The quest to replicate the sophistication of the human brain has led computer scientists to develop artificial neural networks in an attempt to create more powerful artificial intelligence (AI). However, as these artificial neural networks become increasingly complex and powerful, they also consume a significant amount of energy. In contrast, a Swiss start-up company has introduced a ‘biocomputer’ that interfaces with living brain cells and claims to operate more efficiently than traditional computers by leveraging the energy efficiency of nature’s design.
FinalSpark’s innovative biocomputer platform utilizes spherical clusters of lab-grown human brain cells known as organoids. These organoids are integrated into the system within four arrays, each connected to eight electrodes and a microfluidics system that supplies essential water and nutrients to the cells. This unique approach, termed wetware computing, demonstrates the growing capabilities of researchers to culture organoids in the lab and gain insights into the functioning of miniaturized organ replicas.
According to FinalSpark, bioprocessors like their brain-machine interface system are remarkably energy efficient compared to traditional digital processors. While specific data on their system’s energy consumption or processing power is not provided, the research team highlights the substantial energy savings potential of their approach. For instance, training a large language model like GPT-3 requires an immense amount of energy, which contrasts starkly with the energy-efficient operation of the human brain that consumes significantly less energy to power its extensive neural network.
The growing popularity and utilization of artificial neural networks, particularly in AI applications like large language models, have contributed to a surge in energy consumption within the IT industry. With projections suggesting a substantial increase in electricity consumption by the AI sector by 2030, there is a pressing need to enhance the energy efficiency of computing systems. Exploring synergies between brain cell networks and computing circuits offers a promising avenue for developing more sustainable and energy-efficient technologies.
While FinalSpark’s biocomputer system represents a significant milestone in leveraging biological systems for computing tasks, it is not the first endeavor in this domain. Researchers in the United States previously developed a bioprocessor that connected computer hardware to brain organoids, demonstrating its ability to learn and recognize speech patterns. FinalSpark’s system, which enables remote connectivity and continuous monitoring of electrical activity in brain organoids for up to 100 days, has garnered interest from various research groups seeking to conduct experiments in wetware computing.
Future Prospects and Research Applications
As FinalSpark continues to enhance the capabilities of its biocomputer platform for a broader range of experimental protocols, researchers are poised to explore new avenues in wetware computing. From injecting molecules and drugs into organoids for testing to facilitating diverse research experiments, the potential applications of this innovative technology are vast. Whether these advancements lead to energy-efficient computing solutions or further advancements in organoid research remains to be seen, but the collaborative efforts of researchers in this field promise exciting developments in the near future.
Leave a Reply