在被告之uber近期回收了一间名叫几何图形智能化的新成立公司以后，uber将眼光看向了人工智能技术以加速其无人驾驶方案。Geometric Intelligence是一间由科学研究工作人员在纽约市开创的科研型新成立公司。Uber公布，几何图形智能化的创始人兼CeoGary Marcus将领导干部Uber最新消息单位Uber AI Labs的15人精英团队，该单位坐落于该企业坐落于美国旧金山的总公司。
该企业表达，Uber AI Labs的创立是以便深化推动该企业的“人工智能技术和深度学习科学研究”。Uber顶尖商品官Jeff Holden郑重声明说，本次回收将协助Uber 深化拓展其UberEATS，UberPOOL等细分市场的产品知识。霍尔顿说：“它延伸到教授自动驾驶机器来安全和自主地导航世界，无论是路上的汽车还是穿越繁忙的领空的飞机或新型机器人设备。”
uber现阶段在美国匹兹堡和美国旧金山的大城市有着测试车，用以纪录其运输队学习规划的很多统计数据，这种统计数据最后将进到公共性路面优化算法，或许要等候监督机构的准许。
Uber回收Geometric Intelligence的将会缘故其一是，该精英团队准备解决说白了的“深度神经网络”，即从实质上为开发设计中的系统软件出示很多统计数据。
几何图形智能化的创办人，如今是Uber AI试验室的责任人，加里·马库斯(Gary Marcus)表述说，他的企业选用了这种不一样的学习的方法，该实体模型应用的实体模型在语言学习专业技能时候从小孩的脑部发育中吸取设计灵感。她说：
“深度神经网络并非它愿意破译的所有。这对一些难题十分有效，但它不容许人们开展大家常常开展的逻辑推理。人们必须下代技术性。当您解决例如安全驾驶等实际不良影响的难题时，您期待尽量地贴近100%。”加里·马库斯(Gary Marcus)还预料了四轴飞行器将协助出行者防止出远门时的交通堵塞的将来。他在接纳英国广播公司(BBC)访谈表示：“她们将容许大家以75或150公里/钟头的速率开展远途出行，不然您将深陷交通阻塞。在没多久的未来，这的确是行之有效的。”
Uber回收成本费的关键点尚未公布。
英文版
English version
After the defendant Uber recently reclaimed a new company called Geometry Intelligence, Uber looked at artificial intelligence technology to accelerate its driverless program. Geometric Intelligence is a research-based, newly established company pioneered by scientific researchers in New York City. Uber announced that the founder of Geometry and CeoGary Marcus will lead the 15-person elite team of Uber AI Labs, the company's latest news unit, located in the company's head office in San Francisco.
The company expressed that Uber AI Labs was created to deepen the "artificial intelligence technology and deep learning science research" that promoted the company. Uber's top merchandiser Jeff Holden solemnly stated that this recycling will help Uber deepen its product knowledge in UberEATS, UberPOOL and other market segments. Holden said: "It extends to teaching autonomous driving machines to navigate the world safely and autonomously, whether it's a car on the road or a plane or a new robotic device that travels through a busy airspace."
At this stage, Uber has test vehicles in Pittsburgh and San Francisco, USA, to record many statistics of its transportation team's learning plan. This statistical data will eventually enter the public road optimization algorithm, perhaps waiting for the approval of the supervisory agency. .
Uber's recycling of Geometric Intelligence will be due to the fact that the elite team is ready to solve the confusing "deep neural network", which is to present a lot of statistics from the system software in the development of the design.
The founder of geometric intelligence is now the owner of the Uber AI lab, and Gary Marcus stated that his company chose this different approach to learning, the entity to which the entity model is applied. The model draws inspiration from the child's brain development as the language learns professional skills. she says:
"Deep neural networks are not all that they are willing to decipher. This is very effective for some problems, but it does not allow people to carry out the logical reasoning that people often carry out. People must be technical next generation. When you solve the problem of practical adverse effects such as safe driving, You expect to be as close as possible to 100%.” Gary Marcus also anticipates that the quadcopter will assist travellers in preventing the future of traffic jams when they are far away. In an interview with the BBC, he said: "They will allow everyone to travel at a speed of 75 or 150 km / h, or you will be trapped in traffic jams. In the near future, this is indeed the case. Effective."
The key points of the Uber recycling cost have not been announced.
（图/文编译：飞行汽车 http://www.flycar.com.cn/）
声明：未经原创作者飞行汽车网允许不得转载本文内容，否则将视为侵权，本站依法保留追究权！
