AI Agents provide an opportunity for bringing automation to so many categories of work that you could never automate with software before. Most of the time in software you’re competing within existing, largely well defined markets. While the spend may even be growing, there usually are still zero sum dynamics in the space. Given the reset happening with AI Agents, it would be a mistake to largely focus on just building what we’ve always been able to do with software, and just incrementally adding some AI capabilities on top. The big play right now is to build AI Agents for brand new categories by bringing a deep domain expertise, vertical understanding, data from the customer, and tool use. There are endless categories that would have been nearly impossible, or at least far too expensive, to automate before, where we finally can now. If you had tried to do this even 1 year ago it probably would have been too early. And ultimately, these categories will likely meaningfully exceed the total size of current enterprise software spend, as they can be tied with the amount of output generated vs. capped by a per seat price. Incredible times.