Nobody Knows What You're Worth Anymore | The AI Job Market Reality
TL;DR
AI made generation cheap, so output no longer proves value — Nate B Jones argues the old chain of “hard work → visible effort → demonstrated expertise” is collapsing because anyone can now ship code and apps fast with tools like Claude, ChatGPT, and Lovable.
This isn’t just a junior hiring problem — it’s a system-wide talent allocation problem — he frames 2026’s tech job crisis as broader than entry-level pain, pointing to Oracle cutting up to 30,000 jobs, Amazon 16,000, Block 4,000, Dell 11,000, and 60,000+ confirmed tech cuts in Q1 alone.
Comprehension is the scarce skill now, not production — his first principle is that people need to deeply understand what AI helped them build: why it works, what would break, what trade-offs they made, and when they overrode the model.
Explanation has to travel with the work like a modern commit message — Nate says every project should ship with a concise artifact answering four things: what it is, why it was built this way, what might break, and what was learned.
Resumes and credentials are too slow for AI-speed work, so we need ‘microtransactions for jobs’ — instead of relying on a role held for 2-4 years, he wants more granular proof that someone did meaningful work and got paid for it in compressed timeframes.
His product pitch is Talent Board: a public profile for AI-era proof of work — the idea is to gather scattered repos, Claude artifacts, Lovable projects, and explanations into one visible place so you don’t get dismissed as a “slop factory.”
The Breakdown
The real problem: nobody knows what workers are worth anymore
Nate opens with a blunt claim: in AI jobs, “nobody knows what you and I are worth anymore.” The usual 2026 advice — build a portfolio, learn the tools, ship projects — isn’t wrong, but it over-indexes on the one thing AI makes easiest: generating code.
Layoffs made the value question impossible to ignore
He ties the anxiety to the current market, rattling off cuts from Oracle (up to 30,000), Amazon (16,000), Block (4,000), Salesforce, and Dell (11,000), adding up to 60,000+ confirmed tech layoffs in Q1. His point is that companies are no longer just undoing pandemic overhiring — they’re actively recalculating how many humans plus AI they need to get the mission done.
Why the old proof-of-skill system broke
The heart of the argument is that production used to be hard, and that difficulty signaled effort and expertise. Now that something can look impressive at first glance with almost no effort, the old social machinery for hiring, promotion, and evaluating contribution starts to fall apart for everyone, from new grads to mid-career PMs.
Principle 1: optimize for comprehension, not just generation
His first principle is the big one: understand what you build at a deep level instead of just vibe-coding until something works. He warns that tools like Lovable intentionally feel like magic boxes, and says the industry is now producing software at unprecedented speed while comprehending it at unprecedented lows.
The AWS story and the case for slowing down
Nate gives a memorable failure case: an Amazon engineer using a mandated AI coding tool allegedly ended up deleting the entire production environment, leading to 13 hours of AWS downtime before it was labeled “user error.” His takeaway is sharp: if you move fast without understanding, you’re “going quickly without knowing how to steer the car,” and that usually ends in a wreck.
Principle 2: explanation should be part of the deliverable
If comprehension is internal, explanation is how you make it visible. Nate wants every project to ship with a simple, inseparable explanation of what it does, why certain choices were made, what could break, and what the builder learned — basically resurrecting the spirit of a good commit message for the generative era.
Principle 3 and 4: from credentials to microtransactions, and from private work to public proof
He says credentials are inflating away in value, so what matters more is evidence of real labor exchanged for real value — what he calls transactions. But because AI compresses meaningful work into much shorter cycles, he thinks we need “microtransactions for jobs,” plus more public, observable work so people outside your company can actually see your reps.
The Talent Board pitch: don’t let your work disappear into scattered artifacts
The final section is both framework and product pitch: Talent Board is his answer to expired links, dead chat logs, repos without context, and AI work spread across tools like Claude and Lovable. The core message lands cleanly: yes, increase output with AI, but the rare commodity in 2026 is proof that you can still think.