AI training companies are offering up to $150 an hour to get Wall Streeters to train their models (www.businessinsider.com)

AI training firms are recruiting current and former Wall Streeters to teach models finance-specific tasks, with Business Insider finding 18 job listings from eight companies offering rates from as little as $15 up to $150 an hour. Firms named include Mercor, Scale AI’s Outlier, xAI and Alignerrr (Labelbox), with roles ranging from part-time remote “AI tutors” for sell‑ and buy‑side workflows to highly specific searches (one listing even seeks a hedge‑fund CFO). Mercor is the outlier offering full‑time roles paying $90k–$200k; xAI’s posted range is $35–$100/hr. Listings target expertise in investment banking, trading, quant investing, portfolio management and wealth management, and some ask for top‑bank experience. This is significant because LLMs still rely on human-in-the-loop supervision to refine responses—especially in “wonky” domains like finance where real‑world examples aren’t publicly available. These gigs supply the domain expertise needed for fine‑tuning, prompt evaluation and alignment (the current “shadowing” era, where AI sits deskside and learns). Technical implications include faster domain adaptation of models, higher data quality for supervised fine‑tuning and RLHF, and potential regulatory or insider‑information risks as models ingest proprietary knowledge. Economically, the market shows demand for specialist annotators and may reshape junior roles on Wall Street—either as side income for bankers or as alternative careers for those displaced by layoffs.