State of the Art of Biological Computing Ewelina Kurtys & Charles Humble

The podcast explores emerging trends in computing, emphasizing the intersection of AI and bio-inspired technologies. It contrasts current AI systemsreliant on narrow, extrinsic learning methods like gradient descentwith the efficiency and adaptability of biological brains, which utilize intrinsic learning mechanisms. The discussion highlights bio-inspired computing as a potential paradigm shift, with a focus on bioprocessors using living neurons. Dr. Evelina Curtis, a scientist and entrepreneur at Final Spark, explains how her background in neuroscience and biotechnology led her to advocate for rethinking AI through biological systems. She underscores the importance of team culture, remote collaboration, and transferable skills from academia to industry, while addressing challenges like the energy efficiency of living neurons compared to digital hardware and the technical hurdles of scaling biological systems.

Key topics include the high costs of AI, particularly large language models (LLMs), and the potential of living neurons to drastically reduce energy consumption, though they face limitations in speed and scalability. The conversation delves into the complexities of decoding neural signals, the deterministic nature of brain activity, and ethical debates around stem cell use and commercialization. Final Sparks work involves developing prototypes with human stem cells to study neuronal signal processing, though scaling to a functional computing device requires overcoming scientific and logistical challenges, including replicating natural physiological conditions and ensuring reproducibility in neuroscience research. The podcast also touches on the philosophical implications of biological determinism and the need for interdisciplinary collaboration to address unresolved questions in biocomputing.