ChatGPT Work is bringing agentic workflows to a much broader audience, but the transition from using a chatbot to managing agents is more complicated than the labs make it look.
Organizations need to help people learn how to assign work, provide context and tools, evaluate outcomes, and know when to step in.
Mostly Serious tested AI-assisted workflows in Vei, our internal test environment for realistic business systems. In 32 of 36 completed runs, at least one field we checked was still wrong after better information arrived.
The issue was not always that the AI misunderstood the situation. Often, the AI made the summary better while failing to update the fields another person would rely on later.
A website can keep loading, submitting forms, and technically doing its job while quietly introducing people to an outdated version of the business behind it.
Our long-term relationship with OMB shows what happens when a website foundation and the team behind it keep adapting as the business grows into new divisions, new audiences, and new offerings.
The AI industry wraps simple concepts in complex language. Agents, agentic workflows, multi-agent orchestration—it all sounds like it requires a computer science degree. It doesn't.
A practical look at how folder-based AI systems handle real organizational work, from marketing content to client onboarding, and why the barrier to entry is lower than the jargon suggests.
Senior leaders often speak first in meetings without realizing the cognitive consequences. When a leader offers an early opinion, it anchors the group, narrows thinking, and reduces cognitive diversity. Research shows that evenly distributed participation improves collective intelligence, yet power dynamics make that difficult in practice. If leaders want better decisions, they must discipline their share of voice and intentionally create space for others to think and speak first.