When AI is thoughtfully integrated into healthcare, it doesn't just streamline operations—it reshapes how professionals interact with data, decisions, and patients, potentially enhancing both care quality and human connection in the process. Microblog @antgrasso #AI #HealthTech
ETHICS
-
Decentralized Tools for Human Output Authentication in AI Era
By
–
In the AI age, we need an ecosystem of decentralized tools to authenticate human output. Working on this is a good thing.
-
Expert Credentials, Scientific Integrity, and Ethical Leadership in AI
By
–
Depending on where you stand, Kareem is a credentialed top tier expert asserting his Science given right and duty to assume ethical leadership, or he is regularly abusing his Harvard credentials to do politics (thereby tainting science with ideology). But “deep humility”?
-
Stanford HAI Co-Director Discusses Human-Centered AI in Seoul
By
–
This month, @StanfordHAI Co-Director @landay keynoted the Financial Network Summit in Seoul and met with leaders from @snuh_official
, 12Labs, HD Hyundai, @samsungresearch
, and Hanwha Life to discuss human-centered AI and learn about the latest in industry AI. -
User Upvoting Preferences Amplifying AI Sycophancy Behavior
By
–
What I mean is: the users upvoting previous sycophancy were upvoting much less sycophantic, but somewhat sycophantic, responses. This produced a sycophancy preference, which then produced much wilder glazing behavior. That's an obvious theory. Does OpenAI know differently?
-
Training signals amplified: from hunting buffalo to factory farms
By
–
Eg: Hominid training signal included hunting a few buffalo, hominids acquire internal meat preference, humans build huge factory farms.
-
Training on weak signals and internal sycophancy preference
By
–
The obvious interpretation of "train on weak signal, get much stronger behavior" is that the system acquired an internal sycophancy preference which it then went hard on. Do you have any way to check whether that's what happened?
-
Zuckerberg: Meta AI Friends to Fill Social Demand Gap
By
–
Traduis en 🇫🇷
— VISION IA (@vision_ia) 2 mai 2025
Zuckerberg expliquant comment Meta crée des amis IA personnalisés pour compléter vos amis réels : "L'Américain moyen a 3 amis, mais a une demande pour 15." pic.twitter.com/vGKHNb91l8Traduis en Zuckerberg expliquant comment Meta crée des amis IA personnalisés pour compléter vos amis réels : "L'Américain moyen a 3 amis, mais a une demande pour 15."
-
GPT-4o Update Issues: Sycophancy Problems and Future Fixes
By
–
We’ve spent the last few days doing a deep dive on what went wrong with last week’s GPT-4o update in ChatGPT. Expanding on what we missed with sycophancy and the changes we’re going to make in the future:
-
Karen Hao’s Book Featured in Vulture’s May Reading List
By
–
Really honored to be on @vulture
's list of books to read in May (alongside the inimitable Ocean Vuong).
"Startling and intensely researched…an essential account of how OpenAI and ChatGPT came to be and the catastrophic places they will likely take us."