{"id":433,"date":"2026-08-08T11:05:01","date_gmt":"2026-08-08T11:05:01","guid":{"rendered":"https:\/\/nationalconsumerreportss.com\/?p=433"},"modified":"2026-08-08T11:05:01","modified_gmt":"2026-08-08T11:05:01","slug":"the-week-that-was-12","status":"publish","type":"post","link":"https:\/\/nationalconsumerreportss.com\/?p=433","title":{"rendered":"The Week That Was"},"content":{"rendered":"
\n

Ren\u00e9e DiResta and Ronald Robertson analyzed Grokipedia\u2019s update records, finding that the site appears to have stopped considering human edits or updating the site since late April. DiResta and Robertson highlighted how Grokipedia\u2019s lack of transparency harms both the site\u2019s reliability and ability to demonstrate the promise of an encyclopedia supported by artificial intelligence (AI).<\/p>\n

Read more Courts for AI Constitutions<\/a><\/p>\n

Kevin Frazier shared a research agenda for understanding the constitutions that govern AI models. Frazier explained that the agenda\u2019s goal is to develop these constitutions in a way that encourages public transparency, oversight, and dialogue that extends beyond the doors of the frontier AI labs.<\/p>\n

Nathan Darmon and Tom Reed proposed court-like institutions developed by AI companies to guide frontier models through complex decision-making. Darmon and Reed argued that an evolving set of precedential decisions would be preferable to a static constitutional document because it would create a richer training corpus for models, improve transparency and accountability to the public, and allow for a few careful decisions to guide the models\u2019 actions at scale.<\/p>\n

On Scaling Laws<\/em>, Frazier sat down with Daniel Kokotajlo to discuss Kokotajlo’s policy proposal to delay superintelligence, which Kokotajlo thinks is necessary to ensure that the disruptive effects of highly-capable AI systems do not outweigh the benefits. Frazier and Kokotajlo discussed scenario scrutiny, feedback from other AI policy stakeholders, and more.<\/p>\n

\n