Calling nonprofits & grant makers/foundations! HAI and @project_evident are conducting a national survey to better understand the current use of, interest in, and opportunity for AI among the social sector. Please participate or help us spread the word: https://
bit.ly/45niPlC
@stanfordhai
-
National AI Survey for Nonprofits and Grant Makers
By
–
-
Stanford Develops Classroom AI Resources for Underserved Schools
By
–
Many high schools lack access to basic AI tools and training, particularly those in underserved, low-income communities. To address this gap, Stanford researchers collaborated with teachers to develop classroom-ready resources for teaching AI. https://
stanford.io/3RGTDDa -
Stanford Researchers Recommend AI Safety Standards for Biomedicine
By
–
To make sure AI does no harm to patients, it’s time to put in place standards and guardrails around the deployment of AI in biomedicine. Here, Stanford researchers outline recommendations:
-
Generative AI’s Impact on Upcoming Election Season Raises Concerns
By
–
"While much of generative AI’s impact on elections is still being studied, what is known does not reassure." HAI International Policy Fellow @MarietjeSchaake on the upcoming election season and the wild card that is AI:
-
Stanford’s Sophia Cuts Language Model Pretraining Time Half
By
–
With the goal of optimizing the pretraining of large language models, Stanford scholars developed an approach called Sophia which cuts that time in half. Find out how they did it:
-
AI Evaluates Dermatology Training for Skin-Tone Bias
By
–
A new study uses AI to evaluate dermatology training materials for skin-tone bias, suggesting the need to incorporate more images of black and brown skin diseases into the training literature. https://
stanford.io/3RFA7a9 -
Nonprofits Survey on AI Impact and Transformation
By
–
Survey for nonprofits: AI is transforming the world as we know it, creating a future where AI will impact humans in ways we can only imagine today. We invite you to be part of this transformation by sharing your thoughts here: https://
bit.ly/3ZD1hAl -
LLM Trustworthiness Risks in Critical Domain Deployment
By
–
“Everyone seems to think LLMs are perfect and capable, compared with other models. That’s very dangerous, especially if people deploy these models in critical domains.” — @sanmikoyejo https://
hai.stanford.edu/news/how-trust
worthy-are-large-language-models-gpt
… (3/3) -
AI Emergent Abilities May Be Overestimated Says Research
By
–
.
@sanmikoyejo and his research team show current perceptions of AI’s emergent abilities might be overinflated: https://
hai.stanford.edu/news/ais-osten
sible-emergent-abilities-are-mirage
… (2/3) -
HAI researcher discusses AI model evaluation with US policymakers
By
–
Last week, HAI Faculty Affiliate @sanmikoyejo headed to DC to meet with @corybooker
, @RepDonBeyer
, @RepYvetteClarke
, @JayObernolte and several other policymakers to discuss his research on the mirage of emergent abilities and model evaluation. (1/3)