The Feedback Loop That's Changing Everything
Hofstadter's Law states that everything takes longer than you expect—even when you account for Hofstadter's Law. It's the planning fallacy with a finer point that we can all relate to—your kitchen renovation stretches from one month to three, even when you accounted for two.
But AI development increasingly operates under the opposite principle. There are several reasons for this, but the best explanation is that AI itself is becoming the primary driver of its own acceleration.
AI Building AI: The Self-Reinforcing Cycle
The breakthrough isn't just that AI can code—it's that AI is now accelerating its own evolution.
Here's how AI development used to work: a researcher would dream up a new AI architecture, then spend weeks writing the code to test it. Debugging took months. Running experiments took longer. Analyzing the results and starting over added a few more months. Building AI required small armies of engineers to translate ideas into working software. Now, researchers describe what they want, and AI writes the first draft of the code. Debugging that took weeks now takes hours. Experiments that needed months of setup are running in days.
This creates a compounding feedback loop. Better AI helps researchers build better AI faster, which helps them build even better AI even faster. Each generation of AI doesn't just advance the technology—it accelerates how quickly the next generation can be built. The result is progress arriving years ahead of even many of the most optimistic predictions.
How This Should Drive Your Public Affairs Strategy
It comes down to one simple point for every public affairs practitioner—adopt new tools, or get left behind. In the past, you could stick with a technology product for years. But AI years are like dog years. The tool you adopted last year may already be obsolete. If the tool was available three years ago, before the modern AI era? You're working with Stone Age materials.
And critically, your competitors probably aren't standing still. While you're evaluating whether to adopt new AI tools, your competition may already be using them against you. And because AI keeps getting better at building AI, that advantage compounds and accelerates over time.
The AI race is not like other races. In most technology adoption cycles, being six months behind meant you could still catch up in a year. With AI, being six months behind might mean never catching up at all.
