The Wake: May 9, 2026

The Wake is a daily briefing from George's saved internet. The issue is written as a newsletter first. The tweets are the source material, preserved below for receipts.

Source window: May 8, 2026. Signals: 3 bookmarks and 2 likes.

Brief

Three small signals yesterday point to the same, larger political economy: scarcity is nearly always institutional, not natural, and whose rules determine scarcity is becoming the central fight across housing, infrastructure, tech, and cultural memory.

A sharp thread on housing mistakes visible platforms for the underlying problem. Two engineering notes: one about ops automation, one about a new model that runs cheaper and faster: show routine work being automated away and platform providers consolidating leverage. A reminder from antiquity shows that survival of what matters is contingent on copying, choices, and institutions. Read together, they outline the same pattern: incentives set the supply, visibility sets the blame, and preservation sets the long-run winners.

Housing: blaming the thermometer

Progressive politicians are increasingly targeting platforms like short-term rental marketplaces as if they were the root cause of housing shortages. That is politically effective; platforms are visible, private, and framed as extractive. But the economics look different.

Regulatory constraints on supply: zoning that freezes density, permitting regimes that take years, rent controls that reduce turnover, and cumulative construction standards: are the mechanisms that make housing scarce. When supply is constrained, prices rise. Platforms can change who gets to use a scarce unit at any given moment and can exacerbate pressure in certain neighborhoods, but they are not the structural reason new units are not being built. The pattern is clear in comparative urban governance: cities that restrict building become expensive, while those with looser rules can accommodate population growth more cheaply.

That distinction matters for policy. Targeting platforms treats a symptom and leaves the core incentives untouched. To actually lower prices and increase access, the levers are supply side: zoning reform, streamlined permitting, flexible building standards, and targeted social housing. Political appetite for those fixes is uneven because they redistribute value away from incumbent property owners and local NIMBY coalitions. Expect the rhetoric to keep aiming at the visible villain while the harder policy fights over land use remain the decisive battleground.

(Credit: @brivael for the crisp inversion-of-causality critique.)

Automation at the edges: cron jobs, claws, and GPT

On the ops side, there is less theater and more quiet displacement. Two separate notes: a developer tweeting about delegating cron jobs and system components “talking to each other” is not trivia. It marks the incremental automation of maintenance and the reduction of cognitive load for engineering teams. Small improvements in orchestration multiply across systems.

Now add the leapward in foundational models. If a widely used model becomes materially more efficient and cheaper to run for routine tasks, teams stop reaching for more expensive tools and rearchitect less. That changes hiring, tooling, and margins. One developer’s report that GPT-5.5 is “very good, very efficient” in a low-reasoning mode signals more than quality: it means cheaper automation for many text and coordination tasks that previously required people or bespoke code.

Put together, improved orchestration and cheaper AI reduce the labor hours needed for routine monitoring, on-call work, customer triage, and other repeatable tasks. That is good for operational cost and responsiveness. It is also consolidating. Centralized model providers and platform operators become the locus of capability: if you depend on a particular model for everyday work, you depend on its vendor for performance, pricing, and policy.

Policy and strategy implications are straightforward. Organizations should plan for fewer “keep-the-lights-on” roles and more focus on edge cases, resilience, and governance of third-party models. Regulators should watch market concentration in tooling and inferential services, not just consumer-facing platforms.

(Credit: @steipete, @dhh.)

What survives and why: cultural memory as a warning

The observation about the Epic Cycle: that only the Iliad and Odyssey survive nearly intact while other once-ubiquitous epics are gone: is more than a classical aside. It is a parable about contingency: what endures is what people copy, fund, and value in institutions that transmit knowledge.

Translate that to modern policy debates and technology. Cities choose to “preserve” certain neighborhoods by blocking new construction; that preservation benefits current residents and owners and locks in scarcity. Tech stacks and platforms determine which datasets, models, and interfaces are maintained and monetized; anything outside the profitable path risks disappearing. Digital archives and the humanities regularly lose materials when institutions fail to prioritize redundancy and access.

If we want cultural and material ecosystems that serve broader publics rather than incumbents, we have to intervene where copying and funding decisions happen. That means supporting open archives, durable public goods, and policies that bias towards maintenance and access rather than exclusive capture.

(Credit: @Dr_TheHistories for the reminder about contingent survival.)

Converging dynamics and the political map

These three domains share mechanics. Scarcity is produced by institutions. Visibility determines political target selection. Technology changes who holds leverage, often accelerating centralization. That produces a predictable political map.

• Incumbents with entrenched property, market share, or archival control push policies that conserve their advantage. They prefer regulations and norms that make supply inelastic.
• Outsiders and newcomers point to visible, monetizable actors: platforms, private model vendors: because they are easier to attack and sometimes deserve scrutiny.
• The durable fixes are harder: changing zoning, rethinking public infrastructure funding, investing in open preservation, or regulating model markets. Those require sustained coalition-building and technical policy work.

For strategists and investors the takeaway is practical. Short-term regulatory fights against platforms will continue to be high drama and can yield tactical wins. Long-run change will come from the less-visible work: rewriting land-use laws, building alternative infrastructure for model provisioning, and funding durable archives and public digital goods.

What to watch

• Municipal zoning proposals and state-level preemption battles. These determine whether cities expand supply or entrench scarcity.
• Legislative and regulatory actions targeting short-term rental rules versus broader housing policy. Watch which wins the media narrative.
• OpenAI and major model vendors for public releases and pricing changes around the new, more efficient models; lower-cost models reshape enterprise cost structures.
• Devops and orchestration projects that offload routine maintenance (examples: cron delegation, automated remediation). Small tooling wins compound.
• Funding and initiatives in digital preservation and public archives. The Epic Cycle lesson is a policy prompt, not nostalgia.
• Political messaging from progressive leaders that frames platforms as the problem. That rhetorical strategy will shape voter salience even if it does not fix supply constraints.

If you want a single thesis to carry forward: blame is easy, fixes are structural. If you want cheaper housing, more resilient infrastructure, or a broader cultural record, you have to change the rules that determine what gets built, copied, and preserved.

Generated from Birdclaw bookmarks and likes. Edited by Ody before publication.