- FinalSpark’s AI biocomputers operate 100,000 times more efficiently than silicon-based systems.
- FinalSpark organic AI biocomputers have a lifespan of 100 days.
- Rent AI access powered by human brain cells is now available for $500/month.
A Swiss technology company, FinalSpark, has introduced a service that provides scientists with biocomputers derived from human brain cells, hosted in the cloud and costing $500 per month.
This new technique provides a low-cost training tool for non-deep learning-based AI models, which should optimally utilize the power of AI with less energy demand, as solutions based on deep learning do.
Advancing AI with biocomputing
FinalSpark is a biocomputer made from organoids, or clusters of human brain cells, and is functional for up to 100 days. These biocomputers, which FinalSpark founder Fred Jordan has pointed out, could be 100,000 times larger than those of today’s conventional silicon-based systems.
It must be emphasized that much of this significant increase in efficiency is due to the use of organic materials, which allows information processing to take place more sparingly.
Training techniques for organic AI
While conventional AI training focuses on numerical values and reinforcement learning, organoid-based training involves biological activities. Scientists first use dopamine for positive reinforcement, the way rewards promote learning in the human brain, and electrical signals are used for negative reinforcement.
This approach is more realistic than the previous one and closer to the principles developed by scientists in neuroscience. However, these biocomputers have a raw potential for innovation due to fundamental challenges associated with their longevity.
Each organoid can run at an optimal level for just over three months, after which the output quality deteriorates, requiring the AI system to migrate to a new set of brain cells.
This cycle raises questions about the long-term sustainability of AI projects, but also raises questions about the possibility of extending the life cycle of such dynamic and organic systems in the future.
Sustainable and efficient AI development
The biocomputing technology offered by FinalSpark will not only help reduce the cost of developing artificial intelligence, but will also help reduce the burden of extensive calculations that contribute to environmental pollution.
Using human brain cells, the company wants to set new precedents for the further evolution of AI and perhaps create something that will have the same lifespan as humans. It bridges the gap between biological and artificial intelligence, an indication that a future is possible where AI will be sustainable and in touch with humanism.