Things that are part of every tech transition arc: • Hardware is a key enabler (PC, graphics, networking, ARM, Servers, etc) but all hardware is cyclical and prone to leap-frogging because hardware is modular. Software is the "sticky" part of a transition because it is less modular and subject to abstraction that makes it stick around. Software assets are key. • Component shortages are one aspect of the initial hardware wave: RAM, GPU cards, HDD (each form factor), Flash, and on and on. • Synthetic metrics define "too soon". Early in an arc finance and analysis are desperate for metrics that matter but all of them are either going to zero or infinity (meaning they cost ~nothing or the supply is ~infinite). MIPS was a measure for like a year of PCs because that is what mainframes measured. Today's include: gigawatts, tokens, context, etc... • Benchmarks are the technical metrics that define "too soon" or "not what people really do". These take the longest to correct and often remain for decades but continue to be irrelevant. SQL TPS, web page weight, broadband penetration come to mind. • Market share is what everyone wants to know but the market isn't a thing yet. Share of X isn't meaningful when even the distribution channel is unclear or when the numerator is computed differently by vendors. • M&A is the Wild West. There's massive amounts of speculation channeling all sorts of strategery usually based on either vertical integration or horizontal expansion. Many prematurely talk of a "consolidation" as though the market was defined. • There is a "disturbance in the arc" when a player in the context of one of the above either overplays its hand or the narrative shifts to a new metric/measure that isn't in their favor. For example, share -> revenue, revenue -> share, or metric does in fact go to zero/infinity because of some innovation. This reset is critically important to the arc as it usually defines a time when "now things are getting clearer". eg PC BIOS, Windows, iPhone touch, Chrome, Salesforce, etc.