Topic
Essays, profiles, and reading notes on AI products, AI systems, adoption, agents, and company strategy.
Start Here
Start with the control layer for agents, context, tools, permissions, and human review.
Series Index AI Adoption That Actually Changes the CompanyUse the index instead of dropping into a middle chapter of the series.
Series Index Seven Powers in the AI EraRead the strategic power series from the table of contents.
AI The AI Builder’s Edge: 23 Patterns Across 264 ProfilesAcross 264 ai & machine learning profiles and 18,726 lessons, 23 patterns recur often enough to matter. They are not universal rules; recurrence indicates breadth within this collection, not agreement across an entire field.
Research Explainers When Do AI Agents Actually Need Blockchains?Jeremy Allaire argues that AI agents need economic infrastructure for payments, identity, and coordination. This explainer separates the functions that genuinely benefit from blockchains from those better handled by conventional systems, clarifying where onchain architecture earns its complexity.
Research Explainers AI Traffic Is Becoming Workflow TrafficOpenRouter's 100 trillion token usage study suggests AI demand is shifting from simple text generation toward reasoning, tools, code, and context-heavy workflows. This explainer maps what that change means for model providers, infrastructure, and the economics of serving AI.
Related Hubs
A post-training paper shows how interpretability tools can audit preference data, expose unwanted learning signals, and reshape rewards before models absorb bad habits. The method turns interpretability into a practical intervention in data curation and reward design.
Self-Harness shows that agents can use execution traces, targeted edits, and regression tests to improve the scaffolding around their own model. The important constraint is preserving task performance while the runtime changes itself.
A paper separates the model that improves an agent's harness from the model that must use it. Cheap models may propose useful scaffold changes, but task performance still depends on an executor capable enough to load and follow them.
A scaling paper argues that larger models win partly because frequent tasks stop overwriting rare-task features. The proposed mechanism links parameter scale to reduced interference and better retention of infrequent capabilities over time.
Leopold Aschenbrenner's Situational Awareness argues that AGI by 2027, superintelligence, trillion-dollar clusters, lab security, and government mobilization follow from the same trendlines. Reading it as a unified acceleration thesis makes its assumptions and policy implications easier to test.
An essay on Pluralis argues that decentralized AI needs scalable training, defensible monetization, and governance for intelligence outside the major labs. The piece tests whether distributed infrastructure can create a credible alternative to centralized model ownership.
Life-Harness argues that many agent failures come from the runtime interface, not the model, and shows frozen models improving across tasks when the harness changes. It shifts attention toward interfaces, tools, feedback, and execution design.
A security survey uses OpenClaw to argue that agent security is not a patch list but a product, architecture, ecosystem, and governance problem. The framework connects individual exploits to the permissions and incentives built into the surrounding system.