Related: this is the outcome the other three pieces in this series are actually describing — Kingdom: AI, The Scaffolding Prescription, and Knowledge Distribution are what it takes to end up in the ring below.
Picture a cone standing on its point, opening upward and outward. That's most people's relationship with AI right now, and it's a real, valuable thing — being a competent user of a capable model makes you meaningfully faster at the job you already had. Call it 20%, give or take. That cone is wide and getting wider every month, because it's genuinely not hard to get into: use the tool well, and the tool gives something back. More people join it constantly. That's not a criticism. Most value in most industries gets created by exactly this kind of broad, real, incremental gain.
Now picture something separate from the cone entirely — not its tip, not its highest point, a different shape altogether: a ring, floating above the cone's opening, not touching it. That's the small group — maybe five percent — who aren't using AI to do their existing job faster. They're using it to multiply what they can build in the first place. Output that isn't 20% more. Output that's orders of magnitude beyond what one person's hours could produce alone. The cone describes people getting better at their job. The ring describes people who stopped having a single job in the recognizable sense at all.
Two Kinds of Faster
The distinction that actually matters isn't skill level. It's what kind of gain you're stacking. Inside the cone, gains add: you plus a faster version of the thing you already did. Every person in the cone is still fundamentally bounded by their own hours in a day, just spending those hours more efficiently. That's addition, and addition is genuinely good — I'm not being sarcastic about the 20%, it's real money, real time, real relief.
The ring is multiplication, and multiplication doesn't come from doing your existing job faster. It comes from stopping being the bottleneck at all — running systems instead of tasks, agents instead of actions, structure instead of individual effort. Knowledge Distribution already covers the mechanism: addition is what you get leading followers, multiplication is what you get leading leaders. Substitute "agents" for "leaders" and you have the exact same math, running on a different kind of team.
What It Takes to Cross
Getting from the cone into the ring isn't a matter of using the tool longer or harder. It takes a specific, learnable combination: the nerve to build rather than just use — to commit real time to infrastructure instead of taking the immediate, smaller win. The ability to juggle a lot at once without dropping a thread, which is a discipline, not a talent (see: The Scaffolding Prescription). And systems-thinking — reaching for a real structure instead of describing output directly, the exact instinct behind Kingdom: AI. None of the three is exotic. All three are buildable. None of them show up automatically just because you've used a chatbot for a year.
The Jump, Not the Climb
Here's the part that matters most and is easiest to get wrong: the ring isn't the top of the cone. There's no gradual ascent where you get incrementally better at your job until one day you notice you've crossed over. It's a discontinuous jump, closer to what Gestalt psychologists documented decades ago as insight problem-solving — you're stuck, working the problem the old way, and then you're not, all at once, with no gradual convergence in between. Wolfgang Köhler watched it happen in a chimpanzee suddenly connecting two sticks to reach a banana it had been staring at for ages. Something restructures below the threshold of awareness, and what surfaces isn't a slightly-better version of the old approach — it's a completely different relationship to the problem. That's what crossing into the ring actually feels like. Not "I got faster." "I stopped asking it to do my task and started asking it to run my system."
What to Do Right Now
If you're in the cone — and almost everyone reading this is, including a version of me not long ago — the honest move isn't to feel bad about the 20%. It's real, it's worth having, and dismissing it is its own kind of arrogance. But if you want the ring, stop optimizing the task in front of you and start asking a different question entirely: not "how do I get this done faster," but "what would this look like if I built the thing that does this, instead of doing this." That question is the whole jump. Everything else — the taxonomy, the scaffolding, the distribution — is just what it takes to survive on the other side of it.
— J.P. Howlett
Related:
- Kingdom: AI — the systems-thinking third of the jump: instructions that build structure, not just output.
- The Scaffolding Prescription — the juggling-a-lot third of the jump: running multiple threads without losing your place.
- Knowledge Distribution — the multiplication math underneath the ring, applied to people instead of agents.
Sources
- Gestalt's Perspective on Insight: A Recap Based on Recent Behavioral and Neuroscientific Evidence — the discontinuous, non-incremental nature of insight problem-solving this piece's "jump, not climb" section is built on.