Sabine Hossenfelder is one of my favourite science Youtubers - a German physicist who is great at distilling complex topics in a way which is educational, entertaining yet respectful to the audience. Here she is covering a two recent research papers which build the momentum of the argument that AGI is not going to happen - at least, not via LLM’s (we need embodied AI is Hossenfelder’s position). Easily worth
7 minutes of your time.