If AIs are the workers of the 21st century, what is the right recruitment process for them?

Hello everyone, I share with you a little thought I had a few days ago. For the context, we are creating at Eden AI a new standard where we aggregate and harmonize various AIs available via APIs so that consumers can very easily see what is possible to do with them, compare competing AIs, and above all use them with the greatest possible flexibility: change them when you want, make several work together, etc. While talking with our users, I found myself talking about some AI models as if they were more or less efficient workers: "Well, DeepL performs well for translating European languages, but for Chinese, we should get Google or NeuralSpace to work on it", etc. Very disturbing sentence in retrospect. So I wonder if we are not becoming a kind of recruitment agency for AI on behalf of our users, with the particularity that unlike "human workers", we can change "employees" in a fraction of a second (and fortunately by the way). What do you think of all this? Has anyone here ever had a similar thought? Best, Taha