LIVE FROM HUAWEI CONNECT 2018, SHANGHAI: Eric Xu, Huawei’s rotating chairman (pictured), believes more conflict will emerge between technology development and society in the initial phases of artificial intelligence (AI) deployments, particularly as it is adopted more widely in the workplace.

“Existing policies, corporate processes and workforces are built around older technologies. The broader social environment is not yet ready for the AI era,” he explained.

AI development is moving from the first phase, characterised by small-scale exploration, to the second phase where machines and humans will often face collision, he said.

As a combination of technologies forming a new general-purpose technology, AI will change every organisation by making them more efficient, Xu noted: “Every one of us needs to ask ourselves: ‘How will AI reshape or disrupt the industry I’m working in?’”.

Xu acknowledged AI will replace humans in certain tasks and reduce costs through automation, forcing organisational structures to change as demand for repetitive, routine tasks declines.

It will be important to focus on areas where AI can add the most value, and he believes technology and society will reinforce each other as the world moves into the third phase, when the social environment will be more conducive to AI development.

Customer engagement
Vishal Sikka, former CEO of Infosys, agreed AI will have a huge impact on every industry, with enterprises starting by deploying applications to streamline labour-intensive processes.

He believes AI can enable companies to revolutionise the customer experience, as they have access to tools allowing them to better understand customer behaviour: “In addition to improving all aspects of customer engagement, AI will impact the back office by simplifying complicated processes and improving the utilisation of assets.”

A major factor restricting AI development, however, is the limited talent pool. Sikka said there are just 300,000 trained machine learning engineers worldwide: “there should a few million”.

Xu also sees the scarcity of data scientists as a significant limitation. Adding to the problem is AI development requires highly skilled experts, but as the technology matures and tasks are automated he is optimistic staff with basic skills will be able to develop AI apps.

With only 4 per cent of enterprises investing in or deploying AI in their businesses, he said Huawei is excited about the opportunities it will open. In the retail sector, for example, about 2 per cent of companies have invested in AI, while 5 per cent of education institutions use the technology to augment the teaching experience.