EFRI BEGIN OI: Integrating Human-Derived Neural Networks and AI for Information Processing in Brain Organoids

About This Grant

Artificial Intelligence (AI) is poised to dramatically alter our world; however, one of the biggest challenges to progress in AI is the high energy consumption of the hardware and the lack of flexibility for adapting to new environments or problems. By contrast, the human nervous system has evolved to become flexible, nimble, and energy-efficient. This award will support the development of biotechnology to grow cellular brain organoids capable of replicating the network and activity observed in the brain. These organoids will be integrated into an engineered system and, using insights from cognitive science, programmed to solve increasingly complex associative problems. The research team will evaluate how this bioengineering technology compares to state-of-the-art computer chips that mimic brain function. The team will also study the social and ethical implications of using biological tissue to address AI challenges. This project aims to foster a new field at the interface of computer science, neuroscience, and neuroengineering. There is increasing interest in engineering systems that incorporate organoids - self-organized 3D cellular structures in vitro that resemble organs. In the context of the societal need to develop machine learning (ML) and AI for increasingly complex tasks under low-energy consumption demands, brain organoids offer considerable promise. Brain organoids may recapitulate fundamental computational learning and memory functions not achievable with silicon-based approaches. Research supported through this award will use cellular reprogramming tools to convert blood cells into human induced pluripotent stem cells (hiPSCs), with subsequent differentiation into neurons. These will be used to build brain organoids that integrate excitatory and inhibitory neurons in a configuration similar to that found in the neocortex. The motivating hypothesis is that the self-organization of these cortical-like brain organoids can be harnessed to perform computational, learning and memory tasks. However, for brain organoids to be viable for ML/AI applications, it is essential to demonstrate that they can be reliably produced and programed using biological mechanisms of learning and memory. The proposed research will establish a framework to characterize, validate, and ‘program’ brain organoids. Ethical considerations will be embedded throughout the project to enable in-depth analyses of emerging societal and ethical questions surrounding the use of brain organoids in ML/AI. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.