Two Oxford University researchers, Carl Benedikt Frey and Michael Osborne, who have been studying artificial intelligence (AI) for a decade, warn that those working from home will be the first to lose their jobs to AI. Outsourced work is easier for AI to replicate than more varied activities in an office environment where face-to-face interactions are an important part of the work process.
Subscribe now for unlimited access.
$0/
(min cost $0)
or signup to continue reading
Their working paper to be published in The Brown Journal of World Affairs notes, "The potential scope of automation has expanded in that many virtual social interactions can now be automated." Their assessment may prompt workers fighting against a return to the office to rethink their situation. This will be a particular dilemma for those who have adjusted their lifestyle to working at home or are covertly working two or more jobs.
The emergence of powerful new tools such as ChatGPT (Chat Generative Pre-trained Transformer) has led to fears that job losses on a vast scale are imminent. New AI tools can summarise emails, respond to phone calls, write essays, analyse language, converse, translate, make persuasive arguments, and produce images and videos, as well as offer recommendations and compile research findings.
Frey and Osborne say that in the short-term ChatGPT and other generative bots are most likely to improve productivity in creative jobs, rather than replace workers wholesale. However, the threat of what AI is capable of doing has already caused a significant reaction in the US. Thousands of actors and writers have gone on strike in Hollywood amid fears their roles could be taken over by robots or digital avatars.
Meanwhile, Elon Musk has embarked on an unusual downsizing project to increase work productivity at his organisations. Since he acquired Twitter, now referred to as X, Twitter's software development team has been reduced to 15-20 per cent of its former size. Musk claims that the company is now far more productive, and able to release more features in a year than the previous larger team could achieve in five years.
Musk claims that Twitter, at its core, is not as complex as some might think, drawing a comparison with the software team at Tesla, which consists of only 200 workers, despite dealing with much more complex technical challenges. Twitter, in contrast, had 10 times as many employees.
While Musk is prone to exaggeration and self-aggrandisement, his claim seems to show that an organisation can achieve more with less, and that in future the optimal workforce size for achieving desired outputs may involve fewer human resources. Indeed, it seems that having too many staff can actually reduce productivity.
Over the past 15 years, life has become increasingly comfortable for many "office" workers, especially since the COVID lockdowns. One employee working from home claimed to have put in only 30 hours of genuine work in three months - but was celebrated by management for his productivity. Another went on a 4000-kilometre bike tour, taking Zoom calls by the roadside without arousing suspicions about his work commitment.
But it's not always the workers at fault. In Canberra, an IT architecture developer contracted to Defence told me he was paid $1200 a day - but some weeks had nothing to do while management considered what they wanted the IT system to do, or negated work he had already done by moving the goalposts to meet an unforeseen - but foreseeable - requirement.
Emmanuel Maggiori, an author and software developer, blames top management for being too hesitant to challenge IT employees. Over the years, IT teams have acquired an almost priestly status within many organisations, able to define their own success without proper oversight. He says senior managers have feared falling behind technology trends or betraying their own IT ignorance, resulting in a lack of critical questioning.
READ MORE:
The modern office environment has been a problem as well. Instead of rewarding individual excellence, middle managers have introduced unproductive administrative requirements, such as all staff attending "personal development" workshops, and unnecessarily documenting performance indicators and benchmarks. Distraction from outstanding individual performance has caused the most talented and productive staff to become demoralised and unproductive.
According to office researcher TeamStage, other time wasting at the office includes: checking emails (up to 121 times a day); seven-and-a-half hours a week spent browsing social media; daydreaming in meetings, and; taking social breaks. Employees also spend 50 days a year doing menial, repetitive tasks. Unsurprisingly, Friday is the least productive day of the week, especially in the afternoon.
Another public sector distraction from core business can be "woke" management. In the UK this month, the Health Minister has criticised senior bureaucrats at the poorly performing National Health Service for plans to introduce three new departments for "equality, diversity, and inclusion". They will require 244 new staff with 18 managers on six-figure salaries (in pounds). Critics say these aspects should already be part of a line manager's responsibility.
Elon Musk's experiment with Twitter - reducing its software development team while increasing productivity, challenges long-held assumptions about the productivity of large teams in technology companies and other organisations.
The assessment of the two Oxford researchers and Musk's downsizing results suggest that organisations could be more productive if they embraced AI, downsized their support and non-productive staff - particularly managers, administrators, and HR personnel, and took a good look at how their IT staff and home-based workers are spending their time.
- Clive Williams is a visiting fellow at the ANU's Strategic and Defence Studies Centre and former public sector manager.