“Responsible marketing” is top of mind for what’s coming in 2024. SAS CMO, @JennChase makes her prediction about the importance of knowing the impact of AI models that you deploy. Read this prediction + more: http://
2.sas.com/6017RTFnR #ArtificialIntelligence #MachineLearning
SAFETY
-
Responsible AI Marketing: Understanding Model Impact in 2024
By
–
-
AI Extinction Risks: Cult Belief or Legitimate Concern?
By
–
Thanks for the question and the opportunity to clarify. What I refer to cultist is the belief endorsed by a few that extinction from AI is so important that all other problems are negligible in comparison. I’m using a strong word because I don’t think it’s right to leave people
-
AI Model Learns Weird Animal Santa Hat Generalization Pattern
By
–
The animals wearing santa hats in the training data has definitely generalised weirdly.
-
AI Extinction Risk vs Addressing Today’s Problems
By
–
No change of mind, Shane. I feel the argument is more nuanced though. I see many tech people focusing on AI extinction, while disregarding today’s problems. I feel that the best way to prevent extinction by AI or any other for is to address today’s problems, and to advance
-
Low-Resource Languages Enable GPT-4 Jailbreak Research
By
–
Exciting news from #NeurIPS2023! The "Low-Resource Languages Jailbreak GPT-4" paper by Zheng-Xin Yong, Cristina Menghini, and @stevebach from the Snorkel Research team has won the Best Paper Award in the Socially Responsible Language Modelling Research (SoLaR) 2023 Workshop!
-
AI Security and Robustness: Critical Missing Considerations
By
–
Having read some of the comments, I feel it would be helpful to provide more nuance. I hope this helps, and thank you all for your comments. The list of current problems is not comprehensive, eg security and robustness of AI systems are missing. They are of course important
-
OpenAI releases AI preparedness framework beta for safety
By
–
We encourage you to read the full Preparedness Framework (Beta) here for more detail. This is a living document, and we expect to regularly update it as we learn more. https://
cdn.openai.com/openai-prepare
dness-framework-beta.pdf
… -
AI Model Safety Baselines: Deployment and Risk Mitigation
By
–
In our safety baselines, only models with a post-mitigation score of medium or below can be deployed, and only models with a post-mitigation score of high or below can be developed further. We also increase security protections commensurate with model risk.
-
Safety Baselines and Governance for AI Model Development
By
–
Our new safety baselines and governance process will turn these technical findings into safety decisions for model development and deployment. This involves establishing a cross-functional Safety Advisory Group to make safety recommendations.
-
Frontier Model Safety Framework Implementation Details
By
–
We are systemizing our safety thinking with our Preparedness Framework, a living document (currently in beta) which details the technical and operational investments we are adopting to guide the safety of our frontier model development.