Imagine a future of work where, instead of spending hours on tedious and time consuming tasks, you could focus your time and energy on the analytical and creative aspects of your job.

When your predecessors were toiling over spreadsheets to organize reports, you can now spend that time fine-tuning strategies and padding those same reports with better results. Allowing algorithms to take over mundane and repetitive tasks so that the human worker can do the heavy mental lifting of innovation and imagination will soon become a reality across knowledge-based industries.

Process Automation is a term used to describe robotic automated systems that can do tedious business tasks through artificial intelligence (AI) so that employees can focus on more cerebral work that generates more value for the company. According to School of Information Studies professor Kevin Crowston, this could resemble a system that automatically populates large amounts of data into a spreadsheet, or a bot that automatically downloads resumes from a posting and places them in the corresponding files for future review – by a qualified human.

“Artificial intelligence is driven by machine learning, and is driven closely by interacting with the person doing the task,” said Crowston. “Machine learning is already a big part of work, there are so many cases where a task needs to be done quickly and repetitively, and AI can be built in when it doesn’t make sense to have a human do the work.” These machine learning algorithms are already being used in most people’s everyday life and throughout workplaces.

Credit card companies are currently using AI to detect fraud by implementing computer systems that can automatically identify suspicious activity. Features on Google Docs like spell check and predictive typing on our cell phones are also examples of automated processes that most people use on a daily basis. If your thoughts flow more freely into your typing or texting because the device is assisting you as you go, you understand what this kind of technology has to offer.

Crowston himself has secured an NSF grant to study how this type of process automation can be used to help journalists write stories. His team believes that there are a subset of journalistic stories that can be written automatically by bots so that the journalists can focus on more investigative/deeper stories. For example: a bot, after being properly programmed and trained, could write simple and informative summaries using clear data sets, such as earnings or stock market reports.

“Journalism is an information-based task that has always been centered around technology,” he said. “We’ve already seen it transform since journalists can now use their cell phones to report on social media in real time, so the process of how information is reported has completely changed.”

This study will also look at how process automation can be used to aid journalists in the information gathering process. AI could analyze a large number of emails to find patterns and piece together a timeline automatically, so the journalists can focus more on the analytical aspects of fact-checking and weaving together the real story. Crowston also suggests that this type of technology could be used in journalism right now. As public health has become a central story during the current pandemic, process automation can be employed to help trace public health documents as a tool for journalists to establish trends and piece together a narrative.

Artificial intelligence is driven by machine learning, and is driven closely by interacting with the person doing the task.

“We still have a lot of questions about the implications and importance of this research that we’re hoping to answer,” said Crowston. “Will journalists actually find this technology useful? Once the system identifies a story, will that information actually help the writer compose the story? Will journalists even be willing to use this technology?”

Crowston also noted that there are many unknown outcomes of process automation due to the lack of research regarding what happens when the machine makes a mistake. “It’s often unclear,” he said, “how machine-made mistakes will be identified and corrected, and where the responsibility lies when it does happen.” Despite these unknowns, Crowston is excited about how process automation will benefit the workforces of both the present and future. An important takeaway from Crowston’s approach is that the goal is not to replace workers with robots, rather it is to complement and supplement the worker, to free them from the rituals of mindless procedure, and empower them to participate more deeply towards the greater goal.

“We can use process automation to eliminate boring, routine work so that employees can focus on more interesting analytical tasks,” he said. “We’ll also need people to build and monitor these systems, so while some jobs may be replaced by machines, we’ll see a ton of job growth for the individuals who create these automated processes.”