i卡會員

Vinclu, a tech company from Japan, launched Gatebox, a 3D holographic robot, at the end of 2016. The device projects a 3D virtual home assistant, offering natural language dialogues and a range of smart home applications. Gatebox has triggered heated discussions in the marketplace. The 3D anime, formed inside a multi-faceted glass jar, gives a naked-eye view of three dimensions.

The virtual character of Gatebox only lives inside the box. However, if incorporated with AR/VR, the anime could emerge to be life-sized and interact with the user in a realistic manner. By putting on the AR helmet, the user could see the virtual housekeeper standing next to the TV explaining the contents of the show or giving one-on-one lectures as part of an educational program.  

Microsoft is currently working on the application of Cortana (a virtual voice assistant) in its AR helmet HoloLens so as to present Cortana as a person to meet, speak with and answer questions from the user.

Virtual reality started in the gaming market. Japanese gaming developers are currently working on two VR games: VR girlfriends and summertime classes. Players can interact with virtual characters once they have put on VR helmets.

In fact, virtual assistants are digital characters equipped with artificial intelligence. They have access to a massive amount of data and knowledge, are able to communicate with humans with natural language and provide bespoke information and services. The best known virtual assistants in the marketplace are Apple’s Siri, Amazon’s Alexa and Microsoft’s Cortana.

In addition to the expansion of intelligence, the development of virtual assistants mainly depends on the evolution of man-machine interfaces, from the text communications in the early days, to voice communication and interactions with 3D images.

The development can be divided into four generations as follows:

Generation 1: The virtual characters may come with an anime look and communicate with texts. Supported with a back-end archive of corporate documents, the assistants can only provide one answer to users at a time and usually with low accuracy. Clippy in the early versions of Microsoft’s Office is an example.

Generation 2: The virtual characters remain faceless, only capable of responding with texts or speech. They can identify questions from the dialogues with users and reply accurately. Apple’s Siri is an example. This is the best that the voice assistants currently on the market can do.

Generation 3: The assistants have a basic look and are able to provide answers with texts or conversations on the basis of short-term memories. The backend is connected with a wide variety of third-party applications and services, such as taxi ordering, food deliveries, expert tutorials, foreign language learning, legal consultation and medical diagnosis.

Generation 4: The assistants show up as 3D virtual personalities, able to adjust responses and registers according to the difference in the moods and tones of users. They can constantly learn from the dialogues and adjust replies in order to improve correctness. Virtual assistants can take the form of pure voice, AR/VR characters or 3D holograms. 

The AI services in combination with AR/VR include gaming, legal consulting, medical diagnosis, virtual stores, and so on. Once the AR/VR personal assistants are connected with neural networks, users may receive professional services in either a virtual or bricks-and-mortar environment, not from real people but from virtual characters equipped with artificial intelligence. For example, consumers may play ‘Go’ with AlphaGo in the AR/VR setting, consult with IBM’s Ross for legal opinions, discuss penalty avoidance with DoNotPay, listen to IBM’s Watson for an update on their physical condition and diagnosis results, or allow fund management bots to assist in handling their personal finances.

The e-commerce giant in China, Alibaba, has initiated Project Creator by working together with merchants on Taobao to construct the largest VR shopping mall in the world. To this end, Taobao has developed the DIVA technology that allows merchants to take pictures of their merchandise with smartphones to use as the building blocks of 3D models which can be put on their shelves quickly.  

In the future, shoppers may access the 3D VR version of Taobao stores and browse products by wearing VR helmets. Merchants may provide consultation with a 3D virtual image. By that time, the focus for the industry will be on the ways in which consumers interact with AI shop assistants to discuss VR products in the VR world. 

Artificial intelligence has been the talk of the town in the tech field over recent years. Artificial intelligence has been powering ahead and improving the man-machine interactions such as image recognition, voice recognition, semantic analysis and motion sensing technology. Neural network algorithms and other professional domains (such as Go, legal and diagnostics) are also making leaps and bounds.

The technologies, if used in the immersive environment created by AR/VR, can create an excellent user experience. Virtual teachers armed with AI can instruct foreign languages in the VR world, and this may even completely replace the real-world curricula. As an increasing list of apps and services become available in the AR/VR version, the public will engage in more and more economic activities with virtual characters in the virtual world. A brand new market place, the virtual economy, will emerge.

Artificial intelligence is the service provider in the virtual economy, whereas AR/VR renders the virtual world more realistic, natural and intuitive. The virtual economy lowers the costs associated with professional manpower and transportation through AI and AR/VR. This will dramatically reduce service costs and make professional services inexpensive.