"Today we talk to machines in a mechanical way but the future is different. The user interface will be much more natural and will be able to improve the lives of many more people," said Miffy Chen, head of Alibaba's artificial intelligence labs in a plenary lecture held at the CES Asia 2018 conference held in Shanghai in china
"The goal of the division I head is to bring humanity to machines," said Miffi Chen, head of Alibaba's artificial intelligence labs. Chen spoke at a plenary lecture held at CES Asia 2018 in Shanghai, China.
"Today we talk to machines in a more or less mechanical way but the future is different. The user interface will be much more natural and will be able to improve the lives of many more people in remote places. It seems that in the near future innovation will cross the boundaries of functionality towards a design that will encourage user involvement using standardization that will facilitate the establishment of an entire ecosystem," she noted. According to her, "At the end of the process, all our lives will change. We need a technology that will unite everything under one interface, Mash Technology, and end the era of chaos in which we are today."
Chen pointed out that "there are already smart televisions today, but to talk to them you have to separate the voices and talk to the machine, like with crazy people. The adults remember the time in the 80s, when only computer engineers could operate a computer. Then Bill Gates came along, invented Windows and added a mouse to the keyboard, which already existed."
The Tmall Genie digital assistant is being upgraded
"For this purpose, we developed the Tmall Genie digital assistant. It is a machine that understands its surroundings and communicates with us with the help of voice in a fluent language. Now it also has the ability to recognize images thanks to our central systems that it is linked to in the cloud. Its camera is being trained and it has never failed to recognize and I speak About 63 million units. The Chinese market loves innovation and we feel it," she added.
"Anything that can be smart will be upgraded to have a voice interface," Chen said. There are currently over 200 IoT platforms and over 5,000 devices. We cannot cover them all, so we go with the open approach. An important element in utilizing the capabilities of IoT devices is their ability not only to receive the sensor data but also to analyze it and respond accordingly. All these types of machines will be linked together through a Lowtech interface, which will allow humans to use it. It is not enough for the devices to be connected to each other, it will be possible to establish ad hoc networks. In order to accelerate the development of the field, we started a first cooperation of its kind in which we open the API and companies can integrate it into their products."
Tmall Genie is based on a voice processing chip that Chen says a recent version is in mass production and uses the AliGenie voice assistant developed in Alibaba's AI labs. Tmall Genie allows users to control using voice commands. The system is equipped with voiceprint recognition technology that will allow it to better understand the commands, and thus allow not only to buy from Alibaba's shopping sites but also to pay using the device, through the AliPay service. Tmall Genie also provides connectivity to devices and Alibaba's Tmall Box smart home platform.
Soon you will be able to know the status of the engine using voice commands
About a month ago it was announced that Tmall Genie will be embedded in luxury cars such as BMW and Volvo, which will be sold in the Chinese market in the near future. The system will allow drivers to check the fuel status, the mileage. Activity level and engine status via voice commands. Besides diagnosing the condition of the vehicle, it will also allow controlling the doors, windows, air conditioning and other systems.
In the enclosed article, it should be mentioned that the system only understands the Chinese language, and there are two reasons for this, first - adaptation to the Chinese domestic market, but the more important reason is that the Chinese government is trying to avoid introducing foreign technologies into the Chinese market.
In conclusion, Chen said that "it is necessary to improve the voice interface with the machines so that it is more human, or in other words, to introduce humanity into the machines."
2 תגובות
It is beautiful and blessed that machines are becoming more sophisticated and more useful for human use.
But the statement "to bring humanity to machines" is problematic. Either it reflects a distortion in the understanding of what a machine is, or it reflects a distortion in the understanding of what humanity is, or it reflects a distortion in the understanding of both machine and humanity together. In the first case it is bad, in the second case it is worse, in the third case it is a disaster. Machines are not people, and mechanical cannot be human. At the same time, people are not machines, and humanity cannot be mechanical. The machine, however sophisticated it may be, is not able to completely deviate from the algorithm or its products. Humanity is not capable of going through a complete reduction and completely converging into an algorithm, not even close to it, and it must not come to that either.
This is true even in civilizations of the authoritarian variety, asked one of the two main ones to which the speaker belongs. Especially in the civilizations of the West, the beliefs about free will, personal autonomy, freedom of choice, personal and community morality, democracy and beliefs that have a dimension of holiness of some kind.
Bees of the type described in the article are simply a danger, because discourse of any kind, certainly when it is managed by people with cultural (including technological) authority and economic power, has an enormous formative - cultural and social - power, capable of damaging humanity in a decisive and irreversible way.
Everyone deals with image, audio, text processing. Professor Naftali Tashvi from Hebrew, and Amnon Shashua from Hebrew published innovative articles on a global scale in the field.
Naftali in 99-2017 developed a thermodynamic theory of machine learning, and Shashua released a theory of machine learning analysis using quantum mechanics. Not using a quantum computer. It was surprising that Shashua found time to publish 2 such heavy articles while working at Mobileye
In the world, David Singer is a leading scientist at Google Deepmind on game theory in machine learning, he brings out the most scientifically interesting news in the field.