Using an AI model, researchers found just how underrepresented black and brown skin is in frequently used medical training materials. https://
stanford.io/3Rcbulg
ETHICS
-
AI Reveals Underrepresentation of Black and Brown Skin in Medical Training
By
–
-
Cultural Sensitivity in Online Spaces Fosters Global Inclusivity
By
–
4/Cultural Sensitivity: In the online space, you interact with people from diverse backgrounds. Being culturally sensitive fosters inclusivity and respect. This can also be a good opportunity to learn about different cultures from around the world.
-
Empathy Matters: Building Positive Virtual Team Environments
By
–
3/Empathy Matters: Remember, there are humans behind those screens. Empathy and understanding are crucial in remote teams. Check in on colleagues, offer support, and build a positive virtual environment.
-
AI Data Use Ethics: Consent Without Coercion in LLM Training
By
–
Finally, where AGI l, LLMs or any AI system uses human data and said humans don’t fully understand how their data is being used, any faith metaphor must use a forces proselytizing modality versus a shared faith a listener would adopt by choice versus societal duress.
-
AGI Faith vs Innovation: Need for Respectful Disagreement
By
–
Note having people on staff who respectfully disagree with the cult-like stance would provide greater innovation and understanding overall. But blind faith in a heretofore not universally defined definition of AGI makes it a cult versus an open, non proselytizing faith.
-
AGI Belief as Faith-Based Speculation: Cult or Reality?
By
–
Believing AGI is inevitable is at best, faith-based speculation. And at worst, or by another name, a cult. So not being hired is a form of religious discrimination (or maybe vice versa) where a key part of the AGI cult is protesting it’s not faith based speculation.
-
Aligning Large Language Models with Human Values
By
–
Aligning Large Language Models with Human: A Survey by Yufei Wang, Wanjun Zhong, Liangyou Li, Fei Mi, Xingshan Zeng, Wenyong Huang, Lifeng Shang, Xin Jiang, Qun Liu
-
Halo: Reducing Hallucinations in Weak Open-Source LLMs
By
–
Halo: Estimation and Reduction of Hallucinations in Open-Source Weak Large Language Models by Mohamed Elaraby, Mengyin Lu, Jacob Dunn, Xueying Zhang, Yu Wang, Shizhu Liu
-
Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback
By
–
Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback, by @StephenLCasper @alxndrdavies @causalclaudia @sociotiose @jeremy_scheurer @javirandor @FreedmanRach @tomekkorbak @davlindner @5kovt @CRSegerie @MicahCarroll
, et al. -
Culture Interview Should Be Renamed Cult Interview
By
–
The “culture” interview should be renamed to the the “cult” interview.