双语学习
登录注册网站首页

From mind to machine用“人脑”做的“生物计算机”诞生!

来源:融媒体采编平台
作者:张语迎
日期:2024-01-08
【导读】科学家创造出了用“人脑”做的“生物计算机”。

图源:视觉中国

The human brain is a powerful natural computer that has evolved over tens of thousands of years. But what would you get if you combined this biological computer with a real one?

人脑是一台强大的天然计算机,已经进化了数万年。但如果将这台生物计算机与真实的计算机结合起来,你会得到什么呢?
 

Researchers at Indiana University Bloomington in the US have linked human brain-like tissues to an electronic chip (芯片), which can perform simple computer tasks. Their creation, named Brainoware, is part of a growing field called biological computing that might one day outperform (超过) current computers, according to Nature.

美国印第安纳大学伯明顿分校的研究人员将类脑组织与电子芯片连接起来。该芯片可以执行简单的计算机任务。据《自然》杂志报道,他们将这套系统命名为 Brainoware,属于不断发展的生物计算领域的一部分,未来或超越当前的计算机。
 

The brain-like tissues the researchers used, also known as brain organoids (类器官), weren’t part of a living person’s brain. They were “grown” from human stem cells (干细胞), which are capable of producing different types of cells and forming body tissues. It took two months for the stem cells to mature into an organoid.

研究人员使用的类脑组织,又名类器官,不是活人大脑的一部分。它们是由人体干细胞培育而成,能够分化成不同类型的细胞并形成身体组织。干细胞需要两个月的时间才能达到成熟阶段,变为类器官。
 

The researchers then placed one organoid onto an electronic chip with thousands of electrodes (电极). Although organoids are much simpler and smaller than an actual brain, they act similarly to human brains, such as responding to electrical signals, which is what our brain does all the time. These responses lead to changes in the brain, which fuel our ability to learn.

随后,研究人员将一个类器官放置在带有数千个电极的电子芯片上。尽管类器官比真实的大脑更简单、更小,但它们的行为反应与人脑相似,例如对电信号做出反应——这个大脑一直在做的事情。这些反应会导致大脑发生变化,增强我们的学习能力。
 

To test Brainoware, the researchers used it for voice recognition by training it with 240 Japanese recordings spoken by eight speakers. They found that the organoid reacted differently toward each voice. With an accuracy of 78 percent, it successfully identified the speakers by showing different neural (神经的) activities.

为了测试 Brainoware,研究人员使用了8个人所说的240 段日语录音对其进行训练,从而将它用于语音识别。他们发现类器官对每种声音的反应都不同。通过识别神经活动的不同,它可以成功识别说话者,准确率达到78%。
 

What’s special about Brainoware is that researchers can take advantage of the organoid’s complexity without understanding its cell networks, according to Daily Mail. In other words, scientists don’t need to know exactly how the organoid works in order to use it.

据《每日邮报》报道,Brainoware的特别之处在于,研究人员可以在不了解其细胞网络的情况下利用类器官的复杂性。换言之,科学家不用确切知道类器官是如何工作的就可以使用它。
 

According to the researchers’ work published in December, combining organoids and electronic chips could increase the speed and efficiency of AI in the future. Also, such models can be used to study human brains, according to Nature. Using Brainoware to model and study neurological disorders (神经系统疾病), such as Alzheimer’s disease, is one example. It could also be used to test the effects of different treatments for such diseases. “That’s where the promise is: using these to one day hopefully replace animal models of the brain,” Arti Ahluwalia, a researcher in Italy, told Nature.

根据研究人员去年12月发表的研究成果,将类器官和电子芯片相结合可以提高未来人工智能的速度和效率。《自然》杂志称,此类模型还可用于研究人脑。人们可以使用Brainoware来建模和研究神经系统疾病,例如阿尔茨海默症。它还可以用来测试不同治疗方法对此类疾病的治疗效果。意大利研究人员Arti Ahluwalia告诉《自然》杂志:“这就是希望所在——未来或许能用这些技术取代动物的大脑模型。”

 

以上文章内容选自《21世纪英文报》高二932期

分享到


联系我们  |  诚聘英才  |  演讲比赛  |  关于我们
© i21st.cn   京ICP备13028878号-12