>Moore’s law appears to be a special case of Wright’s Law. I.e. it seems well explained by experience curve effects (or possibly economies of scale).
If that’s the case, it makes my points stronger, cheers!
>I don’t see these strong reasons.
Suppose we discover a new trove of documents from the industrial revolution. Do you expect this to shift our view of the agricultural revolution and of future tech development? Or, if we make uploads earlier or later than expected, will that shift our view of the industrial revolution? All these changes seem fully explained by trends that are local to the change (in terms of time and geography).
>Age of Em gives some hints (page 14) that the last three transitions may have been caused by changes in how innovation diffused, maybe related to population densities enabling better division of labor.
The changes can exhibit similar features without being strongly connected (especially in the mathematical-exponential way that Kurzweil and Robin are proposing). It’s like my example with Goodhart’s law: once we have more detailed explanations, we no longer need the outside view. Maybe if all we knew about was the fact that the various transitions happened, then we could usefully conclude “large changes are possible.” But once we start to learn about the various transitions and about human nature, then very quickly the outside view stops having extra explanatory power.
Or to use another example: almost all successful innovations are purchased by the rich first, and then price drops and they access the mass market. This is a feature of almost all these innovations. But I can’t conclude from this common feature that innovations will share other, unconnected features.
>Also, your link to “a previous post” is broken.
Thanks, corrected.
>Moore’s law appears to be a special case of Wright’s Law. I.e. it seems well explained by experience curve effects (or possibly economies of scale).
If that’s the case, it makes my points stronger, cheers!
>I don’t see these strong reasons.
Suppose we discover a new trove of documents from the industrial revolution. Do you expect this to shift our view of the agricultural revolution and of future tech development? Or, if we make uploads earlier or later than expected, will that shift our view of the industrial revolution? All these changes seem fully explained by trends that are local to the change (in terms of time and geography).
>Age of Em gives some hints (page 14) that the last three transitions may have been caused by changes in how innovation diffused, maybe related to population densities enabling better division of labor.
The changes can exhibit similar features without being strongly connected (especially in the mathematical-exponential way that Kurzweil and Robin are proposing). It’s like my example with Goodhart’s law: once we have more detailed explanations, we no longer need the outside view. Maybe if all we knew about was the fact that the various transitions happened, then we could usefully conclude “large changes are possible.” But once we start to learn about the various transitions and about human nature, then very quickly the outside view stops having extra explanatory power.
Or to use another example: almost all successful innovations are purchased by the rich first, and then price drops and they access the mass market. This is a feature of almost all these innovations. But I can’t conclude from this common feature that innovations will share other, unconnected features.
>Do you expect this to shift our view of the agricultural revolution and of future tech development?
A sufficiently large trove might well shift my view of those.
>Or, if we make uploads earlier or later than expected, will that shift our view of the industrial revolution?
A large surprise would shift my view a bit.
>All these changes seem fully explained by trends that are local to the change (in terms of time and geography).
You repeat your conclusion without pointing to strong reasons. I take that as evidence that you don’t have strong reasons.