The new AI Index spots major advances in multimodal models, robotics, generative AI, and more.
The new AI Index spots major advances in multimodal models, robotics, generative AI, and more.
Building a Social Media Algorithm That Actually Promotes Societal Values
A Stanford research team shows that building democratic values into a feed-ranking algorithm reduces partisan animosity.
A Stanford research team shows that building democratic values into a feed-ranking algorithm reduces partisan animosity.
Experts in technology, law, and human rights debate the unique implications of this technology and how we might best direct its potential to benefit humanity.
Experts in technology, law, and human rights debate the unique implications of this technology and how we might best direct its potential to benefit humanity.
Why Large Language Models Like ChatGPT Treat Black- and White-Sounding Names Differently
A new study shows systemic issues in some of the most popular models.
A new study shows systemic issues in some of the most popular models.
A new report analyzes the risks of AI and offers potential solutions.
A new report analyzes the risks of AI and offers potential solutions.
In its fifth year, HAI catalyzed a multidisciplinary community of researchers, industry, policy, and civil society to ensure artificial intelligence prioritizes humans.
In its fifth year, HAI catalyzed a multidisciplinary community of researchers, industry, policy, and civil society to ensure artificial intelligence prioritizes humans.
Hospitals are increasingly adopting AI tools for patient care. They need to be thinking about liability.
Hospitals are increasingly adopting AI tools for patient care. They need to be thinking about liability.
HAI Co-Director Fei-Fei Li discusses President Biden’s State of the Union remarks about harnessing the power of AI.
HAI Co-Director Fei-Fei Li discusses President Biden’s State of the Union remarks about harnessing the power of AI.
HAI Senior Fellow Daniel E. Ho shows large language models hallucinate frequently when used for legal queries.
HAI Senior Fellow Daniel E. Ho shows large language models hallucinate frequently when used for legal queries.