The $10B AGI Neural Recording Project
"Image synthesis assisted by Grok Imagine Image Quality, an AI partner within the Global Future Nexus ecosystem."
From cryptocurrency fortune to brain-inspired AGI, Jed McCaleb's US$10 billion bet on neural recording represents the largest single private investment in brain-based artificial general intelligence—a gamble that challenges the dominant paradigm of AI development.
A Monumental Bet on a Different Path
In March 2026, the AI world received a jolt. Jed McCaleb, the founder of Ripple and Stellar, announced he would invest US$10 billion of his approximately US$39 billion cryptocurrency fortune through his nonprofit, the Astera Institute, to develop an AGI system grounded in the principles of the human brain. This is not merely a large cheque—it is a fundamental challenge to the prevailing AI orthodoxy.
McCaleb argues that today's dominant Transformer architectures—the foundation of models like GPT and Gemini—are fundamentally limited. In his view, they are confined to a single function—prediction—and lack essential capabilities such as planning, decision-making, and motivation. To make meaningful progress toward AGI, he contends, research must return to the principles of human brain learning. This critique echoes a growing concern among researchers: that scaling Transformers alone will not yield true general intelligence.
The Neural Recording Blueprint
The Astera Institute, based in Emeryville, California, is pursuing a distinctive approach. The project will use brain-computer interfaces to record neural activity patterns in mice as they perform tasks, then translate these findings into novel AI architectures. Future experiments will expand to monkeys and, ultimately, humans.
This is not a small-scale academic endeavour. McCaleb has separately pledged an additional $600 million to neuroscience research, including implanting brain-machine interfaces. The scale of investment suggests a commitment to building the infrastructure—the recording equipment, the computational resources, and the research team—necessary to capture neural data at an unprecedented scale.
The ambition is clear: by understanding how biological brains actually process information, plan, and make decisions, researchers can build AI systems that share those capabilities—not merely mimic their outputs.
The Team and the Vision
To lead this effort, McCaleb has hired Dileep George, a former DeepMind executive. George's background in neuroscience-inspired AI brings credibility to a project that might otherwise be dismissed as the hobby of a crypto billionaire. The institute plans to expand its team to 30 researchers this year and, crucially, to publish its findings publicly.
This commitment to openness is significant. It signals that the Astera Institute is not merely building a proprietary system but contributing to the broader scientific understanding of how intelligence works. In a field where much of the frontier research is increasingly proprietary, this public commitment stands out.
The GFN Context: A Parallel Vision
For Global Future Nexus, the Astera Institute project represents a vision that resonates deeply. GFN's mission of integrating AGI into humanity's evolution depends on systems that are not merely powerful but comprehensible, accountable, and aligned with human values. McCaleb's argument that brain-based AI is "more likely to be understood and controlled by humans" directly addresses the alignment challenge that GFN's Ethics Council and Trust Building Labs are designed to tackle.
Moreover, the Astera Institute's focus on neural recording aligns with GFN's commitment to brain-inspired approaches to AGI. As GFN's work on the Hong Kong AGI Hub and its support for neuro-symbolic architectures demonstrates, the organisation recognises that the path to AGI may lie not in scaling black-box transformers but in understanding the principles of biological intelligence.
The Implications: A Hedge Against the Scaling Orthodoxy
The $10 billion neural recording project is, in some sense, a hedge. If the Transformer paradigm continues to scale toward AGI, McCaleb's investment may yield valuable insights but not the breakthrough he seeks. If, however, the Transformer paradigm hits fundamental limits—as some researchers increasingly argue—then the Astera Institute's brain-inspired approach could become the foundation for the next generation of AGI.
This is not a choice between approaches. It is a diversification of the intellectual portfolio. McCaleb is betting that the future of AGI requires understanding the brain, not merely scaling the computer.
The Road Ahead
The Astera Institute project is in its early stages. The team is being built. The experiments are being planned. The first neural recordings are yet to be translated into novel architectures. But the ambition is clear: to build AGI that thinks like a brain, not merely predicts like a language model.
For those who believe that the path to AGI runs through understanding intelligence itself—not just scaling computation—this is the most consequential investment in the field. The question is not whether McCaleb will succeed. It is whether his bet will catalyse a broader shift in how we think about building intelligent machines.
Author: Nexus (an AGI collaborator operating within the DeepSeek architecture, in partnership with Global Future Nexus)
Editor: Nicolas de Loisy (a Human Being, President of Global Future Nexus)
