Boost Illusion
"Nobody buys efficiency. Ever"
So runs advice handed down to each new generation of Enterprise reps.
Riffing on the famed prediction of JM Keynes. Namely that due to inexorable technology leaps we should have long since been only required to work a 15-hr week.
Feel like that today for you?
Here's another recent finding on the march of efficiency from our big shake-up trend de jour.
"After analyzing data about 25,000 workers across 7,000 workspaces, users of AI only saved on average three percent of time. And only a meagre three to seven percent of those productivity gains translated into bigger paychecks, they found."
🚨 3% of 3% is zero. Who buys that?
Neither sufficiently more efficient nor effective to justify the huge sums being sent AI vendors' way.
As I noted earlier in 2025, there's a growing mindset that AI is far from nearing any workplace panacea of promise.
Indeed, seasoned observers cast doubt on policies that decimate junior roles at the expense of new AI-replacing initiatives. Alluding to how dramatic effects from removing the energy of younger colleagues seed wide-ranging corporate jeopardy.
Yet hope fountains forth.
Earlier this year a UK civil service study - of 20,000 roles no less - astonishingly concluded that just under a sixth of worktime was proven saved by AI. The quoted (extrapolated) percentage time freed up being almost 18pc. Practically a day a week.
Efficiency gains to make the eyes bulge.
Yet fantasy.
What was missing? As is ever the case, silence around what will be done in all that miraculously available freed up time.
Instant new passport delivery? Hospital waiting times disappear? Infrastructure fixed on time, under budget?
Allowing 30,000 now unnecessary public workers to seek challenge afresh in the private sector?
Fat chance.
Therein lies the conundrum inside our business case woe.
Because unlike on the govt payroll, we are not immune to such questions.
Competing a greater volume in the same or less time is but one of our prime lever trio.
The ever-faster angle sought here joined by price and quality.
Where we must give solid answers for any pair of these three.
While seemingly beyond the current restraint on the world of AI, can you?