Fetching the paper…

Reading the bibliography…

2023

Aligned: A Platform-based Process for Alignment

Shaotran, Ethan, Pesok, Ido, Jones, Sam et al.

Original
PDF
Post on X

Understand

We are introducing Aligned, a platform for global governance and alignment of frontier models, and eventually superintelligence.

While previous efforts at the major AI labs have attempted to gather inputs for alignment, these are often conducted behind closed doors.
We aim to set the foundation for a more trustworthy, public-facing approach to safety: a constitutional committee framework.
Initial tests with 680 participants result in a 30-guideline constitution with 93% overall support.

Aligned: A Platform-based Process for Alignment · Around

Built on

Toward trustworthy ai development: Mechanisms for supporting verifiable claims
Original
Miles Brundage, Shahar Avin, Jasmine Wang, Haydn Belfield, Gretchen Krueger, et al · 2004
Earlier work this paper cites.
Birdwatch: Crowd wisdom and bridging algorithms can inform understanding and reduce the spread of misinformation
Original
Stefan Wojcik, Sophie Hilgard, Nick Judd, Delia Mocanu, Stephen Ragain, M. B. Fallin Hunzaker, Keith Coleman, and Jay Baxter · 2022
Earlier work this paper cites.
Fine-tuning language models to find agreement among humans with diverse preferences
Original
Michiel A. Bakker, Martin J. Chadwick, Hannah R. Sheahan, Michael Henry Tessler, Lucy Campbell-Gillingham, Jan Balaguer, Nat McAleese, Amelia Glaese, John Aslanides, Matthew M. Botvinick, and Christopher Summerfield · 2022
Earlier work this paper cites.

Similar

Democratic inputs to ai, 2023
Wojciech Zaremba, Arka Dhar, Lama Ahmad, Tyna Eloundou, Shibani Santurkar, Sandhini Agarwal, and Jade Leung · 2023
Cited alongside, same era.
Governance of superintelligence, 2023
Sam Altman, Greg Brockman, and Ilya Sutskever · 2023
Cited alongside, same era.
Collective constitutional ai: Aligning a language model with public input, 2023
Anthropic · 2023
Cited alongside, same era.

Then

How are community notes ranked?
X · 2023
Closest in time.
Using gpt-4 for content moderation
Lilian Weng, Vik Goel, and Andrea Vallone · 2023
Closest in time.

Beyond the bibliography

alphaXiv searches the wider corpus for related work and actual follow-ups.

Open on alphaXiv

alphaXiv is searching for related work…