AI’s Impact: Examining the Effects on High-Paying and Low-Skilled Jobs
The commercial availability of ChatGPT and the subsequent developments have sparked off major debates in academia, the corporate world, and government on:
- Will AI affect white-collar or blue-collar jobs the most?
- How will it impact skill adjacencies?
- Will the tech lead to a major re-skilling of the workforce? Or will it require upskilling?
- (The most important question of them all). Will AI get rid of most jobs because of automation?
- What and how will the tech impact a nation’s overall productivity?
Skill adjacency is a generic though sophisticated term to describe the connections between an employee’s present-day skills and those they need to acquire based on emerging needs.
For example: If you are a professional photographer, is it time now for you to upskill and learn how to use image editing software?
A professional photographer may also choose to reskill by learning videography or even social media marketing to expand his/her offerings. Visual content is his/her forte, and upskilling or reskilling revolves around the core competency. Mostly. One of the duties of HR professionals the world over is to try and understand the linkages that can be leveraged to get workers to get to such upskilling and reskilling opportunities.
Which is all very fine, but the very nature of the technology that AI represents constantly re-routes all doubts and queries to this one destination — the fear uppermost in most employees’ minds — Will most jobs become automated?
On AI-induced job losses, if the views of the doomsday brigade were to be kept aside, the consensus among social scientists and AI experts, for now, is there will be a percentage of jobs that will disappear, but a percentage of this percentage will be replaced with new job profiles. So there’s some hope.
Here’s a thumb rule that many experts ask today’s workers to follow:
Ask yourself whether you currently do a repetitive job. Does it also involve data? If the answer is yes to both, there is a huge probability that your job will be taken over by AI.
Some experts have tried to drill deeper. Which kind of job profiles will become redundant or over-written by others? Also, will it be blue or white-collar jobs, predominantly?
Again, there are no clear answers, but increasingly the conversation is veering to this — highly skilled white-collar workers may be most affected than anybody else in the workforce. Office-based work like customer service, Sales and Marketing, and jobs that rely on cognitive expertise will be most impacted.
So Will AI Significantly Impact Only High-paying Jobs?
A 2019 research report by the American research group Brookings Institution (slightly before ChatGPT came online) has actually spelled out the jobs that could face the highest exposure to AI. The entire report’s forecast is that better-paid, better-educated workers face the most exposure because of AI. Workers with graduate or professional degrees will be almost four times as exposed to AI as workers with just a high school degree, it said.
The jobs listed below have the highest exposure, as per this report:
- Chemical engineers
- Political scientists
- Nuclear technicians
- Physicists
- Occupational therapists
- Gas plant operators
- Administrative law judges, adjudicators, and hearing officers
Reference: Brookings. edu
An even earlier report by McKenzie, however, claimed AI would impact blue-collared jobs like fast food service and machine operations the most.
A report by the World Economic Forum released on May 1, 2023, lists the fastest-growing and declining jobs:
There will be more studies in the future on the impact of AI on jobs, but for now, the haze is slowly lifting. So, even though the picture may not be very clear yet, the coming job displacement wave will, most likely than not, affect the highly skilled and the extremely poorly skilled workers.
Let’s say you are a radiologist interpreting medical scans of a suspected TB patient. As a human, you take at least a day or two to come up with a report. Which is also mostly based on your experience of having studied previous such scans, coupled with your formal education. An algorithm with the right training dataset, which means thousands of scans of other TB patients, will be able to do the job far better than a human. And, within minutes. And with a far greater degree of accuracy. Which will then mean the radiologist will now have to upskill.
Poorly skilled workers will be affected in a slightly different way. For this lot, AI will be the tool to help them learn and bring them to the level of their already well-trained counterparts.
In one interview, Stanford professor Erik Brynjolfsson reveals how AI can identify skill adjacencies to bridge the skills gap. In an article by the World Economic Forum, he extends the example of a forensic expert. If you were to take a forensic accountant and teach them some cyber, they can become a cybersecurity expert, maintains Erik.
Reference: Weforum.org
Microsoft Study Finds Employees Want To Forge Partnership with AI
Here’s something surprising. Despite concerns over job displacement, a study has shown that employees were more eager for AI to ease their workload than they were afraid of being replaced by it.
The study by Microsoft spoke to 31,000 people in 31 countries and analyzed trillions of Microsoft 365 productivity signals, along with labor trends from the LinkedIn Economic Graph.
The report said although 49% of workers feared that AI may take their jobs, 70% would prefer to delegate as much work as possible to AI to alleviate their burden.
Reference: Microsoft
Research paper author and organizational psychology professor Adam Grant called the finding “intriguing”. People are more enthusiastic about AI rescuing them from burnout than they are anxious about it eliminating their jobs, he pointed out.
The survey revealed that people are seeking AI assistance in almost every aspect of their work, with 3 out of 4 respondents indicating that they would be comfortable using AI for administrative tasks (76%), analytical work (79%), and even creative work (73%). Furthermore, people are also interested in AI’s help with finding information (86%), summarizing meetings (80%), and planning their schedules (77%).
Automation introduced by AI is a general apprehension uppermost on everybody’s minds, but the fear is taking attention away from another critical question — the re-engineering of society’s fabric and the role corporates will play in it.
Will Coming Shift in Job Profile Worsen Society’s Disparities or Correct Them?
Productivity growth is a crucial factor in a nation’s economic prosperity. Efficiency because of AI will increase, which almost everybody in the ecosystem seems to agree upon. For now. But will it be the same for productivity?
As shocking as it may sound — productivity growth has remained stagnant in the US and other advanced economies since around 2005, according to this report. This means workers’ wages have, by and large, remained in the same range between then and now.
This lack of growth has also worsened income and wealth inequality. Now, some experts fear that the coming re-jigging of employment and job profiles may widen the income gap.
Will AI break this impasse and raise productivity so much that even as it displaces a lot of jobs, it creates new ones and raises living standards? If yes, how will it tie into a nation’s overall productivity?
Since November 30, 2022, when ChatGPT was announced, and the competitors of OpenAI also jumped into the AI race, it’s becoming clear that this marathon, too, will be run by the ad-revenue motivated Big Tech such as Google and Microsoft. This means that the large language models will be dominated by the same big companies that have ruled much of the digital/internet world for decades.
The concern is that these companies may use this new, AI-led automation to replace comparatively well-compensated white-collar jobs, leading to a shift for these workers towards lower-paying service roles. Meanwhile, a small group of individuals who are most adept at utilizing the new technology may reap most of the benefits.
In a well-argued, well-presented article in MIT Technology Review, Editor David Rotham talks of the “gold rush” sparked off among companies because of generative AI. David says the focus is — how do we (corporates) make money out of this?
The upcoming revolution of large language models, represented by ChatGPT, he points out, is expected to bring significant changes to the economy. The question is how we will shape this transformation. The impact on various jobs is inevitable, and it is up to us to determine whether this change will result in widespread prosperity or not, writes David.
Initial findings, says the report, indicate that ChatGPT and other generative AI systems could potentially provide “upskilling” opportunities for those struggling to secure employment today. Current AI adoption in companies has been limited to optimizing tasks incrementally, leading to only marginal productivity and efficiency gains.
How much, how little? Productivity jump or mere enhancement? Redefinition of skills or their complete loss? Even more disparity or the comeback of some semblance of equality in society? These and many related questions continue to be probed, studied, and debated by our social scientists, economists, technology experts, and psychologists. For now, though, the consensus is only around very few hypotheses.
As for the answers to the remaining questions, subscribe to my newsletter — AI For Real — to keep track and be updated as AI evolves over the coming years.
