Evaluating new technologies
Be mindful of the adoption lifecycle
Is this the right time? * Does the current economics of CPU/memory/bandwidth support adoption of this tech? * Is this the right tech at the wrong time? * What would need to change in the economics of compute/bandwidth etc. to make this work? Because that's what might make it work next time
Is it a platform or a feature?
Is it a big enough improvement to drive change behaviour? a 20% decrease in cost probably won't motivate change 50%+ does
Does it remove significant barriers to how we used to do it?