The Machine Workforce Layer: Why Managing Robot Employees Is the Next Big HR Problem

We're really in the middle of this big paradigm shift in economic value creation. We're going from the traditional economy to the autonomous economy. All the signs are there. You already see it with agents and you see it with all the robots, the announcements from Figure and Tesla with the humanoids. So here is ultimately what it boils down to. This is our investment thesis. If you put together everything I said, this is our AE autonomous economy value stack and where we want to invest going forward for the rest of Fund 1 and all of Fund 2. So at the bottom you have the intelligence and compute sub- Substrate. So this is everything that provides the intelligence, everything that's physical and/or the models. So everything sits there. Then you have all these kind of new layers. They're not looked at typically this way. What are we going to need in this new world? We're going to need the agent runtime and orchestration layer, all these new things that we don't have really to the scale that we need to have them— planning, tool calling, memory, workflow graphs, sandboxes. Then if we go up, we have the autonomy control plane, right? So how are we going to control and make sure that the brains of it, everything is working right. Identity, permissions, policy evaluations, audit, rollback, kill switch. These are the new rails that agents need to be able to play nicely with each other. Then you have the machine workforce layer. Now all of a sudden we had HR departments that you manage just your human employees. Now you have to manage all your robot employees, right? And so you have to procure them. You have to see, is he doing good? Is he doing bad? You have to know what their skills registry is. You have to manage the fleet. They're going to break down. They need to be serviced. And then on top of that, you have the autonomous experiences to the enterprise, right? So these are all the Gentec apps.