https://br.sputniknews.com/20211220/cerebro-ciborgue-e-criado-com-neuronios-humanos-e-aprende-mais-rapido-que-ia-video–20749042.html
‘Cyborg Brain’ was created using human neurons and learns faster than artificial intelligence (video)
‘Cyborg Brain’ was created using human neurons and learns faster than artificial intelligence (video)
The device called “DishBrasin” learned in five minutes to play Pong, while the AI needed an hour and a half to execute… 20.12.2021 Sputnik Brasil
2021-12-20T07:30-0300
2021-12-20T07:30-0300
2021-12-20T07:30-0300
society and every day
Science and Technology
study
Scientists
Artificial intelligence
/ html / head / meta[@name=”og:title”]/@Content
/ html / head / meta[@name=”og:description”]/@Content
https://cdnnbr1.img.sputniknews.com/img/1530/19/15301943_0:175:3018:1873_1920x0_80_0_0_2462835b996776cb7c8eecf962c78be2.jpg
Scientists at Australian Cortical Labs have created a chip containing human neurons, project leader Brett Kagan told NewScientist. The system, called “DishBrain,” consists of a petri dish with brain cells grown on an array of microelectrodes capable of stimulating and detecting signals. “We often refer to them as the ones who live in the Matrix. When they’re in the game, they think they’re the paddle,” Kagan adds. about traditional artificial intelligence. While DishBrain learned to play Pong in five minutes, the AI took an hour and a half to do so. However, they both understood the mechanism of the game, DishBrain showed worse results than artificial intelligence and lost matches played with virtual intelligence. According to the scientists, the model works with both human cells and mice. However, there is a difference between the two models. “Using this DishBrain system, we have shown that only one layer of cortical neurons ‘in the lab’ can automatically regulate and display intelligent and sensitive behavior when entering a simulated game world.” scholars.
Forabozo
You have to try brain cells out of a pocket atom inferior to those of mice.
0
1
2021
News
br_BR
https://cdnnbr1.img.sputniknews.com/img/1530/19/15301943_145:0:2874:2047_1920x0_80_0_0_051394f9ec6552608b6f87710aa14057.jpg
Science and technology, study, scientists, artificial intelligence
The device called “DishBrasin” learned in five minutes to play pong, while the AI took an hour and a half to perform the task.
“Through electrophysiological stimulation and recording, the cultures were introduced into a simulated game world, to mimic the card game Pong,” the scientists write in study.
“We often refer to them as live in a matrix. When they’re in the game, they think they’re the racket,” Kagan adds.
While DishBrain learned to play Pong in just five minutes, the AI took an hour and a half to do so. However, both understanding the mechanics of the game, DishBrain showed worse results than AI and Lost matches played With virtual intelligence.
According to the scientists, the model works with cells from both humans and mice. However, there is a difference between the two models.
The researchers note that “human cortical cells have always outperformed rat cortical cells with subtle game characteristics,” noting that this is the first evidence that human neurons per se are superior to those of rodents.
More Stories
What ChatGPT knows about you is scary
The return of NFT? Champions Tactics is released by Ubisoft
What does Meta want from the “blue circle AI” in WhatsApp chats?