The Wall Street Journal reports that AI agents are producing an unexpected effect inside some startups: instead of reducing founders' workloads, they are encouraging them to work almost continuously. As agents become capable of coding, researching, monitoring campaigns and completing multi-step tasks, founders feel compelled to remain available whenever an agent needs guidance or approval.
Aditya Sharma, co-founder of Keel, describes staying awake until 6 a.m. because his AI agents could finish tasks at any time. His concern is that leaving an agent blocked for eight hours could mean losing a substantial amount of productive work. Other founders interviewed by the Journal describe similar experiences, with AI effectively creating a 24/7 work environment rather than a conventional workday.
The productivity gains can be enormous. Peter Pezaris, founder of Proxon, estimates that his six-person startup operates about 30 times faster with AI agents than it would without them. But the article highlights an important paradox: agent productivity can create more human work. Faster product development means more customers to onboard and support, while faster research creates more ideas to pursue. Instead of eliminating work, AI can expand the amount of work a small team believes it can—and should—attempt.
This creates a new form of AI-driven FOMO. If an agent can accomplish in hours what previously took days, every hour away from work can feel like a missed opportunity. Some founders are even monitoring agents through smartwatches, allowing them to approve actions while exercising or away from their computers. The result is a blurring of the boundary between work and personal life.
The bigger lesson is that AI productivity doesn't automatically translate into a better work-life balance. If companies measure success primarily by how much more can be accomplished, employees may simply use the additional capacity to do more. This creates a potential “productivity paradox”: AI makes work dramatically faster, but instead of requiring fewer hours, it can raise expectations about what must be accomplished.
That makes the future of AI-enabled work partly a management and cultural question, not just a technology question. Companies will have to decide whether AI's ability to operate around the clock means humans should also be constantly available—or whether the technology should instead be used to let people accomplish more within healthier boundaries.
