One of the most current topics in the news today is the advancement of artificial intelligence, otherwise known as AI. AI has existed for many years but has come into prominence recently as it has advanced to the point where it can do certain jobs better than humans. Instead of simply following instructions, AI can now learn on its own and come up with its own solutions to the problems it is given.
This has caused people to be concerned that their jobs will be taken over by AI. Graduates are particularly vulnerable to this as they already have a difficult time getting their first full-time job. There are over 2 million graduates in the UK seeking jobs at any one time!
Although it is true that AI will eventually take over many current jobs, this will take time to be properly implemented, and new jobs will emerge to replace the ones that were lost.
What Type of Jobs Is AI Taking Over?
Arguably the largest category of jobs that AI will replace is white-collar roles. White-collar jobs are jobs that are usually done in an office, working from behind a desk. They are called white-collar because these employees typically dress smartly.
A recent report by Goldman Sachs predicts that other roles at risk of automation include architecture, engineering, legal work and administrative office work. This is because, generally, modern AI can write reports, sort data and design structures more effectively than a human can, and this ability will only continue to grow as AI continues to learn.
A 2013 study by Frey and Osborne predicts that some of the jobs most at risk include clerks, insurance underwriters and tax preparers. All of these are white-collar jobs with a nearly 100 per cent chance of being automated.
It is also cheaper and more efficient for businesses to use AI over human workers when possible, as AI is faster and less likely to make mistakes than a human. This is a concern for graduates especially, as the junior white-collar roles they typically apply for after leaving university are easily replaced by AI since companies are more likely to rely on an AI than graduates in minor roles with little workplace experience.
On the other hand, blue-collar jobs are under less threat from AI than white-collar ones. Blue-collar jobs are roles that usually need more physical labour or require expertise in a particular area.
While the number of white-collar roles has grown in recent times, especially in medicine, the number of blue-collar jobs in the UK has fallen from 41% in 1991 to 29% in 2006. This may be reversed with the rise of AI, as blue-collar jobs are far less likely to include things AI is capable of doing. The same Frey and Osborne study found that some of the least likely blue-collar roles to be automated are police officers and mechanics. An AI cannot arrest criminals or fix a boiler.
As blue-collar jobs do not require a degree nearly as often as white-collar jobs, future graduates may have to do specialised qualifications on top of their degrees to be considered for blue-collar roles. Universities will become less popular as they become less and less able to guarantee work for those who attend, and people are likely to decide that any tuition fees they have to pay to go to uni are no longer worth it.
We predict that Universities will become less of an automatic choice for many as this decade progresses.
Why the Impact of AI is Limited
Despite many predictions that nearly every job will end up being automated, the truth is this is probably incorrect. While AIs have many advantages over humans, they also have their disadvantages. Currently, AIs are rather narrow in what they can accomplish, only being able to do what they are programmed for. People can perform a wider variety of tasks, should they unexpectedly need to.
There are also a number of jobs that AI could theoretically do but aren’t trusted to do. For example, AI is already used on an aircraft for the majority of the flight, with take-off and landing being the exceptions. An AI could be programmed to perform these additional functions, eliminating the risk of pilot error. However, very few passengers would ever be willing to fly on a plane with no pilot, due to the risk of the autopilot malfunctioning.
There are also jobs AI are simply incapable of doing. Although an AI could absolutely create a school curriculum to teach any conceivable subject, it is impossible to believe that schoolchildren would follow instructions given to them on a screen by an AI without adult supervision and mentoring.
Similarly, while AI could be programmed to teach those with depression about cognitive behavioural therapy, no one with serious mental health issues would take much comfort from being assured of their self-worth by an algorithm instead of a trained therapist. No matter how smart AI become, it is no substitute for genuine human connection.
The impact of automation will also be limited, at least in the short term, by how long it will take to implement AI in the workplace. This is partially driven by scepticism towards AI and the automation of work. While the idea of an AI apocalypse sounds like something more out of a sci-fi movie than reality, it is a big enough concern that over 1000 experts, including Elon Musk, have recently argued that AI development should be temporarily halted.
Automation will also be implemented slowly due to humanity’s own self-interest. You do not have to be an expert to realise that the majority of the world’s population being made unemployed in the space of a few short years would be a bit of a problem.
Politicians are slow at the best of times. It will take them years, if not decades, to figure out a plan to simultaneously implement automation and deal with its consequences, and even longer to complete such an endeavour. It is easy to imagine politicians in the not-too-distant future running for election on the promise to protect people’s jobs from automation and to stall the introduction of AIs into the workplace. If you are worried about the impact of AI on the future of humanity, be assured, turkeys will not vote for Christmas.
This should be reassuring to any graduate worried about finding a job. They will have time to adapt their future plans around automation so that they are not caught off guard. They are unlikely to discover that the job they have studied for has become obsolete by the time they graduate.
New Graduate Jobs
Another reason why graduates should not be overly concerned by the emergence of AI is that new graduate jobs will emerge to replace those that will become automated. This is simply another cycle in which humanity identifies new roles that need to be filled as technology advances and old jobs become obsolete. An example of this is the industrial revolution, resulting in new jobs working in factories.
These new jobs will largely be focused on things AIs can’t do and things people don’t trust them to do, such as programming the AIs themselves, as well as doctors and surgeons.
Perhaps schools will alter their curriculums to educate people to make them better suited to these new roles. University courses that are required or preferred for these new jobs will become more popular as students seek to prepare themselves for this new AI-centred working environment.
It is easy to speculate about the future of work, but there are jobs today that people several decades ago could have never predicted would become a reality, such as selling items on e-marketplaces like eBay and Amazon.
People are also used to changing careers. In one article from 2018, Acemoglu and Restrepo show that between 1980 and 2015, half of all employment growth occurred when people changed the tasks they were required to do or had a different job title.
Summary
Overall, it is accurate to say that AIs will eventually take some graduate jobs as the ‘cognitive gap’ between AI and humans will always exist.
It will take time for AI to be implemented in the workplace and for jobs to be automated, leaving enough time for graduates to plan around it, figure out which job they want to do, and ensure they are qualified to do it by the time they leave university.
However, our advice to graduates is simple.
- Start asking employers how the AI revolution will affect you.
- Read articles on AI from as many reliable sources as possible.
- Start using AI in their own lives. We already have grads correcting and improving their CVs using ChatGPT.
- Ask employers how they are using AI and how that will affect current positions in their workplace.
There will also be new jobs that are always impossible to foresee that emerge as a result of jobs becoming automated, and old roles that need more graduates, such as AI programmers, to keep up with increased demand.
We recommend Rex Woodbury’s blog.
His writing on all this is of the very highest quality. All of our grads read him every week.
;)