In this post, we explain what Constitutional AI is, what the values in Claude’s constitution are, and how we chose them:
ETHICS
-
Constitutional AI Enables Explicit Values for Language Models
By
–
Our research on Constitutional AI allows us to give language models explicit values determined by a constitution, rather than values determined implicitly via large-scale human feedback.
-
Constitutional AI: Encoding Values Into Language Models
By
–
How does a language model decide which questions it will engage with and which it deems inappropriate? We use Constitutional AI to more directly encode values into our language models.
-
Constitutional AI Approach and Training Framework Published
By
–
We’ve now published a post describing the Constitutional AI approach, as well as the constitution we’ve used to train Claude:
-
AI’s unprecedented magnitude of benefit and risk in human history
By
–
Yes, and the magnitude of potential benefit and risk for AI is greater than any technology in human history.
-
AI Safety Concerns, Tech Competition, and Code Interpreter Advances
By
–
Marketing AI Show podcast ep 46 dropped today. 00:02:42 — Hinton leaves Google and raises fears about AI destroying humanity
00:11:14 — Google and OpenAI have “no moats?”
00:24:21 — The things about AI that give me hope for the future
00:35:51 — @OpenAI Code Interpreter shows -
SAS Viya Video Series: Cloud Economics, AI Ethics, Performance
By
–
Watch the @SASsoftware short video series "Just the Facts" 1) Cloud Economics with #SASViya: https://
youtube.com/shorts/1bhinga
NO58
… 2) #AI Ethics with SAS Viya: https://
youtube.com/shorts/-kzqYn3
hlqI
… 3) Get Faster Results with SAS Viya: https://
youtube.com/shorts/D8dLmYL
BpJw
…
——
#SASVisionary #SASInnovate #DataScience #ML -
SAS Presents Trustworthy AI and Responsible Innovation Principles
By
–
Great presentation on Trustworthy #AI at @SASsoftware #SASinnovate The Core Principles of Responsible Innovation at SAS: https://
sas.com/en_us/company-
information/innovation/responsible-innovation.html
… Model Cards are a great example!
————
#SASVisionary #MachineLearning #DataScience #ML #TrustedAI #AIEthics #GenerativeAI #MLOps -
SAS Addresses AI Ethics Amid Generative AI Rise
By
–
Live at #SASInnovate 2023 by @SASsoftware It's great to hear how SAS is addressing AI Ethics, especially with the rapid rise in #GenerativeAI, #ChatGPT, and #LLMs (Large Language Models), which have met with both high #AI hopes and doubts!
————— #SASVisionary #ML #DataScience