AI course, part 2 – Potential AI futures

First, we map two visions of the future. On the linear, shorter-term track, AI simply makes processes cheaper, faster and more accurate. On the less-trodden ‘what if?’ track, entirely new asset classes emerge – think data streams or algorithm ‘royalties’, and hyper-personalized portfolios managed by tireless digital concierges.

A centerpiece of that second path is the concept of agentic AI. Picture a team of autonomous sub-agents: research, PM and trading, coordinated by an orchestrator. They can spot a potential merger, run risk scenarios and place trades in minutes, then hand humans a fully auditable report. Scale, speed and explainability converge.

Such power demands robust governance. Human roles won’t disappear; they’ll evolve into oversight, strategy and storytelling functions that keep ethics, risk and client trust front and center. Upskilling will be critical as the boundary between quant and fundamental investing blurs.

So, are we headed for incremental change or a paradigm shift? This long read lays out what needs to happen, technologically and institutionally, for investors to prepare portfolios (and skill sets) for multiple futures.