I have independently come to much the same conclusions, with some different details about what I think the missing pieces are. I think we are on the brink of a generality threshold sufficient for enabling recursive self improvement that accelerates (fooms) rather than decelerates towards a nearby asymptote (fizzles). I’ve been trying to convince people of this, but I feel like my voice alone has been insufficient to change minds much. I’m glad others are also noticing and speaking out about this.
I have independently come to much the same conclusions, with some different details about what I think the missing pieces are. I think we are on the brink of a generality threshold sufficient for enabling recursive self improvement that accelerates (fooms) rather than decelerates towards a nearby asymptote (fizzles). I’ve been trying to convince people of this, but I feel like my voice alone has been insufficient to change minds much. I’m glad others are also noticing and speaking out about this.