Digital business is disrupting organisations with its requirements for artificial intelligence (AI). Results from Gartner, Inc.'s 2017 global survey of more than 3,000 CIOs shows that a wide range of companies have begun AI initiatives. Twenty-five per cent of organisations have deployed AI or are making short-term plans for its deployment. This is up from 10 per cent in 2015.
Gartner analysts presented on the outlook for AI during Gartner Symposium/ITxpo, which took place until Thursday 9th October
AI projects, like any others, may under deliver, stall or fail if businesses ignore AI. However, there is a high risk of becoming uncompetitive, or even obsolete. "Now is the time for experimentation and proofs of concept," said Magnus Revang, research director at Gartner. "There are no best practices, only emerging practices."
What CIOs need most is practical experience. "They must also evaluate and understand vendors’ claims that are inflating expectations of what AI is capable of doing," added Mr Revang.
To succeed with AI, CIOs need to influence its long-term direction. To do that, they should consider three issues in turn: how AI could augment their company's existing services; how they could change their company's existing services to suit AI; and what new services they could create that would be impossible without AI.
Some organisations have already had some success using AI to enhance their existing services. For example, a patent office in Australia uses AI to classify all its patents, to ensure they go to the right specialists for review. Its AI solution, which was made by graduate students and uses readily available cloud-hosted services, not only saves on personnel, but significantly shortens the time taken to process patents. "This example demonstrates how AI can have a high impact even at relatively low levels of investment, if applied to the right cases with the right skill set," said Mr Revang. "In fact, in this case the biggest hurdles were the integration architecture and security policies, not the AI technology."
Other organisations have created new services that were made possible only by the use of AI. "To do this, these organisations first embraced the fact that every company is an IT company, and understood that they needed to become not only a software company, but a hardware company," said Mr Revang. One example is an office supply company that uses an intelligent, voice-driven interface to reorder office supplies with minimal human interaction.
"Overall, AI helps organisations relieve some of the burden on the back office. It also helps them power the front office with applications such as virtual personal assistants that respond positively to the increasing needs of customers and free up customer support staff to provide higher-level support for more complex needs and requests," said Mr Revang.
By 2022, Gartner predicts that 72 per cent of customer interactions will involve emerging technologies such as machine-learning applications, chatbots and mobile messaging, up from 11 per cent in 2017.
2020 Will Be Pivotal for the Creation of AI-Related Jobs
Gartner forecasts that AI will lead to $200 billion in new revenue and 768,000 new positions in 2018. While in most cases, savings and efficiencies resulting from the use of AI will improve productivity, 943,000 jobs will be eliminated in the same year.
"However, there will be a transition period through to 2020, when AI will create 2.3 million jobs and eliminate only 1.8 million jobs," said Helen Poitevin, research director at Gartner. Moderately skilled occupations, for which training is received "on the job," will bear the brunt of the job losses. AI will, however, create millions of new highly skilled positions, managerial positions, and even entry-level and low-skilled positions.
Already, chatbots point to likely developments. For example, they will help recruiters screen high volumes of applications, and remove much of these people's mundane work. Ultimately, it seems likely that AI will take over the task of creating and posting job vacancies to attract the right calibre of candidate. "This will give recruiters more time to build relationships and monitor the effectiveness of chatbots and other AI applications that help find the most appropriate candidates," Ms Poitevin added. "The tasks performed by human recruiters will mainly be those that algorithms cannot perform."
Success Relies on the Symbiosis of People and AI
The cooperation of human capabilities and AI is most apparent in healthcare and retail.
In the healthcare sector, AI excels at using very large amounts of data which could decrease the cost of diagnostics and increase the demand in the market for this kind of service. As a result, AI would create new jobs in the sector.
In the retail sector, an online shopping company exemplifies human-AI symbiosis. It uses AI to narrow its huge selection of clothes to suit customers' preferences, but has human designers make the final choices. The company has created 65 data scientist positions and retains thousands of designers.
By undertaking routine tasks, AI gives employees more time to be productive and creative by doing what humans do best. "To foster and profit from a symbiotic culture of people and AI, business and IT leaders must invest in reinforcing agility and 'soft' skills, and accept variability in skill sets, as bots, the Internet of Things and AI will require a larger percentage of data scientists than programmers," said Ms Poitevin. “We predict that, through to the end of 2020, the number of citizen data scientists* will grow five times faster than the number of data scientists.”
"Now is the time to influence the long-term direction of AI, as we are at a pivotal point. Enrich people's jobs, reimagine old tasks, design new occupations and create new industries. AI can handle patterns it has already seen, but it's humans who break new ground," said Ms Poitevin.
Gartner clients can learn more in the report "Top Strategic Predictions for 2018 and Beyond: Pace Yourself, for Sanity's Sake" and “A Framework for Applying AI in the Enterprise.”
Press release by Gartner
Publié le 14 novembre 2017