Human cognition is a full-body experience, not just a brain function. Current AIs are 'disembodied brains,' fundamentally limited by their lack of physical interaction with the world. Integrating AI into robotics is the necessary next step toward more holistic intelligence.

Dan Siroker outlines a three-part roadmap for achieving mind emulation: 1) a complete brain map (connectome), now feasible by 2040; 2) sufficient, cheap compute power, estimated to be ready by 2047; and 3) rich behavioral data, which the Limitless pendant is designed to capture.

Companies are now growing human brain cells on silicon chips and offering cloud API access for developers to code to them. This bio-compute model, which taught neurons to play a video game in a week, is vastly more energy-efficient than traditional GPU clusters, heralding a new computing paradigm.

Dr. Alex Wissner-Gross argues the ideal path to mind uploading avoids the philosophical "copy problem." Instead of a one-time scan, he envisions a gradual process of replacing individual brain cells with substrate-independent equivalents, ensuring a continuous, uninterrupted transfer of consciousness.

Startups and major labs are focusing on "world models," which simulate physical reality, cause, and effect. This is seen as the necessary step beyond text-based LLMs to create agents that can truly understand and interact with the physical world, a key step towards AGI.

Scientists mapped and simulated a fruit fly's brain. By only providing sensory inputs to the simulated neural structure, it correctly enacted motor responses like walking without any behavioral training or reinforcement learning. This suggests complex behaviors are inherent to the brain's wiring diagram itself.

Companies like Cortical Labs are growing human brain cells on chips to create energy-efficient biological computers. This radical approach could power future server farms and make personal 'digital twins' feasible by overcoming the massive energy demands of current supercomputers.

While today's computers cannot achieve AGI, it is not theoretically impossible. Creating a generally intelligent system will require a new physical substrate—likely biological or chemical—that can replicate the brain's enormous, dynamic configurational space, which silicon architecture cannot.

While discussions about biocomputing often veer into sci-fi fears of consciousness, the immediate, practical danger is biological. The neurons lack an immune system, making them highly vulnerable to contamination from bacteria or fungi, which can kill the cells and halt experiments.