No computer is even as powerful and complex as the human brain. The masses of tissue in our skulls can process information at quantities and speeds that computing technology can barely touch.
The key to the brain’s success is the efficiency of neurons in acting as a processor and memory device, as opposed to the physically separate units in most modern computing devices.
There have been many attempts to make computing more brain-like, but a new effort takes it all one step further — by integrating real human brain tissue with electronics.
It’s called Brainoware, and it works. A team led by engineer Feng Guo of Indiana University Bloomington has equipped it with tasks such as speech recognition and mathematics problems such as predicting nonlinear equations.
It was slightly less accurate than a pure AI computer, but the research demonstrates an important first step in a new type of computer architecture.
However, while Gu and his colleagues followed ethical guidelines in developing Brainoware, several researchers from Johns Hopkins University noted in a related report Nature electronics Comment on the importance of keeping ethical considerations in mind as we expand this technology further.
Lena Smirnova, Brian Cafu, and Eric C. Johnson, who did not participate in the study, to caution“As the complexity of these organic systems increases, it is important for society to consider the myriad neuroethical issues surrounding biocomputing systems involving human nervous tissue.”
The human mind is amazingly amazing. And there is appreciation 86 billion neuronson average, and Up to a quadrillion synapses. Each neuron is connected to up to Another 10,000 neuronsconstantly shooting and communicating with each other.
So far, our best efforts to simulate brain activity in an artificial system have not reached the surface.
In 2013, Riken’s K computer – then one of the world’s most powerful supercomputers – was launched. He made an attempt to imitate the brain. Using 82,944 processors and a petabyte of main memory, it took 40 minutes to simulate one second of the activity of 1.73 billion neurons connected by 10.4 trillion synapses, or only about one to two percent of the brain.
In recent years, scientists and engineers have tried to get closer to the brain’s capabilities by designing devices and algorithms that mimic its structure and workings. known as Neural computingIt’s getting better but it consumes a lot of energy, and training artificial neural networks takes a long time.
Gu and his colleagues sought a different approach using real human brain tissue grown in a lab. Human pluripotent stem cells have been stimulated to develop into different types of brain cells that organize into small, three-dimensional brains called organelles, complete with connections and structures.
These are not real brains, but merely arrangements of tissue without anything resembling thought, emotion, or consciousness. They are useful for studying how the brain develops and functions, without interfering with actual humans.
Brainoware consists of brain organoids connected to an array of high-density microelectrodes, using a type of artificial neural network known as Tank computing. Electrical stimulation transmits information to the organoid, the reservoir in which that information is processed before Brainoware outputs its calculations in the form of neural activity.
Regular computers are used for the input and output layers. These layers had to be trained to work with the organoid, with the output layer reading the neural data and making classifications or predictions based on the input.
To demonstrate the system, the researchers gave Brainoware 240 audio clips of eight male speakers making Japanese vowel sounds, and asked it to identify a specific individual’s voice.
They started with a naive organic. After training for just two days, Brainoware was able to identify the speaker with up to 78% accuracy.
They also asked Brainoware to predict Henon map, a dynamical system exhibits chaotic behavior. They left it unsupervised to learn for four days — each day representing a training epoch — and found that it was able to predict the map with better accuracy than an artificial neural network without a long-short-term memory unit.
The brain programs were slightly less accurate than the long- and short-term memory artificial neural networks, but all of these networks underwent 50 training epochs. Brainoware achieved roughly the same results in less than 10 percent of the training time.
“Due to the high plasticity and adaptability of organoids, Brainoware has the flexibility to change and reorganize in response to electrical stimulation, highlighting its potential for adaptive backup computing.” The researchers write.
Significant limitations remain, including the issue of keeping organs alive and healthy, and the power consumption levels of peripheral equipment. But, with ethical considerations in mind, Brainoware has implications not only for computing, but also for understanding the mysteries of the human brain.
“It may be decades before general biocomputing systems are created, but this research is likely to generate fundamental insights into the mechanisms of learning, neurodevelopment, and the cognitive effects of neurodegenerative diseases.” Smirnova, Cafu and Johnson write.
“It could also help develop preclinical models of cognitive impairment to test new treatments.”
The research was published in Nature electronics.
“Typical beer trailblazer. Hipster-friendly web buff. Certified alcohol fanatic. Internetaholic. Infuriatingly humble zombie lover.”