6 open models, 80 M&A memos, double-blind AI judges: Step 5 Preview claims a new Pareto frontier
Aizkmusic · x · 2026-10-16
Alongside the Step 5 Preview launch, the team ran an experiment: six open models wrote the same M&A memo for 80 companies, with two other AI judges double-blind picking the best memo per dollar, aiming to establish a new Pareto frontier for agentic finance. The grind behind it: rebuilding and rebenchmarking the demo, 1,600 AI verdicts run overnight, and three straight days at AI Engineer NYC — capped by the CEO posting a chart with a single yellow dot marking frontier capabilities.
More from Venture
- CNN signs multi-year AI licensing partnership with Google for generative AI experiments — gaganghotra_ · 2026-10-16
- Zymtrace raises $12.2M and hires ML engineers to build agents that fix GPU bottlenecks — cfregly · 2026-10-16
- FT: Risk of an AI crunch for stock markets is rising — nordicinst · 2026-10-16
- Haight Street Compute Market: a one-day trading floor for GPU deals in SF — ctjlewis · 2026-10-16
- OpenAI's 110 London staff averaged £2.3M each in 2025, double the prior year — rohanpaul_ai · 2026-10-16
- Persona Raises $10M Led by Vine Ventures; $179 Wearable Band Expected in December — heyneighbor · 2026-10-16