One subtle reason why the market size for AI Agents is going to be bigger than we realize is there are many use-cases where running multiple agents in parallel to solve the same problem is valuable. This was effectively never possible with most knowledge work before. It just would’ve been cost prohibitive to throw more people at the same problem. But because we have dramatically lowered the cost of deploying intelligence, you can afford to have more redundancy and consider things from even more angles. We’re already seeing these examples in a number of areas. People are using AI code review agents on the same code changes to offer subtle differences in analysis. The same would be true for having security agents review code, where there’s only more value in additional coverage. In deep research tasks, it’s often useful to have multiple prompts running in parallel to compare different approaches. And we can expect this will show up across many other fields where multiple differing views offer incremental value to a workflow, like legal work, healthcare, financial analysis, scientific research, product ideation, marketing content creation, and more. This is why it’s impossible to estimate the full TAM of many of these markets. If you thought agents only approximate how we work today 1:1, it would be easy to miss all the ways that we will be using agents in workflows. This is why these markets have no natural ceiling in the future.