Ren Zhengfei, the retired founder and CEO of Huawei, the technology giant in China, is challenging about US efforts to hinder his company with lawsuits and restrictions.
“There is no way the US can crush us,” Ren said in a rare recent interview with international media. “The world cannot leave us because we are more advanced.”
It may sound like a bluff and bluster, but these words carry a certain amount of truth. Huawei’s technological road map, especially in the field of artificial intelligence, points to a company that is making faster progress – and more technology fronts – than any other company in the world. Aside from its AI ambitions, Huawei is a rising player in the market of the next generation 5G wireless networks, as well as the world’s second largest smartphone manufacturer behind Samsung (and ahead of Apple).
Sign up for The algorithm – artificial intelligence, demystified
“The (Chinese) approach of the government and the private sector is to build businesses that compete in the full technology stack,” said Samm Sacks, who specializes in cyber security and China in New America, a Washington think tank. “That is what Huawei does.”
But it is Huawei’s AI strategy that will give it a truly unparalleled reach in the entire technical landscape. It will also cause a large number of new security issues. The company’s technological omnipresence and the fact that Chinese companies are ultimately accountable to their government are major reasons why the US regards Huawei as an unprecedented threat to national security.
In an exclusive interview with MIT Technology Review, Xu Wenwei, director of the board of Huawei and the company’s most important strategy and marketing officer, praised the scope of his AI plans. He also defended the company’s record in the field of security. And he promised that Huawei would try to make contact with the rest of the world to address the emerging risks and threats of AI.
Xu (using the Western name William Xu) said Huawei plans to increase its investment in AI and integrate it throughout the company to “build a full stack of AI portfolio.” Because Huawei is a private company, it is difficult to quantify technology investments. But company officials said last year it was planning to more than double annual R&D spending to between $ 15 billion and $ 20 billion. This could catapult the company to fifth and second place in global R&D spending. According to the website, around 80,000 employees, or 45% of Huawei’s staff, are involved in R&D.
Huawei’s vision extends from AI chips for data centers and mobile devices to in-depth software and cloud services that offer an alternative to those from Amazon, Microsoft or Google. The company examines major technical challenges, including learning machine data models more and more energy efficient and easier to update, Xu said.
But Huawei is struggling to convince the western world that it can be trusted. The company faces allegations of intellectual property theft, espionage and fraud, and its vice-chairman and CFO (and Ren’s daughter), Meng Wanzhou, is currently under house arrest in Canada awaiting possible extradition to the US. America and several other countries have banned or are considering restrictions on the sale of Huawei’s devices, citing concerns that the Chinese government might exploit Huawei’s 5G equipment to attack systems or absorb sensitive data.
Xu defended the company’s reputation: “Huawei’s security reputation is clean.”
But AI adds another dimension to such concerns. Machine learning services are a new source of risk because they can be exploited by hackers and the data used to train such services may contain private information. The use of AI algorithms also makes systems more complex and opaque, which means that security audits pose a greater challenge.
As part of an attempt to reassure doubters, Xu promised that Huawei would release a code with AI principles in April. This comes down to a promise that the company will try to protect user data and ensure security. Xu also said Huawei wants to work with its international competitors, including Google and Amazon, to ensure that the technology is developed responsibly. However, it is unclear whether Huawei allows its AI services to be controlled by a third party, such as with its hardware.
“Many companies in the industry, including Huawei, are developing AI principles,” Xu told MIT Technology Review. “We now know three things for sure: technology must be safe and transparent; user privacy and rights must be protected; and AI should promote the development of social equality and well-being. “
Pile up
As Huawei progresses in AI and develops towards a ‘full-stack’ company, it seems to be increasingly powerful for many in the West.
It already offers a staggering offer. Last year, Huawei launched an AI chip for its smartphones, called Ascend, which is similar to a chip from the latest iPhones and is tailor-made for executing machine learning code that enables tasks such as face and speech recognition. The technology for the chip came from a startup called Cambricon, which was spun from the Chinese Academy of Sciences, but Huawei recently said it would design future generations in-house.
Huawei also sells a range of AI optimized chips for desktops, servers and data centers. The chips lag behind those of Nvidia and Qualcomm (both US companies) in the area of refinement, but no other company can boast such a range of AI hardware.
Then there is the software. Huawei offers a cloud computing platform with 45 different AI services – comparable in terms of offerings from Western giants such as Google, Amazon and Microsoft. In the second quarter of 2019, Huawei will also release its first deep-learning framework, called MindSpore, which will compete with Google’s Tensorflow or Facebook’s PyTorch.
AI is also interwoven with Huawei’s ambitions to deliver the 5G equipment that connects everything from industrial machines to self-driving cars. “We need to use AI to reduce maintenance costs,” Xu said. “Telecom networks are becoming increasingly complex – 70% of network errors are caused by human error, and if we use AI for network maintenance, more than 50% of potential errors can be predicted.”
Standard bearers
Xu’s statements about AI ethics are in a way also part of an attempt to lead the AI development of the world. Ensuring ethical AI means drafting technical standards, which will be important for shaping the future of technology itself. The United States has had a major influence on the development of the internet through technical standards.
To this end, the Chinese Association for Artificial Intelligence, a state-run organization, set up a committee earlier this year to prepare a national code for AI ethics. Several major Chinese technology companies, including Baidu, Alibaba and Tencent, also have initiatives to understand the impact of AI.
Agreeing on AI ethics and standards can be challenging, as tensions between East and West escalate. A number of national governments, as well as organizations such as the EU, are also trying to determine the rules of the road. “AI brings value, as well as problems and confusion,” Xu told MIT Technology Review. “Global collaboration is needed to address these issues.”
And international cooperation is now not exactly a US forte. Indeed, outside its own borders, the US government can hinder Huawei so much. Some allies are apparently tired of strong US tactics; the UK and Germany both seem less and less likely to ban Huawei from providing 5G equipment and other products and services.
The importance of the company to deal with circumspect countries also has its limits. In recent comments, his CEO, Ren, argued that the international picture is changing, at least on a technological level. “If the lights go out in the west, the east will still shine,” he said. “And when the north gets dark, there is still the south. America does not represent the world. America represents only a part of the world. “
Anyway, there will be Huawei.
