By TIM WINKLER
When ChatGPT burst onto the global stage in 2022 and inveigled us with the promise of minimised drudgery and enhanced productivity, many higher education institutions eyed it warily.
Three and a half years later, it is clear that AI tools can save time for tertiary staff in a range of areas, which has caused some to fear that their roles won’t exist in future, and others to be thrilled that their roles might be changed forever for the better.
However there are other questions that require deeper consideration, relating to the impact of AI adoption on workload – and also what happens to workload when time is saved, noting the tendency for institutions to heap more work onto staff if they are perceived to have some spare minutes.
Natalia Tsybuliak and Yana Sychikova from Ukraine’s Berdyansk State Pedagogical University have written a fascinating article questioning what happens to time saved when LLMs relieve staff of a little administrative burden or research grunt work.
“If academia assumes that any efficiency gain benefits the institution rather than the individual researcher, then generative AI will inevitably amplify pressure, regardless of how advanced the tools become,” the authors write.
“That’s why the question: ‘Who owns the time saved?’ is not just a rhetorical one. It’s also an institutional issue.”
Associate Professor Tsybuliak and Professor Sychikova refer to QUT thought leader Adrian Barnett’s work on slowing down the research publication treadmill, to preference quality over quantity, and warn that if universities are structured to focus on publication numbers as proxies for performance, then windows of time freed up by AI may just lock staff into more unrewarding work.
“What is emerging is not merely a technological shift, but a subtle reconfiguration of the political economy of academic labour – one in which efficiency gains risk being institutionally absorbed rather than individually experienced,” the write
“So before you automate any task, automate the boundaries first. Ask the question out loud: ‘Who owns the time saved?’ If you can’t provide an answer, the default will be ‘not you’. In that case, generative AI won’t reduce your workload; it will simply accelerate an already overwhelming system, making it run faster but with smoother sentences.
“The real risk is not that AI will take our time, but that academia will learn how to absorb every minute it frees.”
The article builds on work by Associate Professor Tsybuliak and Professor Sychikova looking at drivers of academic burnout during the Ukrainian war.
Comment:
The challenges in this article will have strong echoes from academic staff working at the coalface across Australia.
The deceptively simple question of who owns time is significant. Surely if the time is freed up during working hours using tools funded by the institution, then the institution owns the hours? But given academic staff are generally fully grown adults and academic freedom connotes some degree of autonomy with time and focus, shouldn’t we expect that staff will use time wisely rather than burn up the extra minutes on another round of Solitaire, or a long decadent lunch?
This issue is significant because performance management and staff time allocations are patently failing in many academic and professional areas; letting down both staff and institutions.
If staff performance metrics were linked to quality or impact rather than empty citation numbers; if accountability was expected and equitably enforced and if role description recruitment and promotion systems were fit for future use, this article may not have relevance. In the current environment, however, it is critically important.
Join the conversation on the future of HE People & Performance at our conference in Melbourne on 22-23 June.