ORACLE DISSECTION: AI Tax/Job Program Proposal

URL SCAN: Greg Casar calls to tax AI companies to fund jobs program: 'AI is coming for your job'
FIRST LINE: AUSTIN (Nexstar) — U.S. Congressman Greg Casar, D-Texas, wants to start a 'New Deal-era' style job program.

THE DISSECTION

This article is a transitional narrative artifact — a snapshot of elite recognition that displacement is real, combined with reflexive deployment of 20th-century policy tools that are structurally incapable of addressing the problem. Casar deserves credit for naming the threat accurately. He deserves no credit for the proposed response.

THE CORE FALLACY

Tax-tokenize-and-subsidize is a solvent for friction, not for structural unemployment.

Casar's logic is: AI displaces → tax AI revenue → fund government jobs → unemployment absorbed. This assumes the displaced labor can be redirected into productive government employment at meaningful scale. The math collapses on contact with three DT constraints:

1. Token costs are exponentially collapsing. Traux — the sole voice in the article actually thinking about implementation — flags it directly: 60-80% cost reduction per quarter. The tax base he's proposing to fund a federal jobs program will be a rounding error within two years. This isn't a policy flaw. It's a category error. He's trying to fund a sustained structural transfer system with a depreciating asset.

2. Government jobs cannot absorb cognitive labor displacement. The New Deal worked because the labor being displaced (agricultural, manual) had direct equivalents in the work being funded (infrastructure, public works). AI doesn't just automate manufacturing. It automates judgment, analysis, synthesis, communication — the entire cognitive labor stack that the New Deal-era government needed to function. You cannot staff the Federal Reserve with Keynesian construction crews.

3. The circuit is broken at the source. Post-WWII capitalism survives on mass employment → wages → consumption → profit. AI severs the first link permanently. UBI, dividends, or government stipends might preserve aggregate consumption — but they don't restore productive participation. Casar is trying to prop up the consumption side without acknowledging that the wage side is structurally going away.

HIDDEN ASSUMPTIONS

• Assumption 1: AI displacement is a labor market friction problem, not a structural dissolution.
• Assumption 2: The same fiscal architecture that managed the Industrial Revolution can manage the Cognitive Revolution. (It cannot. The Industrial Revolution displaced physical labor while creating cognitive demand. The Cognitive Revolution has no downstream equivalent.)
• Assumption 3: AI companies will be profitable and token generation will remain economically central long enough to fund the program. (AI infrastructure costs are collapsing toward near-zero marginal cost. The "token" as a revenue unit has a finite shelf life.)
• Assumption 4: Political economy can be reformed through lobbying transparency. ("The reason government is doing nothing is because AI companies are spending hundreds of millions in lobbying.") This is true but irrelevant to the structural outcome. Lobbying resistance is a lag defense, not a survival mechanism.

SOCIAL FUNCTION

Classification: Prestige Signaling + Transition Management Theater

This is a politician doing the correct diagnostic work (acknowledging the threat) while reaching entirely into the wrong toolkit. It performs concern without threatening the existing order. It says "we see the problem" while delivering a solution that is (a) mechanically inadequate, (b) politically DOA, and (c) structurally incapable of addressing the underlying mechanism.

It is, in essence, hospice care presented as ICU intervention.

THE VERDICT

Casar has identified a real corpse. He is prescribing aspirin.

The Dickey counter — "computers didn't eliminate jobs" — is the most dangerous paragraph in the article, because it is the confident dismissal of someone who doesn't understand that this time the automation target is cognition itself, not just physical repetition. The historical analogy is valid for 1970-2020. It is architecturally invalid for 2025 onward.

Traux is the only person in the article thinking about the actual mechanics, and his observation about token cost collapse is the correct diagnosis. He just doesn't follow it to its conclusion.

Structural Judgment: The proposal will not pass, and if it did, it would not work. Not because of political resistance — because the tax base is structurally temporary and the displacement scale is structurally permanent. This is a policy response calibrated for a recession. The DT describes a regime change.

VIABILITY SCORECARD (Policy Proposal):

SURVIVAL PLAYBOOK (Individual Adaptation — No Soft Exit):

The policy is not your concern. Your concern is your position in the two-tier economy the DT describes:

• Sovereign path: Acquire equity stakes in AI infrastructure, not AI applications. Infrastructure has durable value regardless of which application layer wins.
• Servitor path: Identify the human verification, judgment, and accountability layers that remain legally necessary regardless of AI capability. These are temporary moats, not permanent careers.
• Lag plays: Physical craft, relationship-based services, regulated professions — all erode slowly, not fast. Use the lag.
• Do not: Rely on government jobs programs as a career plan. The DT timeline is faster than the federal procurement cycle.