As AI tools continue to advance in capabilities and scale, they challenge the traditional notions of fair use and pose complex legal issues. Stanford scholars offer recommendations to address this growing concern.
ETHICS
-
DEFCON Talk Addresses Risks in Technological Transition
By
–
Thank you for informing my talk at the @defcon @xrvillage this week! I think everyone gathering there is excited about the shifts underway and want to help reduce the risks that are coming as part of this transition.
-
LLM Capability and Reliability: Orthogonal Vectors
By
–
This. is. so. true. #LLMs are great – Remember the vectors of capability and reliability and they are often orthogonal. It's true with people as well.
-
Instruction Tuning Risks When Models Lack Training Data Knowledge
By
–
Glad to see this work on alignment; an interesting takeaway for me here is that instruction tuning on data that the model doesn't know can be really, really bad. Basically when you finetune on data the model doesn't know, in addition to teaching it that particular training
-
Stanford scholars advocate proactive LLM governance in medicine
By
–
Stanford scholars urge the medical industry to be more proactive in shaping large language models. The question isn’t how LLMs will change medicine, but what we do so that they're useful for medicine, says HAI faculty affiliate @drnigam
. https://
stan.md/3quLu9F -
Artificial Intelligence And The Law: An Early Warning
By
–
Artificial Intelligence And The Law: An Early Warning
#AI #AIio #BigData #ML #NLU #Futureofwork @oluskayacan @richardkimphd @MariaFariello @johnpearcenews5 @ResurgentAV @rohanmuralee @globaliqx http://
ow.ly/7F2Z30swAB8 -
Les vraies compétences clés selon Jason Wei
By
–
A somewhat amusing personal revelation: a friend recently asked me what were the best skills I had. I think he expected me to say something like "prompt engineering", "writing papers", or "making evals." But my answer surprised him a bit: I said prioritization and communication.
-
Stanford Toolkit Addresses NLP Performance Gaps in English Variants
By
–
The field of natural language processing is fraught with equity challenges. Stanford researchers developed a toolkit to help address performance gaps with English language variants.
-
Latest AI Model Improves Safety and Reduces Hallucinations
By
–
Our latest model has also made improvements on safety. It hallucinates less and is more resistant to jailbreaks, as shown in our automated red-teaming evaluation.
-
AI and Power: Ethical Challenges of Automation and Centralization
By
–
https://
bit.ly/2wqaIXz – AI and Power: The Ethical Challenges of Automation, Centralization, and Scale https://
bit.ly/3OiO1eY #AI #MachineLearning #DeepLearning #LLMs #DataScience