中國(guó)科學(xué)家使用創(chuàng)新芯片為自動(dòng)自行車提供動(dòng)力

中國(guó)科學(xué)家使用創(chuàng)新芯片為自動(dòng)自行車提供動(dòng)力

While tech giants like Tesla and Google are working hard to perfect autonomous cars, a team of scientists in China is developing a self-driving bicycle and their results are pretty impressive.

當(dāng)特斯拉(Tesla)和谷歌等科技巨頭正在努力完善自動(dòng)駕駛汽車時(shí)，中國(guó)的一個(gè)科學(xué)家團(tuán)隊(duì)卻在開發(fā)一款自動(dòng)駕駛自行車，他們的成果令人印象深刻。

Showcased in a short video presentation, the autonomous bicycle developed by scientists at Tsinghua University, in Beijing, can balance itself while also avoiding obstacles, following certain trajectories and obeying vocal commands. It can do all these things thanks to a revolutionary type of computer chip that combines two different architectural approaches to computing that hadn’t really worked well together until now, due to communication difficulties between the two systems. However, if this self-driving bike is any indication, the Chinese seem to have solved those issues.

在一段簡(jiǎn)短的視頻演示中，清華大學(xué)的科學(xué)家們研發(fā)的這款自動(dòng)自行車能夠自我平衡，同時(shí)還能避開障礙物，沿著特定的軌跡行駛，并服從聲音指令。多虧了一種革命性的計(jì)算機(jī)芯片，它將兩種不同的計(jì)算體系結(jié)構(gòu)方法結(jié)合在一起，而由于兩種系統(tǒng)之間的通信困難，這兩種方法直到現(xiàn)在還不能很好地協(xié)同工作。然而，如果這款自動(dòng)駕駛自行車是一個(gè)跡象，那么中國(guó)似乎已經(jīng)解決了這些問題。

This autonomous bicycle is not the first of its kind – A US University has a similar project underway – but the computer chip powering it certainly is. It combines computer science-based machine learning and human brain-inspired networks, allowing it to perform several actions independently at the same time. Because the new chip, called Tianjic, allows for easy communication between its network, it can balance the bicycle, while also avoiding obstacles in its path and follow a number of verbal commands.

這款自動(dòng)自行車并不是同類產(chǎn)品中的第一款——美國(guó)一所大學(xué)也在進(jìn)行類似的項(xiàng)目——但驅(qū)動(dòng)它的電腦芯片肯定是第一款。它結(jié)合了基于計(jì)算機(jī)科學(xué)的機(jī)器學(xué)習(xí)和人類大腦啟發(fā)的網(wǎng)絡(luò)，允許它在同一時(shí)間獨(dú)立地執(zhí)行幾個(gè)動(dòng)作。由于這款名為“天啟”的新芯片可以方便地在網(wǎng)絡(luò)之間進(jìn)行通訊，它可以平衡自行車，同時(shí)還能避開道路上的障礙物，并遵循一些口頭指令。

“Using just one chip, we can demonstrate the simultaneous processing of versatile algorithms and models in an unmanned bicycle system, realizing real-time object detection, tracking, voice control, obstacle avoidance and balance control,” Chinese scientists from the Center for Brain Inspired Computing Research at Tsinghua University wrote in a paper. “Our study is expected to stimulate AGI development by paving the way for more generalized hardware platforms.”

清華大學(xué)腦智能計(jì)算研究中心的中國(guó)科學(xué)家在一篇論文中寫道:“僅用一個(gè)芯片，我們就可以在無人駕駛自行車系統(tǒng)中演示多功能算法和模型的同時(shí)處理，實(shí)現(xiàn)實(shí)時(shí)目標(biāo)檢測(cè)、跟蹤、語音控制、避障和平衡控制。”“我們的研究有望為更通用的硬件平臺(tái)鋪平道路，從而刺激AGI的發(fā)展。”

While claims of accelerated Artificial General Intelligence (AGI) – a term used to describe artificial intelligence on a par with human intelligence – are too bold to be taken seriously, the new Tianjic chip is proof of the strides Chinese scientists have made in the field of artificial intelligence. They may not yet have the most advanced AI chips in the world, but if their latest achievements – like those eerie AI news anchors – are any indication, they are getting there and fast.

盡管有關(guān)加速人工智能(AGI)的說法過于大膽，不值得認(rèn)真對(duì)待，但新的天基芯片證明了中國(guó)科學(xué)家在人工智能領(lǐng)域取得的進(jìn)展。AGI是一個(gè)用來形容人工智能與人類智能水平相當(dāng)?shù)男g(shù)語。他們可能還沒有世界上最先進(jìn)的人工智能芯片，但如果他們的最新成就——比如那些令人毛骨悚然的人工智能新聞主播——能夠說明問題的話，那么他們正在迅速取得進(jìn)展。