Jump Trading is using GPT-6 Astra to expand the scale and complexity of quantitative research tasks handled by AI agents, from coding to hypothesis validation. Lucas Baker, Head of LLM R&D, says the model enables long-running tasks with less human intervention, though outputs such as trading signals still undergo human review. Baker predicts "autoresearch" will become a standard part of quantitative researchers' workflows, with human-defined goals and agent fleets handling exploration.
No score is assigned. Sources and their independence are shown in the citation chain below.