What Will Office Workers Do In The Future If AI Continues Getting Better?

Recently, the IMF and other think tanks have been predicting the end of the office worker. White-collar professionals will apparently soon see their jobs taken over by machines. 

The tales of doom, though, are probably overrated. Even if AI gets to the point where it is cognitively equivalent to a human worker, it may still only perform well in certain domains. It could take decades for machines to have the same performance scope in the workplace as people. 

So what does that mean for people who are currently employed in office jobs? The answer isn’t totally clear. However, we can use reason and logic to peer into the future and look at what might happen to this segment of the economy if AI continues to get better. 

People Will Partner With AI, Not Compete With It

The idea that people will compete with AI is a misnomer. Like any machine, AI does some things better than employees and other things worse. 

Because of this, the most likely outcome is that people will partner with artificial intelligence systems. It will become a tool workers can use to get things done faster, just as the internet and computers help workers today.

This notion comes back to the idea that most technology is complementary. That is, it works best in combination with people, not without them. 

That’s been true of technology for most of industrial history. It works best when it combines with workers to produce an output. 

Proponents of AI say that artificial intelligence changes this paradigm. Whereas phones and computers need people to perform at their best, AI systems with full agency don’t. 

The counterargument is that AI’s agency remains limited. While some systems can direct their activities, it is not yet something that they can do universally. 

Even companies that believe they have systems that can do the work of people will still likely keep conventional employees on-site to supervise. Many bosses won’t be willing to allow untested machines to get on with work unless they feel confident that they can operate independently for months or years on end. 

The first autonomous AI systems to enter the proverbial office will likely make mistakes or require directions, even if we assume an exponential increase in the rate of progress. That means that office managers will need to hedge their technology bets. Even if vendors claim systems will work unsupervised and deliver desired outcomes, companies will still need trained workers to step in if there are problems or unexpected bugs and errors. 

Workers Will Become AI Specialists

Image credit: Mario Gogh on Unsplash

Following on from this trend, it is also likely that we will see many workers become AI specialists. People will still be required to build AI systems

While some vendors might achieve full autonomy, it is unlikely that these systems will then be able to go on and fill in the remaining gaps left by non-AI tools in the regular economy. 

Already, we are seeing this pan out in the vendor space. For example, when document processing technology company Apryse acquires LeadTools and brings it into its team, it will give it the ability to leverage artificial intelligence technologies. 

“The fact that LEAD Technologies is now a part of the firm means that we have unprecedented opportunities at the intersection of AI and document-focused applications,” according to Cassidy Smirnow, Apryse’s CEO. “Bringing in this development team will assist with all aspects of our business and transform our capacity to partner with businesses to accelerate their transition towards the new digital economy.”

Interestingly, Apryse’s acquisition actually brought 85 new employees into two new offices. In other words, the company remains a net supporter of employment in the office sector, primarily because of the demands of AI and the paucity of talent available to build systems. 

Of course, the number of AI specialists available is small, which will undoubtedly push up the wages for individuals with appropriate expertise. However, the higher salaries will encourage many office workers to simply retrain, moving from automatable tasks, like accounting, into hard-to-automate prospects, like AI. 

The process will likely follow a similar trajectory to what happens in commodity markets. When the price rises, more investment flows into the sector, causing greater capacity and falls in prices. The same dynamic could easily play out in the AI labour market. As demand increases, salaries will rise and more people will reskill. Once they have the necessary skills, they will hop back into the market, pushing down wages with higher supper. 

Worker Interactions Will Remain Valuable

On top of this, the human element of office work will likely remain valuable. Even if robots can perform many tasks, business executives or customers could demand that a human remains in the loop. 

Interestingly, the capacity of firms to do this will increase as AI-led productivity and output improves. Higher margins and larger revenues will enable more use of staff and, in some situations, could improve employment levels overall (especially if firms feel confident). 

It could also lead to positive feedback effects that stimulate the economy further. Higher output and wages could result in more demand, encouraging firms to continue adopting pro-growth policies. 

Even if this dynamic doesn’t play out, there are plenty of other areas we could see people remain. For example, we might see workers continue negotiating on behalf of firms. Such skills are highly people-focused and unlikely to have the same effect as an AI, even if executed identically. 

It is also hard to imagine how AI could replace management roles. While software might reduce cognitive load in offices, it probably won’t replace field workers or businesses that require physical manipulation of reality, like construction. 

Firms in this category will probably keep most of their office staff shortly. While full automation could be possible with genuinely agentic systems, it is unlikely to emerge first. What is more likely is a gain of capabilities of head office staff, while management requirements remain. 

Finally, employees involved in client relations will likely continue working in the office. These individuals need to build rapport with customers and make them feel like there is someone on the other end of the phone line who cares about their wellbeing. 

Chatbots could possibly provide some functionality to enable this to happen. However, it is unlikely that they will be able to recreate the full value of human interactions while justifying the cost. 

Workers May Simply Learn New Skills

Another possibility is that AI won’t replace office workers but simply require them to generate new skills. That’s the pattern of the past and, until recently, was assumed to be the pattern of the present, too. 

The idea is that AI will replace some functions but not others. The automated teller machines (ATMs) of the 1970s and 1980s are a good example. These didn’t replace workers but instead forced them to do something different. 

If you went into a bank in the 1960s and asked to take out cash, a teller would go to the safe and then hand you your money. If you do the same thing today, somebody will greet you at the door, sit down with you and discuss your financial options. In other words, bank tellers went from being clerks to customer relations agents. 

A similar transformation could occur for office workers as a whole. For instance, report writers could see their old jobs taken away and replaced with new jobs that lean more on presenting findings to senior management.

“We think we’re going to see more of these up-shifts in skill utilisation,” says Apryse’s Smirnow. “It’s obvious that things like report writing are going to become simpler and take up less time. Because of this, the staff who work on these tasks will likely spend more time interacting with colleagues and sharing information through more accessible channels.”

The extent of this effect will depend on company policy. It might be possible for AI to produce a report in a few seconds, but if senior management doesn’t approve of the process, it could mean that workers will remain in their existing capacity for longer. 

The Office Could Change

Finally, if AI continues to get better, it might change the fabric of the office, adjusting how it functions in profound ways. For example, you can imagine AI systems optimising office layouts to make workers more efficient. You can also anticipate companies using office spaces to promote worker well-being. 

Consequently, workers may enjoy working in the presence of AI. Systems could enhance productivity and make life on the job more enjoyable. 

Wrapping Up

As we have seen, AI could have profound effects on office workers in the future. However, we need to temper our expectations. While media narratives support the idea that employees will be replaced, the reality is likely to be far more complex. AI isn’t yet human-level, meaning there is still a role for office workers. Furthermore, even if it does achieve a human level of capability, there are still office work elements that bosses won’t want to replace. 

(Devdiscourse's journalists were not involved in the production of this article. The facts and opinions appearing in the article do not reflect the views of Devdiscourse and Devdiscourse does not claim any responsibility for the same.)