Durable Sessions AI UX vs Buried City Evidence: Which Framework?
// TL;DR
These two frameworks solve completely different problems and should never be compared as alternatives. Use Durable Sessions if you are building or fixing an AI chat product with streaming, disconnection, or multi-device issues. Use the Buried City Evidence Excavation Framework if you are documenting physical anomalies that contradict official historical records. There is zero overlap in audience, domain, or application. Pick whichever one matches your actual problem.
// HOW DO THEY COMPARE?
| Dimension | Christensen Durable Sessions AI UX Framework | Buried City Evidence Excavation Framework |
|---|---|---|
| Best For | Engineers building AI chat/agent products that need resilient, multi-device streaming | Independent researchers documenting physical evidence that contradicts official historical records |
| Domain | Software architecture, AI product UX, real-time infrastructure | Urban archaeology, investigative history, anomalous architecture documentation |
| Complexity | Moderate — requires understanding of streaming protocols (SSE, WebSockets), pub/sub, and agent architectures | Low-to-moderate technical complexity, but demands years of disciplined physical observation and documentation |
| Time to Apply | Days to weeks for an architecture audit and redesign | Months to decades — evidence accumulation is inherently long-term |
| Prerequisites | An existing AI product with a streaming architecture to audit | Physical access to anomalous sites, a primary observation contradicting an official record |
| Output Type | Architectural redesign: a Durable Sessions layer with resilient delivery, cross-surface continuity, and live control | A documented evidence archive with pattern stacks, cross-city corroboration, and a two-reading presentation |
| Creator Background | Mike Christensen (Ably) — real-time infrastructure and AI engineering | Based on the investigative methodology of Earl Whitcomb — decades of underground urban observation |
| Risk of Misapplication | Low — clearly scoped to AI product architecture problems | High — can drift into conspiracy thinking if the documentation discipline is not rigorously maintained |
| Collaboration Model | Engineering team effort with clear architectural deliverables | Lone investigator or small network of independent researchers |
| Verification Method | Testable: drop connections, open second tabs, send cancel signals — pass/fail criteria | Accumulative: pattern stacks and cross-city corroboration build weight over time, not binary pass/fail |
What does the Christensen Durable Sessions AI UX Framework do?
The Christensen Durable Sessions framework diagnoses why AI chat and agent products break under real-world conditions — network drops, multi-device usage, user-initiated stop buttons — and provides a concrete architectural fix. The core insight is that most AI products use direct HTTP streaming (typically SSE), which couples the response stream to a single client connection. When that connection drops, the stream is gone.
The framework introduces Durable Sessions: a persistent, shared layer between agents and clients. Agents write events to the session; clients subscribe to it. This decoupling unlocks three foundational capabilities: Resilient Delivery (streams survive disconnects), Continuity Across Surfaces (sessions follow users across tabs and devices), and Live Control (users can steer or cancel agents mid-generation). The framework includes a 10-step workflow to audit your current architecture, identify failure modes, and rebuild around Durable Sessions using pub/sub as the foundation.
What does the Buried City Evidence Excavation Framework do?
The Buried City Evidence Excavation Framework is an investigative methodology for documenting physical and documentary anomalies that contradict official historical records. It was built from the decades-long work of Earl Whitcomb, a utility worker who observed underground structures in Portland, Oregon, that did not match any official schematic or historical explanation.
The framework provides structured methods for measuring anomalies (the Wash Line Method), functionally analyzing underground spaces (the Built-for-Foot-Traffic Test), accumulating evidence across categories (the Pattern Stack), comparing findings across cities (the Cross-City Corroboration Test), and documenting institutional refusals to investigate. It emphasizes physical redundancy of documentation, non-sensationalized presentation, and a two-reading framework that presents both the official and evidence-based interpretations without demanding a conclusion.
How do they compare?
They do not compare in any meaningful way. These frameworks exist in entirely different domains and serve entirely different audiences.
The Durable Sessions framework is a software architecture methodology for engineers building AI products. It produces technical deliverables: architectural diagrams, protocol changes (SSE to WebSockets), and testable pass/fail criteria. You can validate it in an afternoon by dropping a connection and checking if the stream resumes.
The Buried City Evidence Excavation Framework is a historical investigation methodology for researchers documenting physical anomalies. It produces evidence archives, pattern catalogs, and cross-city comparison reports. Validation takes years and depends on accumulating independent corroboration.
The only structural similarity is that both frameworks are systematic — they provide step-by-step workflows with named principles and clear terminology. But this is true of any well-constructed methodology.
Which should you choose?
If you are building an AI product — a chatbot, a coding assistant, a customer support agent, a multi-agent research tool — and your users experience broken streams, lost responses, or an inability to control the agent mid-generation, use the Christensen Durable Sessions AI UX Framework. It is the right tool. The Buried City framework has nothing to offer you.
If you are investigating physical or documentary evidence that contradicts an official historical record — buried architecture, anomalous construction materials, sealed structures that appear on no schematic — use the Buried City Evidence Excavation Framework. It provides the documentation discipline and evidence-stacking methodology you need. The Durable Sessions framework is irrelevant to your work.
There is no scenario where these two frameworks are alternatives to each other. If you are seeing them compared, it is because they were surfaced together in a database or search result, not because they solve related problems.
Can either framework benefit from the other?
Theoretically, you could use real-time streaming infrastructure (the kind Durable Sessions addresses) to build a collaborative evidence-documentation platform where multiple researchers subscribe to a shared session of findings. But this is a stretch. The frameworks operate at fundamentally different layers of human activity: one is about software plumbing for AI products, the other is about investigative epistemology for historical anomalies.
Choose based on your actual problem. If your AI chat breaks when users switch Wi-Fi networks, you need Durable Sessions. If you found a vaulted brick chamber under your city that appears on no map, you need the Buried City framework. It is that simple.
// FREQUENTLY ASKED QUESTIONS
What is the Christensen Durable Sessions framework used for?
It is used to diagnose and fix AI chat and agent products whose streaming architecture breaks under real-world conditions — network disconnections, multi-device usage, and user-initiated stop or steering commands. It introduces a persistent session layer between agents and clients so streams survive connection drops and work across devices.
What is the Buried City Evidence Excavation Framework?
It is an investigative methodology for systematically documenting physical and documentary anomalies that contradict official historical records. It provides structured methods for measuring construction features, accumulating evidence into pattern stacks, comparing findings across cities, and presenting evidence without sensationalization.
Are Durable Sessions and the Buried City framework alternatives to each other?
No. They solve completely unrelated problems in different domains. Durable Sessions is a software architecture framework for AI product engineers. The Buried City framework is a historical investigation methodology for researchers documenting anomalous physical evidence. There is no scenario where you would choose between them.
Which framework should I use if my AI chatbot loses responses when users disconnect?
Use the Christensen Durable Sessions AI UX Framework. It directly addresses the Single-Connection Trap where stream health is coupled to a single client connection. It will guide you to introduce a persistent session layer so responses survive disconnections and clients can resume exactly where they left off.
Is the Buried City Evidence Excavation Framework a conspiracy theory methodology?
The framework itself emphasizes rigorous documentation discipline, measurement-based analysis, non-sensationalized presentation, and a two-reading framework that presents both official and alternative interpretations. However, it references the Tartarian Hypothesis and frames institutional silence as evidence, which places it outside mainstream academic archaeology.
What technical skills do I need for the Durable Sessions framework?
You need working knowledge of streaming protocols (SSE, WebSockets), pub/sub messaging patterns, and AI agent architectures. Familiarity with tools like the Vercel AI SDK or LangChain streaming is helpful but not required. The framework is aimed at software engineers and product architects building AI-powered products.
How long does it take to apply each framework?
The Durable Sessions framework can be applied in days to weeks — you audit your current architecture, identify failure modes, and redesign around a session layer. The Buried City framework is inherently long-term, often spanning months to decades, because its evidence accumulation and cross-city corroboration depend on sustained physical observation.
Can I use both frameworks together on the same project?
In practice, no. They address fundamentally different problems. One is software infrastructure for AI products; the other is investigative methodology for historical anomalies. The only conceivable overlap would be using real-time collaboration infrastructure to build a shared research documentation platform, but that is a significant stretch.