{"version":"1.0","type":"rich","provider_name":"Acast","provider_url":"https://acast.com","height":250,"width":700,"html":"<iframe src=\"https://embed.acast.com/$/64e8d7428a084400127b05c3/6a2ef2ad252d86e846ff5ab3?\" frameBorder=\"0\" width=\"700\" height=\"250\"></iframe>","title":"Short Stuff: AI Companies Are Making Their Chatbots Less Smart!","thumbnail_width":200,"thumbnail_height":200,"thumbnail_url":"https://open-images.acast.com/shows/64e8d7428a084400127b05c3/1781461659931-2d76704f-d5b3-4ae5-8d67-c3b7c55c1f42.jpeg?height=200","description":"<p>In this Short Stuff episode, Kieren and Riku unpack the week Anthropic quietly made its newest model, Claude Fable 5, worse at AI-development questions... without telling anyone, until a backlash forced an apology within 48 hours. They demystify recursive self-improvement (RSI), the safety worry behind the move, ask whether it was real caution or competitive advantage, and explain why a limit you can see is a safeguard while a hidden one edges towards manipulation. Tune in to learn how to spot when your AI is holding back!</p><p><br></p><p>Birds for Scale, Smarties is licensed under <a href=\"https://creativecommons.org/licenses/by-sa/4.0\" rel=\"noopener noreferrer\" target=\"_blank\">CC BY-SA</a>.</p>","author_name":"Kieren Sharma & Riku Green"}