The key to progress is improving how we improve
Making sure GenAI doesn’t replace our curiosity
Ease Comes at a Price
Every generation of technology promises to free us from drudgery. But each step forward carries a subtle side effect: we risk doing less on our own and becoming more dependent on the very tools we created. Recent examples show that spell-check dulled our proofreading. GPS weakened our sense of direction. And now, GenAI tempts us to hand over even more of our work to opaque machines.
Two Systems, One Choice
Douglas Engelbart saw this coming long before machine learning was a buzzword. In the 1960s, he asked a deceptively simple question: How can we use technology to augment human intellect rather than replace it? His answer wasn’t smarter machines. It was smarter humans, amplified by their tools.
He drew a line between two systems that I’ll call System A (for automation tooling) and system H (for human thinking)
System A — the automation system, where machines do the work.
System H — the human system, where people learn, decide, and evolve.
Our job, Engelbart said, is to strengthen the bridge between the two, not let the A-system take over for the H-system.
That’s the problem with how GenAI is being offered today. The danger isn’t that it will take our jobs, but that it will dull our curiosity. The risk isn’t sudden replacement; it’s gradual atrophy.
If we’re not careful, we’re likely to find ourselves, as E. O. Wilson observed: “... drowning in information while starving for wisdom.”
Retaining the Work of Thinking
The goal right now should be to use AI to help humans see and understand the work more clearly. To ask better questions. To explore more possibilities.
Used in this way, AI can act like a powerful mirror: reflecting how we currently think, showing us where we are missing important elements, and exposing hidden assumptions for us to address. That’s applying System-A and System-H together to help us improve. That’s augmentation.
A Lesson That Still Holds
Engelbart’s lesson still stands:
“The key to progress is improving how we improve.”
I write Signals for people designing the next generation of APIs and intelligent systems. Each issue gives you clear examples, frameworks, and tools to use right away. Subscribe now and design smarter with AI, not just around it.