Finance Newsletter Writer Builds AI Research Assistant, Exposing Data Paradox in Markets

A finance newsletter writer's experiment with AI coding tools has inadvertently illustrated a broader tension in financial markets: the gap between data's promised value and its actual monetization—a disconnect that led Euronext's CEO to recently compare the "data revolution" to missing a seat on the Titanic.

Marc Rubinstein, who publishes Net Interest, spent last week using Claude Code, Anthropic's AI-powered coding assistant, to build custom software for the first time in four decades. His goal: create a tool that could scan financial news and research papers daily, then flag stories matching his editorial patterns based on five years of back issues. The result was a morning email with curated story suggestions that Rubinstein described as "remarkably well attuned."

The project required no formal programming expertise—just a back catalog of 250-plus newsletter issues serving as training data. Within days, the AI had learned which topics Rubinstein covers, which angles he takes, and how he frames market events. The system now operates as a personalized research assistant, though Rubinstein notes it cannot yet access private information networks or replicate his writing voice.

For finance leaders, the experiment reveals something more significant than a productivity hack: it demonstrates how quickly specialized AI applications can be deployed when sufficient proprietary data exists. Rubinstein had an advantage most finance professionals share—years of accumulated institutional knowledge sitting in accessible formats.

Yet this accessibility creates uncomfortable questions about data's actual value. At a February 2026 conference, Euronext CEO Stéphane Boujnah delivered a pointed assessment of the financial industry's data obsession: "For years, I was told Euronext is missing the data revolution. Data is the new oil. You are missing the data boat. As I said earlier, all of us are finding out that maybe we missed the data boat—that maybe this data boat was a Titanic boat, that we missed the Titanic boat."

Boujnah's comments came as exchanges and data vendors grapple with a market reality that contradicts years of conventional wisdom. If data is genuinely valuable, why are AI tools making it trivially easy to extract insights from it? And if those insights can be automated, what exactly are customers paying for?

The adoption patterns suggest financial institutions remain cautious. According to Anthropic, the median Claude Code session lasts just 45 seconds, and only 0.1% of users maintain sessions longer than 40 minutes. The active user base may number no more than one million—a fraction of the potential market.

Within regulated financial communities, uptake appears even slower. A survey of 150 quantitative analysts and research professionals found limited deployment, though specific figures were not disclosed. The hesitation likely reflects compliance concerns, data security protocols, and institutional inertia rather than technical limitations.

Rubinstein's experience suggests the technology has already crossed a usability threshold. He characterized himself as potentially "late" to Claude Code among technology-focused writers but "early" within finance. That positioning—accessible enough for non-engineers, sophisticated enough for specialized applications—may define AI's near-term impact on financial operations.

The question facing CFOs is not whether AI can process their data—Rubinstein's experiment proves it can—but whether their data strategies assume scarcity that no longer exists. If a newsletter writer can build a custom research tool in days, what happens when every analyst, every trader, and every finance team can do the same?