Anthropic is groundbreaking a concept referred to as “constitutional AI” – that's correctly a way to make an LLM safer by ensuring it “gains” values by using a constitution provided. OpenAI denied stories that it's intending to launch an AI design, code-named Orion, by December of the year. An OpenAI https://titusctiwh.onesmablog.com/the-smart-trick-of-ai-bot-that-no-one-is-discussing-71811700