The real economic value in robotics isn't walking and talking, but predicting structural failure. After thirteen years, Jake Lusararian's thesis is gaining traction as AI hype meets industrial necessity. His company, Gecko Robotics, builds deterministic systems for inspecting refineries and bridges - high-stakes environments where a data error means catastrophe, not a chatbot glitch.
On This Week in Startups, Lusararian argues that the AI boom has spotlighted the desperate need for non-hallucinatory data from the physical world. The goal is reliability, not novelty.
Jake Lusararian, This Week in AI:
- The models are putting a huge spotlight on the importance of valuable data sets that don't hallucinate.
- Especially with things that if they do hallucinate, could cause an explosion and kill people.
This drive for precision slams into a fragmented hardware landscape. Chris Latner, CEO of Modular, explains the deployment crisis. Chipmakers like NVIDIA, Apple, and AMD build proprietary software stacks, forcing developers into a mess of incompatible toolkits. This lack of portability creates a bottleneck for the specialized, optimized systems that industrial robotics require.
Latner's company aims to be the unifying layer, replacing vendor lock-in with flexibility. The convergence point is clear: applied AI's future depends on robots that reliably maintain infrastructure and software that reliably runs on any chip inside them. The industry is shifting from speculative demos to deployment where errors have physical consequences.
