Eleven OpenAI employees and two from Google DeepMind signed an open letter saying leading AI companies cannot be trusted with keeping society safe from AI's harms.
Posts published in “AI Models/Tools”
French startup Mistral AI has released its first code-generation model, Codestral. It is open source with a 32k token context window, which it said is larger than those of its competitors.
MIT's 'The Consensus Game' shows promise for improving the performance of base models, which could result in more reliable and factual outputs from ChatGPT and similar language models.