AI Is Killing 16,000 Jobs a Month — Here's How to Survive

The numbers are no longer theoretical.

Goldman Sachs just dropped a bombshell in their April 2026 U.S. Daily note: AI is erasing roughly 16,000 net jobs per month in the United States alone. That's not a projection for 2030. That's happening right now, in the last twelve months.

The breakdown is brutal:

• 25,000 jobs destroyed per month through AI substitution — machines doing what humans used to do
• 9,000 jobs created per month through AI augmentation — humans working alongside machines
• Net loss: 16,000 jobs every single month

And here's the part nobody wants to say out loud: Gen Z is taking the hit disproportionately. Entry-level roles — the ones that used to be the on-ramp to a career — are exactly the tasks AI handles best. Resume screening, data entry, basic copywriting, first-line customer support, junior code review. Gone.

The inconvenient truth about "AI will create jobs"

Yes, AI creates jobs. The data says so. But there's a trap in that sentence that most commentators miss: the jobs AI creates and the jobs AI destroys are not the same jobs, and they don't go to the same people.

The 25-year-old whose data entry position was automated doesn't magically become the prompt engineer or AI ethicist that fills the augmentation column. The skills don't transfer. The networks don't transfer. The geographic locations don't transfer.

This is what economists call a structural displacement — and it's the cruelest kind, because the aggregate numbers look fine ("jobs were created!") while individual lives are upended.

The survivor profile: what the data shows

I've been tracking AI career displacement for over a year. The people who survive — and thrive — in this environment share a specific pattern. It's not about learning to code. It's not about becoming an AI whisperer. It's something more fundamental.

The survivors own the system, not the task.

Here's what I mean:

• The person who entered data got replaced. The person who designed the data pipeline did not.
• The person who wrote articles got squeezed. The person who built the content engine did not.
• The person who managed ads got automated. The person who architected the growth stack did not.

The pattern: task executors are replaceable. System architects are not.

This isn't about being technical. It's about positioning yourself at the layer of abstraction above the task. The moment you can describe your job as "I do X with tool Y," you're a target. The moment you can describe it as "I designed the process that produces X," you're resilient.

The 3-step repositioning protocol

If you're reading this and feeling the pressure, here's the actionable framework:

1. Audit your task dependency

Write down everything you did last week. For each item, ask: "Could a reasonably capable AI tool do 80% of this within 18 months?" Be honest. If the answer is yes for more than half your list, you're in the danger zone.

2. Move up the abstraction ladder

For every task that's automatable, ask: "Who decides WHEN this task runs? Who defines WHAT good output looks like? Who connects this output to the NEXT step in the value chain?"

Those "who" questions reveal the architectural layer — and that's where you need to be standing.

3. Build your decoupled stack

Start building systems that you own. A newsletter you control. A data pipeline you designed. A client workflow you architected. These are assets — they compound over time, and they're not sitting inside someone else's platform waiting to be automated away.

The Intelligence Way philosophy has always been about decoupling from dependency. This is the ultimate expression of that principle: your career cannot be decoupled from you if you are the architect, not the operator.

The uncomfortable question

Here's what keeps me up at night: the Goldman data shows this accelerating. The 16,000/month net loss was 12,000 six months ago. The substitution curve is steepening while the augmentation curve is linear.

We are not in a steady state. We are in the early phase of an exponential.

If you're not actively repositioning right now, you're not falling behind slowly — you're falling behind exponentially.

The builders who read this site already understand this. The question is whether understanding converts to action fast enough.

The clock isn't ticking. It's already past midnight for some.

---

Data source: Goldman Sachs U.S. Daily Note, April 2026, authored by economist Elsie Peng. The research separates AI substitution (job destruction) from AI augmentation (job creation) for the first time at monthly granularity.