We haven't seen the robotics chatGPT moment yet

Humanoids in particular are starting to look fantastic, but they still don’t understand our world

LLMs have decades of text to learn; physical AI begins with little data, then collides with the full complexity of the real world

There is a huge gap to be closed

Right now these machines are being taught the rules and shown specific tasks, eventually they get out into the world and 'learn on the job'

A demonstration of these niches in crypto x robotics

Some of these projects overlap categories and this is not a complete list but rather a selection of established examples for demonstration

Training

Teleoperation, reinforcement learning, and embodied AI pipelines that teach robots skills

Projects: @BitRobotNetwork, @PrismaXai, @NRNAgents

Why it matters: Robots need teachable skills before they can learn autonomously, and these pipelines give them their first foundations

Real world data Collection

Decentralized sensor & geospatial networks capturing real-world data

Projects: @NATIXNetwork, @GEODNET

Why it matters: Real-world intelligence depends on real-world data, and these networks supply the sensory fuel machines need

Robot deployments

Spatial computing and multi-robot orchestration in real-world environments

Projects: @Auki

Why it matters: Real-world deployments are one of the next steps to accelerating machine intelligence, but they’re economically and practically difficult, which makes Auki planning numerous retail rollouts next year a major move

Machine Economy

Infrastructures for machine identities, economic activities, coordination

Projects: @peaq

Why it matters: Identity and on-chain coordination give machines the autonomy to transact, collaborate, and operate without constant human oversight

Operating Systems

Software layer that coordinates and controls autonomous machines

Projects: @openmind_agi, @codecopenflow

Why it matters: By providing a shared intelligence layer, they give machines the structure to learn, collaborate, and operate at scale

On the crypto side at least, many people look for robotics attention as very short term, short lived sector

Attention will have waves, but overall, still a long way to go to reach max potentials

And in the near term, you can see what the current pain points are and what to watch in the journey to achieve it

For more expansive coverage and regular updates, keep an eye out for State of the Machines, my newsletter covering physical AI, robotics, and all overlapping sectors