Jed McCaleb, that guy who founded Ripple and Stellar, is now heavily investing in AGI. And it's not a small amount — he's putting 1 billion dollars of his fortune (which is around 3.9 billion) through the Astera Institute, a non-profit organization he created. Additionally, he committed another 600 million specifically for neuroscience research.

What’s happening at Astera is quite interesting. They are mapping neural activity of rats while performing tasks, using brain-machine interfaces, and then turning these discoveries into new AI architectures. The idea is to expand this to primates and then humans. The project is led by Dileep George, who left DeepMind, and they plan to grow to 30 researchers this year.

Jed McCaleb has a very clear critique of the Transformer architecture that dominates now. To him, it only makes predictions, lacking planning, decision-making, motivation — things that the human brain does naturally. He believes that an AGI based on principles of the human brain will be much easier to understand and control.

It’s interesting that Yann LeCun, the former chief AI scientist at Meta, is heading in the same direction. His AMI lab just closed a seed funding of 1 billion dollars to research 'world models.' Meanwhile, the team at OpenAI, represented by Sam Altman, thinks differently — for him, AGI will require multiple medium-sized breakthroughs, not just a new direction.

Jed McCaleb’s bet on biology and neuroscience as the foundation for AGI is quite different from the mainstream, but it makes sense if you consider that the brain is the only system that has truly achieved general intelligence.