SOFTBANK Group Corporation

ARTIFICIAL INTELLIGENCE

SoftBank is dedicated to advancing artificial intelligence (AI) research and development. Its robotics division works to embed AI functionalities into robots such as Pepper, launched in 2015 as the world’s first personal robot with emotion.

In 2016, the company acquired UK-based Arm Holdings, which produces semiconductor designs largely used for processors that power smartphones and other AI-enabled and smart devices. Arm’s work in AI also encompasses an AI Platform (aimed to help accelerate advances in AI), AI technology for cloud computing and Internet of Things (IoT)  applications, and AI solutions for intelligent computing, among others. In 2019, Arm published a Trust Manifesto outlining a series of ethical principles of trust in AI systems:

• AI systems should employ state-of-the art security.
• Every effort should be made to eliminate discriminatory bias in designing and developing AI decision systems.
• AI should be capable of explaining itself as much as possible.
• Users of AI have a right to know who is responsible for the consequences of AI decision making
• Human safety must be the primary consideration in the design of any AI system. 

SoftBank is a major investor in AI, in particular through the Softbank Vision Fund. The first Vision Fund was launched in 2019, with a budget of USD$100 billion dedicated to investments in companies focused on AI and other emerging technologies (e.g. IoT, robotics, computational biology, and financial technology). In addition to Softbank’s own investment, the fund was also backed by companies such as Apple and FoxConn and sovereign funds from countries such as Saudi Arabia and the United Arab Emirates. In 2019, Softbank announced the launch of Vision Fund 2, with a projected USD$108 billion value (for the 2019–2029 period) solely dedicated ‘to facilitate the continued acceleration of the AI revolution through investment in market-leading, tech-enabled growth companies’. Due to difficulties in raising the planned USD$108 billion, the exact launch of Vision Fund 2 remains uncertain (as of March 2020).

Softbank has been investing in multiple companies that work with or on AI, such as Uber (which is developing self-driving cars) and WeWork.

In December 2019, SoftBank partnered with the University of Tokyo to establish the Beyond AI Institute, a business-oriented AI research centre dedicated to turning AI research into profitable business ventures. The Institute is also tasked to actively utilise the Collaborative Innovation Partnership (CIP) system that was newly formulated by Japan’s Ministry of Economy, Trade, and Industry (METI), which enables universities and companies to rapidly set up joint ventures. The Institute is expected to ‘contribute to Japan’s AI research and AI-related business to lead Japan’s AI revolution’.

SoftBank Chief Executive Masayoshi Son backs up Japan’s advancements in AI. He is an advocate of AI as a mandatory subject for college entrance exams to overcome the wide gap between the USA and China in this field.

ROBOTICS

SoftBank’s robotics subsidiary (with teams in Tokyo, Shanghai, Paris, San Francisco, and Boston) works to develop robotics solutions ‘that help make people’s lives easier, safer, more connected and more extraordinary’.

The company’s robots are deployed in more than 70 countries around the world and  provide applications in areas such as government services, education, retail, tourism, healthcare, and finance. SoftBank’s most widely known products are Pepper and NAO.  AI-powered Pepper is optimised to engage with people through conversation and a touch screen, and is able to recognise faces and human emotions. The smaller NAO is primarily used as an assistant in education and research.

In 2017, SoftBank acquired robot manufacturer Boston Dynamics from Google’s parent company Alphabet. Boston Dynamics produces a wide range of robots equipped with perception, navigation, and intelligence. Some of the company’s legacy robots were developed in the framework of contracts granted by the US Department of Defense (DoD); one example is BigDog, which was reportedly rejected by the DoD for being ‘too loud’.