Software vs People vs AI Agents — Pricing, Security, UX

We are in the middle of a transition. Companies are buying AI agents. Vendors are selling them. But nobody has quite figured out the rules yet.

How do you price an AI agent? Answers range from tokens to subscriptions to outcome-based pricing — and all of them have problems. How do you secure one? How do you actually get value from it, and what does onboarding even look like when expectations rarely match reality?

In this talk, Eduards maps the chaos by comparing AI agents against two things we already know how to manage: software subscriptions and human employees. Each operates on a different model — different pricing logic, different security assumptions, different user experience. Understanding those differences is the first step to making better decisions about when and how to deploy AI agents in your organization