– Higher compute and confidence were extra important for model performance when wrong answers had serious consequences (like when they were weighted 20x). You wouldn't want a 51%-sure diagnosis of a terminal illness.
ETHICS
-
Johns Hopkins: LLMs Must Know When to Abstain
By
–
You can't just be right, you have to know you're right. Good advice for LLMs, according to new Johns Hopkins research. Sometimes no answer is better than a wrong one – life or death choices in medicine, for example, or big financial decisions.
-
AI Privacy Concerns and Cognitive Capabilities
By
–
AI knows our deepest secrets.
— Linus ✦ Ekenstam (@LinusEkenstam) 19 mars 2025
Dreams of flight.
pic.twitter.com/piVsLAeaK0AI knows our deepest secrets. Dreams of flight.
-
How LLMs Expand Thinking and Reveal Hidden Insights
By
–
LLMs like Gemini and ChatGPT are giving me more time to think, and prompting me to see things I might otherwise overlook.
-
California releases draft policy principles for frontier AI models
By
–
The Co-Leads and collaborators of the Joint California Policy Working Group on AI Frontier Models released a draft report to develop policy principles that can inform how California approaches the use, assessment, and governance of frontier AI. The working group is seeking public
-
Stanford Scholars Urge Four Policy Priorities for AI Action Plan
By
–
Scholars from @StanfordHAI
, @StanfordCRFM
, and Stanford RegLab urge policymakers to prioritize four areas of policy in the development of an AI Action Plan. Read more about their response to @WHOSTP
’s request for information: https://
hai.stanford.edu/policy/respons
e-to-ostps-request-for-information-on-the-development-of-an-ai-action-plan
… -
AI Writing 90% of Code Within Months Says Anthropic CEO
By
–
Are coders actively and happily coding themselves out of a job over the next several years? CEO of @AnthropicAI said last week: “We are not far from the world—I think we’ll be there in 3 to 6 months—where AI is writing 90% of the code. And then in 12 months, we may be in a world
-
Differentially Private Synthetic Data Generation via LLM Prompting
By
–
Today we describe an inference-only approach to generating differentially private synthetic data via prompting off-the-shelf large language models with many examples in parallel and aggregating their responses in a privacy-preserving manner. Learn more at https://t.co/lPoft9PXpM pic.twitter.com/YtLRqc2FyI
— Google AI (@GoogleAI) 18 mars 2025Today we describe an inference-only approach to generating differentially private synthetic data via prompting off-the-shelf large language models with many examples in parallel and aggregating their responses in a privacy-preserving manner. Learn more at https://
goo.gle/4izufJT -
Stanford RAISEHealth Awards Seed Grants for Ethical AI Biomedicine
By
–
Just announced: @Stanford
's #RAISEHealth initiative, a collaboration between @StanfordMed and @StanfordHAI
, has awarded seed grants to projects focused on the ethical, responsible, and safe development of AI in biomedicine.