Editor's Note
How Will GenAI Affect Jobs?
This is a great article that explains how GenAI may affect jobs and why it will have a bigger impact on women and people of color.
It's based on a McKinsey study that assumes that employers will replace humans with computer programs any time they can. Computers don't need health care benefits.
I'm not sure that's right. The way GenAI works is that it takes a huge database of things people have written and then predicts the next word, one word at a time, until it builds sentences and paragraphs. There are lots of ways to give the programs instructions, but it's all based on predictions.
As anyone familiar with weather predictions can tell you, predictions are only as good as the information you have at the time. A prediction that there is a 99% chance of sunny skies means that there is a 1% chance it won't be sunny. And even if it rains or snows, the prediction will be "correct."
With GenAI, the information it relies on is changing all the time. In some cases, these programs are, as my friend Kate Bischoff says, eating the internet. What could possibly go wrong?
What's going wrong already is that biases are imbued in the data because humans. Some GenAI programs are also designed to vary the responses to the same questions, which gives you alternatives to choose from, but also increases the chance the program will make things up. These systems have no idea what is correct, true, or good. They just pick the next word from the options available in the data and whatever instructions they have.
The problem that I think is coming is that now that we've let these things loose on the internet, they will start eating content created by GenAI, which means the systems will make more things up, only faster. Once the machine-learning systems are using information they just made up, you can't just take it out. It's more like trying to take shortening out of a pie crust than pulling cherries out of the filling.
What I absolutely agree with is that the people least able to deal with the problems that arise will be affected first. I also agree with the suggestions. And there's even a downloadable ethics and AI policy at the end.
- Heather Bussing
The rise of artificial intelligence – especially Generative AI (GenAI) – will cause women and diverse groups to lose their jobs at disproportionate rates through the year 2030, according to a report published last week by McKinsey Global Institute. The culprit: the projected automation of many office support, customer service, and food services positions – disproportionately held by women and people of color. But the good news is that you can take proactive steps to address this concern and meet your workforce diversity priorities. Below are seven steps you can take to alleviate this trend.
Putting it All in Context: Quick Recap
While AI tools have been around for years (think rudimentary customer service chatbots and predictive text tools), it’s the rise of GenAI that has now caused most businesses to realize how automation will change the shape of the workplace forever. Exploding onto the scene earlier this year with the release of Open AI’s ChatGPT, we’ve now seen numerous GenAI tools on the market that are already transforming the way natural language text, images, video, and audio can be created and used.
As an increasing number of businesses – and people – use GenAI products, the accelerated pace of the AI revolution has caught many by surprise. But like it or not, GenAI will affect a wide spectrum of work activities across all industries in transformational ways. In fact, McKinsey’s July 26 report estimates that nearly 30% of the hours that Americans now work will be taken over by AI by the year 2030.
The good news for those of you worried about losing your job to a robot: the report concludes that AI will actually fuel overall employment and economic growth, and will not “wipe out” jobs in the long term, particularly in healthcare, STEM, banking, insurance, pharmaceutical, and transportation. The bad news: there will be short-term job losses as the overall market adjusts to the necessary occupational shifts brought about by this technology. And the even-worse news: the job losses will largely occur in positions predominantly held by women and diverse groups.
Where Will the Losses Occur?
According to McKinsey, the following industries will see the highest acceleration of GenAI use by the end of the decade, with the figures below showing the decline in future job growth as employees are replaced by AI:
- Office support (-18%) – including administrative assistants, receptionists, information clerks, and bookkeepers
- Customer service and sales (-13%) – including retail salespersons and cashiers
- Food services (-2%)
- Production work (-1%)
Many of these jobs share some common traits: they involve repetitive tasks, data collection, and data processing, all of which can be efficiently and quickly done by GenAI products. And McKinsey’s report notes that those jobs are disproportionately held today by women and people of color.
The bottom line? “We estimate that 11.8 million workers currently in occupations with shrinking demand may need to move into different lines of work by 2030,” says the report.
“Roughly nine million of them may wind up moving into different occupational categories altogether.” And the solution is not as easy as simply saying that these workers should just change their occupations to ones that will be bolstered by AI. This process (as opposed to finding a new job within the same occupation) often requires adding new skills. It is thus a more challenging exercise for lower-wage workers, who often face educational and other resource barriers to acquiring new skills.
What Can Employers Do? Your 7-Step Plan
It will be incumbent upon employers to take affirmative steps to ensure that the nascent AI revolution does not disproportionally impact women and people of color. Here are steps you can take to join the effort.
- Offer widespread access to training programs. Larger organizations will have the capacity to retool their own training programs to open up opportunities for their workers, while smaller companies may need to team up with community organizations, educational institutions, industry groups, and government agencies to identify reskilling prospects.
- Adapt your hiring and training practices. Since there is expected to be an overall net positive to the labor market thanks to GenAI (especially for new roles calling for social-emotional and digital skills), you can meet your diversity efforts by expanding your outreach to include more women and persons of color in the applicant pools for these roles. Moreover, rather than solely focusing on candidates with prior experience, also consider applicants based on their potential, their ability to learn and adapt, and their transferable skills. Here are five questions you can ask as you rethink your job requirements to meet this new era.
- Provide micro-upskilling opportunities. Rather than approaching the upskilling process as a burdensome task, you can facilitate a process that ensures workers receive training and education on a more frequent basis—weekly or quarterly—to learn a new skill or hone an existing one. Here’s a checklist on accomplishing this in your own organization.
- Diversify your candidate pools. There are many ways for employers to seek out candidates from a broader grouping of applicants. You can consider whether functions can be performed remotely to increase the geographic mobility of your pool and identify candidates with disabilities, in rural or other urban settings, or those with family responsibilities. You could also seek out retirees who want to return to the work environment or provide opportunities to those with past criminal records seeking a second chance.
- Create job transition programs. These programs can provide support for workers who need to change occupations due to AI automation. They could include skill assessments, retraining programs, job search assistance, and financial planning.
- Demonstrate AI’s benefits to demystify the technology. This recent Insight published last week identifies some tips you can follow to ease the pathway and allay the fears of your workforce as you introduce AI to your organization.
- Develop ethical AI workplace policies: Don’t forget that your work starts at home. Make sure you adopt policies to ensure that the very AI applications you implement are used responsibly and that the technology does not contribute to bias in hiring or other work decisions. (You can read an overview on how to create such a policy – and receive a free template you can put to use right away – by clicking here).
