What to build when robotics enters the smooth exponential (www.giete.ma)

Recent discussions in the AI and robotics community highlight a transformative period where advancements in robotics are anticipated to follow a "smooth exponential" trajectory, driven by enhanced machine intelligence. Eric Jang's insights suggest that robotics is reaching a critical mass, characterized by increased participation, the integration of modern large language models (LLMs) to evaluate robotic capabilities, and access to extensive training data. A notable example is Astra, a model that recently demonstrated impressive robotic control by creating paintings, hinting at future improvements in speed and efficiency. Predictions suggest that within two years, such models may execute tasks up to 10 times faster, thus lowering barriers to entry for developers in the field. This evolution is significant for the AI/ML community as it not only democratizes access to robotics through reduced hardware costs but also generates vast amounts of real-world data that can enhance model performance. As companies like Skild achieve significant revenue milestones, the focus shifts to finding deployment strategies that harness this technological momentum. Emphasizing deep integration into existing workflows and capturing specific customer needs will be crucial for new entrants in an increasingly competitive landscape. The anticipated convergence of LLM capabilities with robotics not only enhances task execution but could potentially revolutionize entire industries, paving the way for substantial economic growth in the sector.