Many industry experts have made predictions about the future of artificial intelligence (AI), machine learning, and automation in the workplace–but little research exists on how skilled and unskilled employees view the coming technology disruptions.

Gartner surveyed more than 2,700 employees in the US and the UK across multiple industries and skill levels, including unskilled workers, skilled manual workers, clerical workers, and professionals, to determine how different groups view AI in the workplace.

The majority of employees (52%) said they would prefer AI to be deployed as an on-demand helper–essentially, acting as their own employee–rather than as their manager (9%), coworker (11%), or proactive assistant (32%), the report found. In this role, AI could help by limiting their routine work tasks, as well as reducing mistakes, respondents said.

SEE: IT leader’s guide to the future of artificial intelligence (Tech Pro Research)

Workers differ significantly in their opinions about how AI will impact jobs, the report found. Improving safety and quality of work, the pace of work, and how interesting the work is were the areas where many said they believed AI could help. However, workers held more negative views toward AI’s role in customer contact, coworker contact, and job security.

Most workers said that their jobs were fairly low in terms of structure and repeatability, implying that certain tasks could be automated, but the entire job would not be able to be replaced by a machine, the report found. Many workers also reported that their role required empathy and other special skills that AI currently struggles with.

AI’s capabilities will increase steadily, most workers predicted, to the point that the technology will be able to replace twice as many jobs in 10 years as it is currently–eventually leading to more than four out of 10 jobs being replaced by AI.

Workers in communications, energy, and insurance industries had higher relative acceptance and ideas of how AI could help them, which means those industries may have the broadest opportunity for the technology, the report found. The industries that were the most pessimistic about AI were banking, energy and utilities, retail, and transportation.

The report offered the following tips for technology managers in charge of AI products and services who want to improve their go-to-market effectiveness:

1. Prioritize assistive AI over AI meant to replace entire job functions: Offer workers examples of assistive AI, such as natural language queries or virtual personal assistants.

2. Improve adoption and engagement by sending buyers messaging that makes workers more optimistic about how AI could improve their jobs: Describe how you can improve the safety and quality of work and make tasks more interesting, and address and calm fears by emphasizing that workers will have more time to interact with customers and coworkers.

3. Prioritize AI efforts in industries where workers are relatively accepting and see many ways AI could help them: These include communications, energy, and insurance.

The big takeaways for tech leaders:

  • When it comes to adopting AI, the majority of employees (52%) prefer AI to be deployed as an on-demand helper rather than as their manager (9%), coworker (11%), or proactive assistant (32%). — Gartner, 2018
  • Workers in communications, energy, and insurance are most optimistic about AI, and those in banking, energy and utilities, retail, and transportation are most pessimistic. — Gartner, 2018