End Game Thinking
A compounding strategy skill I built for myself (and my team).
The spark
In early February 2026, the internet was saturated with a particular kind of fear. AI is coming for technology jobs. Your career is not safe. It was scary then, and it's still scary now. It was scarier before I started educating myself and actually using the tools enough to understand their constraints and possibilities firsthand.
Then I came across Michael Bloch's End Game Positions, riffing on Will Manidis's essay End Game Play. The core argument: across chess, war, and technology, the middlegame has collapsed. Everyone reasons backward from the terminal state. Bloch applied it to investing: if AI becomes as powerful as it could possibly be, what do you want to own? Energy. Atoms. Capital. Regulatory permission. Trust.
I read it and thought: the same logic applies to me. Not as an investor, but as a design leader watching AI reshape everything around me. What positions should I hold?
But sitting with the worst-case scenario didn't produce paralysis. It produced a different question: if you could change everything about what you're doing, what would you actually do? That reframe (from scarcity to abundance, from "what am I losing" to "what could be possible") turned out to be the most valuable thing the whole exercise produced. So I built a skill to run it again, and again, and again.
Scarcity to abundance - the reframe that turned out to be worth running on repeat.
Scarcity to abundance - the reframe that turned out to be worth running on repeat.
What I built
End Game Thinking is a structured strategic foresight exercise, packaged as a skill I can run on demand. It stitches together five moves I kept reaching for separately: backward induction, inversion, premortem, regret minimization, and scenario stress-testing. Each run pulls fresh industry research, applies those five lenses to my career and my team, and produces a written artifact.
A single run generates:
- A synthesis of what shifted in the AI, design, and enterprise landscape
- A positioning read on my role, my team, and the design org I'm part of
- A "five positions to hold" brief - the bets worth defending this quarter
- A "what remains" artifact - the design skills that compound when software gets cheap
- A resilience tracker logging how each run's calls aged, captured in Notion so the data compounds across runs
That last one is the whole point. Anyone can run a one-off strategy exercise. I wanted a system that got sharper every time it ran.
The tracker is the point. The memo is a byproduct.
The tracker is the point. The memo is a byproduct.
The compounding layer
A single premortem is a session. Twelve premortems across a year is a dataset. The tracker surfaces patterns I can't see in any single run: which of my predictions aged well, where the industry moved in a direction I didn't expect, which "unexpected competitor" roles actually converged on design territory (and which ones I overreacted to). The April check-in is already showing things my January self couldn't have caught.
This skill and its compounding effects are an example of the second-order value you can get from AI tool use. I'm not using AI to write strategy memos faster. I'm designing a system that accumulates a longitudinal read on my positioning, mine alone. That's not something you can buy off a shelf or read in a newsletter.
Second-order value: not faster memos, a longitudinal read on your own positioning.
Second-order value: not faster memos, a longitudinal read on your own positioning.
What it's surfaced
A few things I didn't expect:
- The "unexpected competitors" showing up in design aren't who the consensus says they are. My list is different, and I'd have missed that without forcing myself to write it down each month.
- The skills that hold up in a post-software world are less about craft and more about judgment: which framing of the problem, which pattern, which constraint to defend.
- The scarcity-to-abundance reframe doesn't stay fixed. What my January self flagged as a threat, my April self reads as a capability, and the reverse is just as true. The reframe isn't a one-time move, it's what the cadence keeps teaching me.
Want to run this on your own career? The skill is open source. Clone it, point it at your work, and tell me what breaks, or what it surfaces that I missed.
Why I think this matters
I started working in technology (and stayed in it) because I genuinely love making cool things with tech, and using AI tools is the most fun I've had in a long time. My HCC training at UMBC taught me to think in systems, and this skill is what happens when those two instincts show up in the same place: a personal operating system for staying ahead of a market that's moving faster than any of us can track by reading about it.
I'm not trying to predict the future. I'm trying to have a durable point of view when it arrives, and to give my team one too.
Let's connect.
Open to conversations, collaborations, and meetups.