The Compression Trap
· 2 min read
A company celebrated last quarter. Analyst reports shipping in half the time, AI doing the heavy lifting, six hires off the plan. The earnings deck called it a win.
They had used a capability that could have reshaped the company to run the same company slightly cheaper. The plan hadn't moved; they were simply executing it with fewer people.
When capability gets dramatically cheaper, two questions sit next to each other: how much of the current work can now be done with less, and what just became possible that wasn't before.
The electric motor spent its first thirty years pretending to be cheaper steam. Invented in the 1880s, widely available by 1900, and factory productivity barely moved until the 1920s. Managers kept the old layout and just swapped the power source.
It took a generation for someone to notice that each machine could now have its own motor, which meant the factory could be rebuilt around the work instead of the driveshaft. That's when the assembly line became possible, and the companies that spent those decades optimising the old layout never caught up.
AI is in the same window. Some leaders are treating it as cheaper steam, others as the new motor, and both readings hold up. What the second group builds now is what the rest of us will be living inside ten years from now.