I don't think most people know what to do right now. That's the honest read. AI is reshaping the economy in real time and the playbook hasn't been written yet. You see it everywhere. The quiet reorgs. The headcount that just sort of evaporates between quarters. Nobody is speculating anymore. We're in it.
The default reaction is defensive. People look for hard assets. They want to buy land or gold and wait for the social upheaval to pass. This is a fatalistic exit. It treats the next decade as a blast to be survived. It skips the middle entirely.
In chess, they'd call this the middlegame. It's the messiest part of the board. The opening theory is over (AI is going to be massively impactful on society) but the endgame is far off (we don't know exactly how it's gonna play out). The pieces are tangled and the ground shifts under every move. This is where the game is won, even if the king hasn't fallen yet.
We are in the Great Repricing of intelligence. For fifty years, Western society ran on the scarcity of cognitive labor. We built a middle class around credentials and "knowing things." That was the moat. Commoditized intelligence is filling it in faster than most people want to admit.
The Triangulation
To see the Middlegame clearly, you have to watch multiple signals at once. Here are three I keep coming back to, but they're not the only ones. The important thing is reading them together so you don't misinterpret any single one.
Watch the capital. Real money is moving into energy, data centers, and physical infrastructure. These are massive, illiquid bets on the new floor of the global economy. Capital moves first.
Watch the layoffs. Companies staffed for a world where white-collar labor was expensive and hard to find. Now that it's abundant, headcount is repricing to reflect the new supply. Some of this is just bad management seeking cover. Viewed in isolation, you could explain any individual round of cuts away. But the aggregate trend is structural.
Watch the emerging roles. Ghost roles that exist before HR has a name for them. Translation layers and system architects. People doing the work before the category is settled.
Any one of these could be a false positive on its own. Triangulate all three and you see one connected shift. The old barriers of skill and knowledge are collapsing.
The Case for Speed
It used to be that we optimized for precision. You spent six months on a strategic plan. You spent four years on a degree. Precision is a luxury of stable systems. It works when the rules of the game stay the same while you are planning your move.
In the Middlegame, precision can trick you into moving slow. And moving slow costs more than being slightly wrong. The information environment is always on. Everyone has access to the same tools and the same commoditized intelligence at the same moment. The playing field is level. Execution is the only remaining edge.
If you wait for complete information, you are arriving at the endgame having built nothing. The cost of leaning in and being wrong is a few wasted weeks. The cost of waiting is a portfolio of assumptions priced for a world that no longer exists.
This is the 80/20 era. Good enough and shipped beats perfect and planned. You don't think your way through the Middlegame. You take shots on goal. You move before you feel ready because the tools let you adapt on the fly.
Three Grounded Examples
The machine does not need to fully replace a person to change the economics of a job. It only needs to make the output cheaper and easier to standardize. Here is how that looks in practice.
The Systems Architect. Consider the mid-level professional whose value was tied to a specific cognitive output. Financial modeling. Research synthesis. In the Middlegame, they realize that output is now a commodity. So they stop performing the task and start architecting the system that automates it. They become the person who manages the efficiency engine. They trade "knowing how to do X" for "owning the system that does X."
The Translation Layer. AI benchmarks are high, but real workflows are messy. There is a massive gap between what a model can generate and what a business actually needs to ship. The Translation Layer sits in that gap. They don't have an AI title yet. They just identify where the model fails the workflow and provide the human refinement that makes the product actionable. They turn raw intelligence into finished business results at ten times the traditional speed. That's where the value is.
The Iterative Executive. The old executive model was built on certainty. In the Middlegame, that leader is a bottleneck. The new executive treats every integration as a low-cost experiment. They shift the team's focus from accuracy on the first attempt to the number of iterations per week. They accept a high failure rate because the cost of failure is negligible compared to the cost of being slow. Their moat is cycle time.
The Agency Differentiator
Access to tools is universal. Access to commoditized intelligence is universal. So what's left?
The people I watch pulling ahead right now share one trait. They move before they're comfortable. They pick a problem, point the tools at it, ship something ugly, and learn from the output. Then they do it again on Tuesday. They don't wait for a job title that gives them permission. They don't wait for a course that makes them feel qualified. They just start building and let the role form around what they've built.
The people I watch falling behind also share one trait. They're still preparing. They're reading about AI. They're "keeping an eye on things." They're waiting for the right opportunity to present itself cleanly. It never does. The Middlegame doesn't reward preparation. It rewards contact with the problem.
Here's the thing nobody tells you. The first version is going to be bad. The first automation you build will be clunky. The first workflow you redesign will half-work. That is the point. The person who has shipped ten bad versions has a better map of the territory than the person who has been planning their first move for six months.
The gap between people who thrive in this transition and people who get swept up in it is not intelligence. It's not access. It's the willingness to look stupid in public while you figure it out.
Start before you're ready. Build the thing. Show someone. Fix it. That's the whole framework.