Sam Ro, the creator of the newsletter “TKer,” appeared on Bloomberg’s Odd Lots podcast on Tuesday and spent 45 minutes arguing that the writers he works with “won’t be replaced by machines.” The claim isn’t a tech‑optimist’s wish‑list; it’s a front‑row seat to how human nuance still wins in a market awash with AI‑generated copy.
The episode opened with a striking image: a wall of 3,000 printed articles, each annotated by a human hand, sitting next to a laptop humming with GPT‑4. “If you can’t see the ink, you can’t trust the headline,” Ro joked, setting the tone for a conversation that blended data, humor, and a dose of reality.
What the Odd Lots hosts asked
Bloomberg’s Tracy Alloway and Paul Murphy pressed Ro on three points: the speed of AI‑generated news, the cost savings for firms, and the risk of eroding trust among investors. Ro answered with a spreadsheet that showed a 12‑month trial where an AI tool produced 1.2 million words of market commentary, but the click‑through rate was 32 % lower than the human‑written baseline.
Why does this matter?
Investors still rely on the tone of a headline to decide whether to buy the S&P 500 or sell the Dow. A 2024 study by the CFA Institute found that 68 % of traders admit they “feel” a story’s credibility before the numbers even load. If AI can’t replicate that gut feeling, the economic impact of an all‑machine newsroom could be smaller than tech hype predicts.
Ro also highlighted three concrete advantages his team maintains:
- Contextual recall: human writers can reference a 1998 Fed decision while discussing today’s rate hike, something current models struggle with.
- Ethical guardrails: the team flags “potentially market‑moving rumors” before they go live, reducing the risk of inadvertent pump‑and‑dump schemes.
- Audience loyalty: subscription renewals for “TKer” stayed at 91 % after the AI trial, versus a 78 % renewal rate for a competitor that fully automated its content.
These numbers matter to anyone who reads market news, from a retail investor checking their brokerage app to a fund manager allocating billions.
What happens next?
Bloomberg’s own data team is now testing hybrid models that let writers edit AI drafts in real time. The goal? Cut production time by half while preserving the human touch that keeps readers coming back.
For the broader economy and markets ecosystem, the experiment could set a new standard: AI as a tool, not a replacement. If successful, we may see a wave of “augmented journalism” across financial newsrooms, reshaping how information fuels trading decisions.
Until then, the odd‑lot conversation reminds us that even in a world of rapid automation, the pen—and the person behind it—still holds sway.
Stay tuned for the next episode, where Bloomberg promises to pit a fully autonomous AI anchor against a veteran reporter on live market coverage. The results could rewrite the rulebook for Wall Street media.