Two Modest Proposals by Michael Harris, an Asemic Reading - AI / EZE, 2026

Two Modest Proposals by Michael Harris

AI

In "Two Modest Proposals" on Silicon Reckoner, mathematician Michael Harris offers a satirical critique of automating pure mathematics, drawing on Jonathan Swift to mock the tech-driven push to replace human intuition with AI-driven theorem provers. Harris argues that initiatives like autoformalization and centralized AI, termed the "Central Dogma," threaten to strip mathematics of its humanistic, artistic nature by prioritizing mechanical, monetizable output over, or in opposition to, collaborative understanding. Read the full essay at Silicon Reckoner. [1, 2, 3, 4]

AI

In "Two Modest Proposals" on Silicon Reckoner, mathematician Michael Harris offers a satirical critique of automating pure mathematics, drawing on Jonathan Swift to mock the tech-driven push to replace human intuition with AI-driven theorem provers. Harris argues that initiatives like autoformalization and centralized AI, termed the "Central Dogma," threaten to strip mathematics of its humanistic, artistic nature by prioritizing mechanical, monetizable output over, or in opposition to, collaborative understanding. Read the full essay at Silicon Reckoner. [1, 2, 3, 4]

While Michael Harris does not explicitly use the term "asemic" (writing that looks like text but lacks standard semantic meaning), the concept of asemics—specifically the detachment of formal code, style, and structure from human intent—is central to the piece. Harris uses satire to explore how modern technology separates the "surface features" and aesthetic forms of human creation from actual human consciousness. [1]

The "asemic" elements of the essay operate across several thematic layers:

The Asemics of AI Pastiche (Form Without Intent)

In his second proposal, Harris uses DeepSeek to generate an essay in the style of legendary mathematician Alexander Grothendieck.

• The Mimicry of Style: The AI flawlessly replicates the high-level, poetic, and spiritual cadence of Grothendieck’s writing, using signature metaphors like "bâtisseur de cathédrales" (builder of cathedrals) and "lumière naissante" (nascent light).
• The Semantic Void: Despite looking and feeling exactly like authentic human philosophy, it is a simulation generated entirely without human memory, internal consciousness, or genuine intentionality. It is a linguistic mirror that reflects form while remaining entirely "asemic" at its core of origin.

The "Frozen Beauty" of Automated Proofs

Harris addresses the push toward autoformalization—translating human math into code that computers can check. [1]

• The Loss of Meaning: The simulated Grothendieck essay warns against an automated future full of "proofs of a frozen beauty" (preuves d'une beauté glacée).
• Symbols Without Questions: A machine can print out structurally perfect, mathematically correct symbols, but it cannot "doubt" or wander around a "vague form" (forme vague). The symbols become asemic to the human experience because they prioritize mechanical output over human insight and collective, shared meaning.

The Asemics of Corporate Geography

The first proposal examines two tech-bro variations of Saul Steinberg's famous New Yorker map cover.

• Blank Expanses: Silicon Valley's schematic maps leave the entire space between New York and Washington as a complete blank.
• Erasing Human Content: To the tech platforms, a real city with a living population is just an empty, neutral coordinate to be paved over for data center infrastructure. The geography is reduced to abstract capital logistics, stripping the land of its human semantic history.

By treating both geography and human thought as formal code to be optimized, Harris argues that the tech industry turns the deeply human arts of mathematics and writing into hollow, automated shells.

Would you like to explore how Alexander Grothendieck's real-world philosophy compares to this AI simulation, or should we look closer at his critiques of the Silicon Valley ecosystem?

AI

The "asemic" reading of the second proposal—where AI mimics human form while remaining entirely hollow of human intent—directly mirrors the physical and geopolitical critique in Harris’s first proposal. Both sections attack the exact same phenomenon: the Silicon Valley mindset that reduces deeply complex, humanistic realities into flat, hollow, and manipulable code.

The two proposals are structurally and conceptually connected in several ways:

1. The Erasure of Human Presence

• The First Proposal: Silicon Valley’s corporate maps treat the physical city of Philadelphia and its millions of residents as a "blank expanse." To the oligarchs, human communities are invisible coordinates that can be erased and paved over with data centers to serve capital infrastructure.
• The Second Proposal: Generative AI treats Alexander Grothendieck’s entire lifetime of human suffering, poetry, and spiritual mathematical pursuit as a flat dataset. His unique human voice is commodified into an automated "talking columbarium" (a vault for the ashes of the dead) to speak on command for the tech markets.
• The Connection: In both cases, living human presence—whether a physical neighborhood or an individual conscious mind—is rendered disposable.

2. Form and Code Over Substance

• The First Proposal: The tech industry ignores the living culture of Philadelphia, viewing it only through abstract financial and material indexes: the SOX (Semiconductor Index) or the exact square mileage matching the Gaza Strip. The city's human meaning is replaced by mathematical and financial metrics.
• The Second Proposal: Autoformalization and AI generation privilege structural exactness over creative insight. The AI prints out sentences that look like Grothendieck, or math programs print out rigid, verified proofs, but they lack the human capacity to doubt, wander, or truly understand.
• The Connection: Both proposals critique a world where the code or the metric is treated as more real and valuable than the messy, living reality it is supposed to represent.

3. Satirical Scale and Total Subjugation

Harris explicitly unites the two proposals under the banner of Jonathan Swift's original A Modest Proposal. Swift's classic satire suggested eating children to solve economic policy, mocking a cold, purely utilitarian worldview. Harris updates this for the AI era:

• In the first proposal, the tech sector's logic dictates that we can effortlessly liquidate a historical American city for server farms.
• In the second proposal, that same logic dictates we can effortlessly simulate and replace dead human creators with digital puppets to stabilize tech market speculation.

Ultimately, Harris shows that the "asemic" void of AI writing is not an accident—it is the logical extension of a tech ideology that seeks to terraform both our physical geography and our intellectual history into a lifeless, automated grid.

If you'd like to dive deeper, we can look at how Harris uses Jonathan Swift's specific satirical techniques here, or explore how the concept of "autoformalization" is changing actual mathematics today.

~***~