Independent questions. Shared human futures.Powered by The Human Question ↗
Work & everyday lifeContext, evidence, perspective
Where AI Fits.

Understand the change. Choose the part you play.

Work

Buying the tool is only the first investment

Research on general-purpose technologies shifts attention from the machine to the changes needed around it.

A subscription invoice is an unusually clear number in a murky business decision. A manager can see what an AI tool costs each month. It is harder to price the effort of changing a process, teaching colleagues and deciding which outputs deserve trust. Our editorial starting point is that a useful adoption decision should consider that surrounding work. A purchase alone cannot show whether an organization has learned to use a capability well.

Timothy Bresnahan and Manuel Trajtenberg offered a broader lens in their 1992 work on general-purpose technologies. They examined technologies used across many sectors, capable of improvement and linked to innovation by their users. Electric motors were among their historical examples. In this account, opportunity depends partly on developments on both sides of a relationship: the enabling technology and the activities built around it. This is an economic model, not a promise that every new product transforms an economy.

Paul David approached the issue historically. His 1990 article, The Dynamo and the Computer, examined factory electrification and the productivity puzzle surrounding computers. Individual electric motors permitted changes in factory construction, machine placement and materials handling. The significance extended beyond exchanging one power source for another. The article makes organizational adaptation central to its comparison; it does not tell us how many years an AI transition must take.

A later study gives measurement a role too. Erik Brynjolfsson, Daniel Rock and Chad Syverson model how complementary intangible investments can affect measured productivity. Their productivity J-curve describes how early investment may be missed and later benefits recorded differently. It should not be converted into a guarantee that any disappointing rollout will eventually become profitable. A measurement explanation and a successful business strategy are different claims.

One practical implication we draw is to make the surrounding work explicit. What information must be prepared? Who checks the result? Which failures require a person to intervene? What must employees learn, and how will the organization know that the output is useful? These are proposed questions for a review, not a universal checklist validated by the studies. They keep the focus on an actual process rather than the prestige of a product name.

The electricity comparison is valuable when it directs attention to complements and weak when it becomes a borrowed timetable. A small experiment can test a clearly specified task; a larger investment needs evidence about a larger system. That is the distinction worth carrying into the next budget meeting. The first visible payment may buy access to a tool. The more consequential decision concerns the work people choose to build around it.

Sources & further reading

Original editorial analysis informed by the sources below. A source’s date, method and limits matter.

  1. General Purpose Technologies: Engines of Growth? Scope & limits · S07
  2. The Dynamo and the Computer: An Historical Perspective on the Modern Productivity Paradox Scope & limits · H01
  3. The Productivity J-Curve: How Intangibles Complement General Purpose Technologies Scope & limits · H02