What’s next if AGI does not happen?

Is all the talk about robotics, automated vehicles, and world models an acknowledgement that the LLM scaling era has plateaued? Is it time to focus on more realistic use cases than the AGI / Super-intelligence hype?