The Ghost Shift

The ghost shift is the nicest team you ever had. It shows up on time, it does not complain, it does not want remote Fridays. Three bots, one desk: at seven the first draft is there, the meeting recapped, the inbox empty. You are more efficient. Do you get a raise? Of course. In tasks. The firm sends flowers in the shape of a badge: AI-ready. Goydan and Piccolo Brunelli ask, without irony, which roles are being redesigned and which are simply asked to do more (Goydan & Piccolo Brunelli, 2026). The intern has a subscription instead of a contract. You have a promotion nobody asked you about. From making to nodding.

They sort, phrase, book, generate eighty variants. Inside the machine’s jagged frontier that goes faster and better. Just outside it, on a complex managerial task, those using AI were less likely to be right (Dell’Acqua et al., 2026). We sign. That is called leadership and means quality control at piece rates. Eight contracts become forty, and you get to find the interesting ones. Niederhoffer and colleagues call what then lands in the inbox workslop: it looks finished before anyone has understood it (Niederhoffer et al., 2025). Nobody reads the minutes, so a bot recaps the minutes, and someone recaps the recap. With the first sentence, the bots work for us. With the second, which reads “please review”, and the third, “by yesterday”, we work for them.

A bot as CEO is charming: it answers at three, it has no mood, in the morning the strategy is in the inbox. The board nods. Liability remains a human topic, like the dishes. Singapore insists humans stay accountable, even as continuous oversight of every workflow becomes impractical at scale (IMDA, 2026). He and colleagues watched a firm that wanted to double output: review doubled with it, and after a while bots were reviewing the bots (He et al., 2026). We have tempo enough now. Is it a pay grade, or only a shift you proof yourself?

Sources

Dell’Acqua, F., McFowland, E., Mollick, E., Lifshitz, H., Kellogg, K., Rajendran, S., Krayer, L., Candelon, F., & Lakhani, K. R. (2026). Navigating the jagged technological frontier. Organization Science, 37(2), 403–423. https://doi.org/10.1287/orsc.2025.21838 (abstract only)

Goydan, P., & Piccolo Brunelli, J. (2026, May 5). Making AI productivity pay off. BCG. https://www.bcg.com/publications/2026/making-ai-productivity-deliver-real-value (full text)

Niederhoffer, K., Kellerman, G. R., Lee, A., Liebscher, A., Rapuano, K., & Hancock, J. T. (2025, September 22). AI-generated “workslop” is destroying productivity. Harvard Business Review. https://hbr.org/2025/09/ai-generated-workslop-is-destroying-productivity

IMDA. (2026). Model AI governance framework for agentic AI (Version 1.5). Infocomm Media Development Authority. https://www.imda.gov.sg/assets/d1e45a56-d898-4b9f-972e-8ae94be1a5a5.pdf (full text)

He, H., Agarwal, S., Denisov-Blanch, Y., Azaletskiy, P., Koyejo, S., & Vasilescu, B. (2026). AI writes faster than humans can review. arXiv:2607.01904. Preprint. https://arxiv.org/abs/2607.01904 (full text)

Previous
Previous

Problem and Solution Wander Together

Next
Next

Design as Checkbox