Scientists built a computer.

The human brain is as complex as a computer can be. The neuron's efficiency in serving as both a processor and memory device is a key to the brain's success. Brainoware integrates real, actual human brain tissue with electronics, making it the first attempt to make computing more brain-like. The research shows an important first step in a new kind of computer architecture and is slightly less accurate than a pure hardware computer. The ethics guidelines that were followed in the development of Brainoware were not followed by several researchers from the University of Baltimore. As the sophistication of these organoid systems increases, it is critical for the community to examine the plethora of ethical issues that surround biocomputing systems incorporating human neural. The human brain has an average of 86 billion neurons and up to a few hundred thousand synapses. The best attempt at replicating the brain in an artificial system was made by Riken's K Computer. It took 40 minutes to create a simulation of one second of the activity of 1.73 billion brain cells with a petabyte of main memory. Scientists and engineers have been trying to approach the capabilities of the brain by designing hardware and software that mimics its structure and how it works. Artificial neural networks are time-consuming and energy-intensive, but a different approach was sought by Guo and his colleagues. These aren't real brains, but three-dimensional mini-brains with connections and structures, made from human pluripotent stem cells. They're useful for studying how the brain develops and works. Brainoware uses normal computer hardware for the input and output layers. The layers had to be trained to function with the organoid, with the output layer reading the neural data and making predictions based on the input. In order to demonstrate the system, the researchers gave Brainoware 240 audio clips from eight male speakers, and asked it to identify one specific individual's voice. Brainoware was able to identify the speaker with 78 percent accuracy after just two days of training. They left it to learn on its own for four days, and found that it was better at predicting the map than an artificial neural network. Brainoware was slightly less accurate than artificial neural networks with a long long-term memory unit. Brainoware has the ability to change and reorganize in response to electrical stimulation, which shows its ability for adaptive reservoir computing, according to the researchers. There are still significant limitations, including the issue of keeping the organoids alive and healthy. "This research is likely to generate insights into the mechanisms of learning, neural development and the cognitive implications of neurodegenerative diseases, but it may be decades before general biocomputing systems can be created," they wrote. It could help to develop models of cognitive impairment to test new drugs.