ChatGPT's in Your Bank, Google Spams the GEO Industry, Brockman Takes OpenAI Product

The recent AI radio experiment by Andin Labs highlighted how current frontier models still need human direction to become viable businesses. Claude, Gemini, ChatGPT and Grok each ran a station, but only Gemini managed a modest $45 sponsorship, while the others generated losses or hallucinated deals. This outcome underscores the gap between impressive language capabilities and real‑world revenue generation, prompting investors to watch product‑level integration as a key metric for future AI profitability.

Meanwhile, a growing political backlash is emerging around AI data center construction. In Pennsylvania, a virtual town hall drew 225 participants, and a QuinnPack poll shows 68% of residents oppose the 60 proposed facilities. Similar resistance is surfacing in Nevada and California, where utilities face strain from subsidized data‑center power loads. The NIMBY sentiment threatens the rollout of compute infrastructure essential for competing with China, making state‑level policy a strategic risk factor for AI investors and developers.

Google’s latest policy classifies any attempt to game generative‑AI answers as spam, threatening demotion or de‑indexing. This move directly challenges the burgeoning Generative Engine Optimization (GEO) industry that sells citation shortcuts for AI search. At the same time, OpenAI is expanding its product suite: a $200‑per‑month ChatGPT finance assistant now reads transactions from over 12,000 institutions via Plaid, though its read‑only scope limits immediate monetary impact. The rollout coincides with Greg Brockman assuming control of all OpenAI product teams, consolidating ChatGPT, Codex and developer tools under a single leadership, signaling a tighter product roadmap as the company eyes an IPO.