New AI Promises to Assign Your Workload and Predict Who Might Quit

A researcher named Jay Barach has published a paper describing a tool called TalentOptima, which uses AI to help companies manage people. In plain terms, it watches your workload, your projects, and your performance data, then suggests who should do what — and quietly flags who looks likely to burn out or leave. It's designed to slot into the HR software a company already uses, so managers see recommendations instead of spreadsheets.

The author tested it by having 40 managers run through simulated scenarios. According to the paper, the results were clear: HR costs went down, productivity went up, and fewer employees quit. The paper also claims other HR tools only track whether work gets done, while this one pushes managers toward smarter decisions. If that holds up in the real world, it could mean less time on scheduling and admin for managers — and fewer surprises when a good employee suddenly resigns.

But here's the catch, and it's a big one. This is an early preprint — a paper posted online that hasn't been checked by other experts — and it reads more like a sales pitch for the tool than a neutral study. There's one author, one small simulation with 40 managers, and no outside verification. Real companies, real budgets, and real people are messier than a simulation.

There's also a human cost worth thinking about. Software that predicts who might quit is really software that scores you. If the data is biased — say, against caregivers, older workers, or people who took leave — the AI could quietly punish them. And if your workload is assigned by an algorithm, you deserve to know how it decided. The direction is clear: AI is moving from writing emails to managing people. The rules and safeguards are lagging behind.