The Fed delivered the kind of move that keeps both mandates formally intact and practically at war. A 25 bp cut with three dissents is not a pivot; it’s a reluctant concession to mounting growth risk, wrapped tightly in language that still treats inflation as the career-risk variable and employment as the convenient shock absorber.

The statement was marginally less hawkish. But the real message was clear:

“Modest easing is fine. Wholesale regime change is not.”

At least not yet.

More important than the cut was how Powell delivered it. His repeated comfort with inflation breakevens was code for: we can be flexible without losing credibility. And then came the kicker: the Fed’s commitment to provide $40B per month in reserve-management operations (RMOPS) — QE-light by another name — including outright purchases of short-term T-bills.

This is liquidity. Real liquidity.

And it marks the start of the next quiet expansion of the Fed’s balance-sheet footprint.

Market commentators framed this as the opening round in a battle over the Fed’s structure and purpose. And that’s the right lens. Because beneath the monetary tweaks lies a much bigger structural crack:

The Fed is running a dual mandate for an economy whose definition of “employment” is evaporating.

AI, robotics, and embodied intelligence are rewriting what work is. And if work becomes a task rather than a job — a process rather than an identity — then the dual mandate becomes unstable fiction. You can’t optimize for “maximum employment” when the job itself is dissolving under you.

This is the fault line that runs through the future of labor, capital, and the Fed’s entire framework.

Let’s walk straight into it.

When the Plumber Disappears From GDP

Picture the classic weekend failure mode (credit to Jordi Visser):

Your garbage disposal dies.

You call the plumber.

A van shows up.

One hour is billed.

GDP ticks up.

In the old model, that’s “work.”

Now picture the new model:

You point your phone at the sink.

An AI agent walks you through the fix.

Ten minutes later, you’re done.

No invoice.

No van.

No billable hour.

GDP falls.

Your actual outcome — time saved, no cost, problem solved — is strictly better. But national accounts treat it as a loss of economic activity.

This is the uncomfortable truth behind all the breathless productivity forecasts. Yes, generative AI could add trillions to global output. Yes, annual productivity could rise a few percentage points.

But those projections still assume the “freed-up time” gets reallocated toward other paid work.

The Jensen Huang / Elon Musk version of the future is weirder:

A world where tens of billions of micro-tasks evaporate into software, and the invoice never gets written at all.

“Work” dissolves.

GDP undercounts.

The standard of living rises anyway.

Tasks vs Jobs: Where the Real Risk Lives

AI does not replace jobs.

It replaces tasks inside jobs.

If your economic value is tightly coupled to a single, well-defined activity — transcription, formatting, basic triage — then the job is the task. And the task disappears the moment an AI agent can do it at scale.

That’s exactly what the plumber example shows: narrow tasks are fragile.

But jobs made of loosely structured tasks — coordination, judgment, communication, relationships — get reshaped, not erased. Early firm-level evidence supports this:

• Significant task replacement
• Very small net employment cuts — so far
• Jobs “bent but not broken,” with humans becoming expensive safety nets for automated workflows

As the task-share of a role rises, the job title survives — but the human inside it feels more like the override button for a machine.

This is radiology’s story in miniature.

Radiology: A Warning and a Template

A decade ago, everyone predicted AI would wipe out radiologists.

Instead:

• AI ate narrow tasks — nodule detection, bone-age estimation, metastasis screening
• Throughput increased
• Finding rates improved
• Report complexity expanded
• Demand for radiologists rose

The system generated more scans.

More documentation.

More expectations of precision.

AI didn’t eliminate the job.

It made the job bigger — and more leveraged.

This is the blueprint for high-skill domains:

AI eats tasks → the workflow expands → humans move upstream → demand increases for those who can supervise, integrate, and sign off.

But notice the catch:

The barrier to entry skyrockets.

This Week’s Deep Dive

In the premium section:

• What Powell’s QE-light tells us about Q1 liquidity conditions
• The sectors that gain when tasks disappear but jobs remain
• The AI-infrastructure choke points investors should be watching
• Why GDP will “fall” even as living standards rise — and how to position for it
• The 2030s macro regime: capital-dominant, labor-thin, energy-heavy
• Portfolio implications for stocks, bonds, energy, AI, and alternatives

If you want the full framework — not just the story — upgrade and continue.