Value impact is always hard to predict.

1) It is, of course, hard to predict major capability advances.

2) But it is also hard to predict capability -> value thresholds. Some large advances won't cross a threshold. While some incremental advancements do.

3) This is all made infinitely harder, because the value chain has many layers and links, each with their own thresholds.

Major capability advances upstream may cross dramatic thresholds, generating reasonable hype, yet still hit downstream thresholds that stymie their impact.

And crossing a few small downstream thresholds can unlock massive latent upstream value, resulting in cascades of impact.

(This is something Apple aims to leverage. Not over-prioritizing major advances vs. the careful identification and clearing of numerous "trivial", but critical, downstream bottlenecks.)