Perhaps in the light of recent announcements about the real and potential dangers presented by large AI models, Sivasubramanian's blog post emphasises the safety provisions and the pursuit of ...