The new Eval Agent in Terminal X grades AI output against a 54‑point rubric and catches 25% more mandate deviations.
NoteBy OossaPublished by Oossa: 1 min read
OpenAI released Terminal X on 2026-10-20. The platform’s Eval Agent uses GPT‑6.1 to score outputs against a 54‑criterion rubric. In tests it matched senior investor judgments 88.6% of the time and identified 25% more breaches of a fund’s investment mandate.
Asset managers can rely on Terminal X to spot more rule‑breaks in AI‑generated analysis, potentially reducing costly investment errors.
Business
AMD's FSR 4.1 upscaling now runs on Radeon RX 6000 cards, Xbox Series X|S, and select Ryzen AI chips, with a modest performance hit.
Business
Three idle reactors are back online, adding 2.25 GW of electricity for AI workloads, according to a report on Oct 20 2026.
Business
ByteDance, Alibaba and Tencent placed orders for over 3 million RTX Pro 5500 workstation cards, but Nvidia says it cannot meet the demand.
Oossa · Newsletter
The week in AI, explained
Every Monday: the stories worth knowing, in plain language. Free, no spam.
