Dive Brief:
- Workers on average spend an hour and a half talking to or using artificial intelligence daily, adding up to about 47 days per year, according to a recent BambooHR survey. But nearly half of that time — equivalent to about 20 days per year — is spent on troubleshooting AI errors or on prompt iteration.
- Nonetheless, 65% of employees said they felt confident and enthusiastic about using the technology at work, and 58% said they were motivated by time savings gained from AI. BambooHR’s survey, which included more than 1,600 full-time, salaried U.S. workers, also found that those at the vice president and C-suite levels spent nearly twice as much time with AI compared to individual contributors.
- As AI takes up more of workers’ time, it could leave less room for human interaction and collaboration. Respondents spent an average of 41% of their time with AI versus 59% of their time talking to people. Mentorship is particularly vulnerable to this trend, BambooHR noted, and more than one-third of workers said that knowledge transfers within their organizations primarily happen through AI tools.
Dive Insight:
Faulty or low-quality AI outputs are so prevalent that they have been given their own moniker: “workslop.” Though BambooHR found that employees spend a great deal of time fixing such outputs, other research suggests workers may be increasingly inclined to forgo this work entirely.
Survey data published last February by Zety found that only 39% of workers said workslop was completely unacceptable and corrected, whereas 31% said it was unacceptable but tolerated. An additional 21% said workslop is somewhat acceptable and overlooked so long as deadlines are met. Zety’s survey estimated that most workers spent up to six hours or more per week correcting workslop, which closely tracks with BambooHR’s estimate that employees spent about an hour and a half each day on average troubleshooting AI.
The findings demonstrate that, as AI advances, organizations are still lagging in terms of their ability to assess how it can best be deployed. Talent management firm Talogy said in an August report that the tech has progressed at a rate that effectively outpaces workers’ capacity to determine its long-term value.
In HR alone, 4 in 5 professionals queried by Paylocity said they were actively managing at least one issue with their AI tools, even as 43% said AI saved their recruiting teams the equivalent of one working day each week.
BambooHR’s findings on the use of AI for workplace knowledge also echo previous findings. In May, Workday published research in which it found that 33% of employees said they rarely or never talked to co-workers during a given week apart from task-related conversations. Conversely, 76% had used AI tools for advice and 52% used them for brainstorming.
Workday said respondents primarily found AI to be less judgmental and more available than human co-workers. BambooHR found that employees may actively avoid getting help from a co-worker due to a “mix of self-sufficiency and self-consciousness,” noting that 27% of respondents said they would rather ask AI than admit needing help.
As workplaces continue to hash out AI’s many effects on workers and work itself, managers are likely to be critical to building employee confidence and reducing anxiety about the tech. But CHROs are split on whether managers are ready to manage through AI, according to a recent poll of Gallup’s CHRO Roundtable.
