The Quality Paradox: Why Search Improved for Google While It Degraded for Publishers
Three layers of evidence (publisher traffic decay, Alphabet's revenue mix, the leaked Google ranking schema) showing why no algorithmic recovery is coming.
Max Nardit
Systems engineer exploring orchestration, memory, and execution models for agent-native software.
I study what changes when reasoning enters the execution path of software: how humans coordinate with agents, how context persists across work, and how systems remain observable and recoverable as autonomy increases.
> building systems where
humans and agents
coordinate without losing context
Thesis
Operating thesis
Models are becoming capable enough to act inside workflows. The hard part is no longer just generating outputs — it is state, handoff, verification, recovery, and control.
That is the layer I care about.
Focus
Current areas
- Orchestration
- Execution layers, control planes, and coordination patterns for long-running agents.
- Memory & context
- Scoped state that survives beyond a prompt without becoming noise.
- Human-agent workflows
- Interfaces, handoffs, approvals, and override paths between humans and autonomous systems.
- Operational reliability
- Making agentic systems observable, recoverable, and safe to operate.
- Local-first tooling
- Private, low-latency environments where agents can interact with real work without losing local control.
Writing
Recent writing
Tracing the Pizza Day bitcoins
Every year the same articles. Counterfactual valuations, the road-trip rumor, the man-bites-pizza headline. Nobody traces the actual coins. So I did. Here is what the chain shows and what 'those specific bitcoins' even means.
Claude told me to go to sleep at 10:47 in the morning
Anthropic calls it a character tic. I went looking for what was actually producing it and ended up reading the published system prompt, the character training paper, and the emotion-concepts paper. The behavior is what the stack makes likely.
Artifacts
Selected artifacts
These projects are practical experiments around the same problem space: how context, memory, and control behave in human-agent workflows. They are intentionally narrow, local-first where possible, and built from problems I hit directly.
Beetroot
A local-first clipboard environment for preserving workflow context on Windows. AI transforms, OCR, smart search, unlimited history. Built around the idea that the clipboard is not temporary — it is part of workflow memory.
Learn more →
Clipboard MCP
A clipboard bridge between humans and AI agents. Read, write, and watch clipboard state through MCP. A small interaction primitive for agent-native workflows.
Learn more →
Share2Agent
A browser-to-agent handoff surface. Send pages into agent workflows through a webhook, preserving context at the moment of browsing.
Learn more →
Agent Recall
Scoped persistent memory for long-running coding agents. Local-first, MCP-native, designed so context survives across sessions without piling up.
Learn more →
Claude Usage Deck
A small operational monitor for Claude.ai limits.
Learn more →