from Human Feedback
rlhfbook.com/
(Engineering aspects of fine-tuning models to objectives from human feedback or verifiable rewards)
4/5
Reinforcement Learning: Theory and Algorithms
rltheorybook.github.io/rlthe…
(Theory behind RL techniques used in RLHF)
Theoretical Perspectives on Knowledge Distillation: A Review
wires.onlinelibrary.wiley.co…
(Introduction to model distillation)
5/5
Jetzt offiziell verfügbar, in deutscher Übersetzung:
"Die Mittel der Prädiktion
Wie KI wirklich funktioniert (und wer davon profitiert)"
suhrkamp.de/buch/maximilian-…
In guter Gesellschaft im aktuellen @suhrkamp.de Katalog :-)
Maximilian Kasy retweeted
Heute in der FAZ (@faznet):
„Die entscheidende Frage der KI lautet: Wer entscheidet, was optimiert wird?“
faz.net/aktuell/wirtschaft/u…
Heute in der FAZ (@faznet):
„Die entscheidende Frage der KI lautet: Wer entscheidet, was optimiert wird?“
faz.net/aktuell/wirtschaft/u…
Maximilian Kasy retweeted
Aujourd'hui dans L'Echo (@lecho, Belgique):
lecho.be/opinions/general/ma…
Aujourd'hui dans L'Echo (@lecho, Belgique):
lecho.be/opinions/general/ma…
Great interview with @triofrancos
on the geopolitics of the renewable energy transition:
newleftreview.org/issues/ii1…
Maximilian Kasy retweeted
New paper with @p_ganong & Avik Garg:
From literature review to effect prediction for a new context - allowing for effect heterogeneity & publication bias.
maxkasy.github.io/home/files…






