Ken Griffin, the billionaire founder of Citadel, told Goldman Sachs that most corporate productivity gains attributed to AI are actually from machine learning and digitization, not true agentic AI. His own team built an agentic system that reproduced and verified academic finance papers in two to three hours versus the six to eight weeks normally required by expert researchers, a distinction that matters because it reframes how transformative AI will actually be to the economy.
Summaries like this, in your inbox every morning.
Sign up free →What happened
At Goldman Sachs's Apex Symposium, Citadel founder Ken Griffin distinguished between what he calls true AI—agentic systems that can plan and execute multistep tasks—and machine learning or optimization tools that companies often mislabel as AI. He cited a Citadel team project that built an agentic system to reproduce and verify academic finance papers, completing work in two to three hours that normally takes expert researchers six to eight weeks.
Why it matters
Griffin argues that most corporate claims about AI-driven productivity gains actually come from machine learning, optimization, or digitization—not the kind of general reasoning systems that triggered his earlier concerns. This distinction shapes how we should think about AI's economic impact: the real transformation may be simpler and narrower than the current hype suggests. For companies assessing their own AI investments, the gap between true agentic reasoning and familiar machine learning tools carries implications for where real competitive advantage lies.
What to watch
Griffin has stated that Citadel has not cut a single job over these efficiency gains and frames them as expanding what existing staff can tackle. He predicted that competitive moats between companies are "being filled in at lightning speed," potentially enabling small teams with agentic systems to challenge larger incumbents.
Griffin's framing reflects a recurring theme in his recent public remarks: a skepticism toward broad narratives and a preference for concrete, measurable evidence. At a dinner about two years ago with multiple CEOs, he initially dismissed their claims that AI was transforming their businesses, then asked them to share specific stories. When they provided "four or five incredible stories" about productivity gains, he determined that none actually involved AI—they described machine learning, optimization, or digitization instead. This distinction matters because the terms are used interchangeably in corporate America, creating confusion about what is actually driving economic change.
The agentic system his team built to verify academic finance papers represents, in his telling, a genuinely different kind of AI—one that moved him from dismissing the technology as "garbage" at Davos to genuine concern. The system's ability to autonomously execute a complex, multistep research process mirrors the kind of general reasoning that separates true AI from narrower machine learning tools. Griffin's broader economic interpretation is that while a technological revolution is underway, AI is only a component. Workers with narrow, hard-to-redeploy skills (he cited translation) face real risk and require retraining infrastructure. At the same time, he believes competitive moats are "being filled in at lightning speed," potentially creating what he calls a "golden age of entrepreneurial activity" where small teams with agentic tools can compete with incumbents. Yet Citadel itself has not downsized; it has instead retasked the gains toward attacking a "huge swath of problems" the firm is pursuing.
No discussion yet for this article
Get curated AI news from 200+ sources delivered daily to your inbox. Free to use.
Get Started FreeFree · takes 30 seconds · unsubscribe anytime
1 minute a day. The AI essentials.
200+ sources · Email / LINE / Slack
