Asset-Light Model Layer for Robotics

This structure makes General Intuition look more like a compute heavy software layer than a robotics operator. The company pays for GPUs and model serving, but it does not have to buy and depreciate robot fleets, staff technicians to install and repair machines at customer sites, or carry the working capital burden of manufacturing hardware. That keeps fixed assets lower and lets new customers be added through API usage and model integration instead of robot rollout.

• Vertically integrated players like Figure and 1X are building robot bodies and control stacks together, which gives them tighter hardware software feedback loops but also means manufacturing, deployment, and service complexity sits inside the business. General Intuition avoids that layer by selling models rather than robots.
  1 sacra 2 sacra 4 sacra 5 sacra
• In robotics, field service is not a minor add on. It means technicians, spare parts, debugging, and customer site support when a robot jams, drifts, or breaks. Other embodied AI companies describe deployment, maintenance, and customer specific integration as major operational drag that slows scaling.
  2 sacra 6 sacra
• The tradeoff is that General Intuition gives up some of the best proprietary data loops. Companies that own deployed robots can capture failure cases from real jobs and feed them back into training. General Intuition instead leans on Medal data and partner integrations to improve models without owning the physical estate.
  1 sacra 2 sacra 6 sacra

Over time, the strongest model layer companies will keep pushing toward robotics revenue without fully becoming robot operators. The likely path is more OEM and platform partnerships, where General Intuition supplies the brain while hardware partners own installation and service. If that modular market structure holds, it can scale faster and with less capital than full stack robot manufacturers.