AI-powered economic analysis that follows the money
Every news article tells two stories: the one on the surface, and the one about where the money is actually going. We use AI to reveal the second story.
Our system analyzes news articles through an economic lens focused on how money moves (or stops moving) through economies, and who benefits when it pools instead of circulates.
We analyze articles about Supreme Court rulings, AI policy, climate deals, healthcare mergers, and military spending. That doesn't look like "economics" in the textbook sense. So why the name?
Because the Velocity Economics framework starts from a different premise: economics isn't a subject — it's the substrate everything else runs on. Every political decision, every technological shift, every social crisis either moves money through the economy or helps it pool in place. That movement — or lack of it — is the real story.
Currency is born (government spending), lives (circulating through the productive economy — wages, purchases, investment), and dies (taxation). That cycle isn't optional policy. It's thermodynamic maintenance. When the cycle runs, economies function. When it gets suppressed — when wealth pools instead of circulates — systems start breaking in ways that look like seventeen separate problems but are actually one disease.
A healthcare story is really about whether capital circulates through care delivery or gets extracted before it gets there. A tech story is about whether investment flows toward productive capacity or toward stock buybacks. A trade war is about which country's currency cycle gets disrupted. A climate deal either redirects capital flows or it doesn't.
That's why we run the same framework on everything. It's not that we think every article is secretly about GDP. It's that every article is about where money is moving, who's moving it, and who benefits when it stops. Once you see that pattern, you can't unsee it — regardless of whether the headline says "politics," "technology," or "foreign relations."
Fair question. Most people's experience with AI right now is watching tech companies use it to replace workers, harvest data, and concentrate profits. That's a legitimate reason to be skeptical. If that's all AI did, we'd be skeptical too.
But there's a difference between AI deployed as extraction and AI deployed against extraction. The same technology that corporations use to eliminate jobs and surveil populations can also give one person the analytical power that used to require an entire newsroom.
That's what's happening here. This site is built and operated by one person with a full-time day job. There is no staff. There is no newsroom. There is no institutional funding. Without AI, this operation simply wouldn't exist — not because AI replaced anyone, but because the work wasn't being done in the first place. No outlet was producing daily structural analysis across every domain through a consistent economic lens. AI made it possible for one person to do what no one was doing.
The AI doesn't decide what to analyze. A human selects every article based on editorial judgment about what reveals the pattern.
The AI doesn't decide what to publish. Every analysis goes through human review. Nothing appears on the public feed without approval.
The AI doesn't provide the framework. The Velocity Economics lens — the questions about where money flows, who benefits, what mechanisms enable concentration — that's the knowledge base the AI draws from. The framework existed before the tool. The AI applies it consistently; it didn't invent it.
The AI doesn't hide its work. Every analysis shows the source material, a neutral summary, and the opinionated analysis side by side. You can read the original, read our take, and decide for yourself. The transparency is the point.
The knee-jerk reaction to AI is understandable — and convenient. If people reject the technology entirely, they reject the tool that could give individuals the analytical power that used to require institutional backing. The people who own the institutions keep their advantage. Everyone else stays in the dark.
We use AI the way it should be used: as leverage for people who don't have a newsroom, a research department, or a billion-dollar platform. The framework does the thinking. The AI does the heavy lifting. The human makes the calls. And everything is transparent enough for you to verify.
A news article is submitted for analysis—either by pasting the text or providing a URL.
An AI call generates a factual summary of what the article claims—no opinion, just the key facts and assertions presented as clearly as possible. This uses Google's Gemma 3 27B model for fast, efficient summarization.
The article is classified into 1-3 topic categories (e.g. Trade Policy, Labor, Finance) so analyses can be filtered and browsed by subject. This also runs on Gemma 3 27B.
A RAG (Retrieval-Augmented Generation) system searches indexed reference material for relevant concepts, patterns, and historical examples to inform the analysis.
Anthropic's Claude Sonnet 4 analyzes the article using the retrieved context, identifying extraction mechanisms, power dynamics, and historical parallels. This is the opinionated "What's really going on" section—the most demanding call uses the most capable model.
Every analysis is reviewed for quality and accuracy before being published to the public feed. Only approved analyses appear on the site.
Our analysis system combines several AI technologies:
Each article goes through multiple AI calls with different models matched to the task. Summary, topic classification, and social sharing text run on Gemma 3 27B via Ollama—a fast, efficient open-source model ideal for structured extraction tasks. The main analysis runs on Claude Sonnet 4 via Anthropic's API—a more capable model suited to nuanced reasoning about economic patterns and power dynamics.
Reference material is indexed in a vector database. When an article is submitted, semantically similar passages are retrieved to provide context for the analysis. This ensures the AI applies a consistent analytical lens rather than improvising.
The analysis is thermodynamic, not ideological. It asks the same question of every story—whether the source is Fox News, MSNBC, Reuters, or anywhere else: "Where is wealth concentrating, and what mechanisms are enabling that concentration?"
This analysis system and its outputs are licensed under the PolyForm Noncommercial License 1.0.0.
You are free to use, share, and adapt the analyses for any non-commercial purpose. This includes personal use, research, education, and non-profit journalism.
Commercial use—including use by for-profit news organizations, financial services, or any revenue-generating activity—requires a separate license. Contact us for commercial licensing inquiries.