The AI genie is out of the bottle. If you have followed news headlines these days, you have probably heard the buzz around ChatGPT, an AI-based language model that can imitate human language and reasoning to a scary degree. As I was playing around yesterday a bit more with ChatGPT, marveling as an AI expert at how far we have come since the early days of machine learning, I was also reminded of the importance of preparing our society and our economy for what is to come.
Jobs in the ChatGPT Age
While the current state of AI is still somewhat away from taking over human jobs, I think it’s safe to say that today, compared to a year ago, we have a better understanding that the speed and scope of this transformation are well beyond our initial imagination. Just search for ChatGPT prompts and you will get a sense that it’s not just manual jobs that will be lost to automation. As AI systems become more advanced, they will likely be able to perform many tasks currently done by humans, such as data entry, customer service and even some forms of decision-making. Soon, an AI will do better market research, copywriting, content creation for social media, journalism, fiction and non-fiction writing, and even legal research and contract drafting (probably only needing minor customizations by an actual lawyer). Not too far behind will be non-scientific coding, consulting/problem solving and many other areas currently paying six-figure salaries.
Granted, the use of AI is also expected to lead to the creation of new jobs, particularly in the areas of data science, machine learning and software engineering. As AI is used more widely across industries, organizations will need experts who can develop, customize and manage these systems. Additionally, with the increasing amount of data being generated, there is expected to be a high demand for data analysts, scientists and engineers. In general, it’s expected that jobs in the future will require a combination of technical skills, creativity and human qualities such as emotional intelligence and the ability to work in a team. So what happens to everyone in their 40s to 60s who might not be able to relearn a lot of these skills in a short period of time? I would argue that we will see increasing unemployment for a section of the population that today does not (yet) feel threatened by these trends.
Universal Basic Income as an Enabler of the New Economy
However given a basic level of economic security, much of the workforce can have a safety cushion and enough time to find their way in the new economy and adapt to new opportunities. And universal basic income (UBI) is one such potential option. UBI is an economic policy surprisingly supported by many progressives and conservatives alike, that proposes providing all citizens with a basic income, regardless of their employment status or income level. When Andrew Yang suggested that the increasing automation of jobs due to AI could make UBI a necessary policy to address job loss and income inequality, most people laughed it off. I don’t think we have the luxury of ignoring it for too long.
UBI could provide a safety net for those who are unable to find work and help ensure that everyone has access to the basic necessities of life. This would provide a cushion for people whose jobs may be replaced by automation, allowing them to retrain for new careers or pursue other opportunities without the burden of poverty. Additionally, UBI could also increase consumer demand, leading to an increase in economic activity and job growth.
U.S. Experiences With UBI
It’s worth mentioning that UBI is still largely a theoretical idea and has not yet been implemented on a large scale, and therefore it’s difficult to say what the specific effects of UBI on the U.S. economy would be. In the United States, there have been two recent small-scale studies on UBI. Starting in 2019 in Stockton, California, a group of unemployed individuals received a monthly stipend of $500 over the course of 24 months through a citywide initiative, resulting in an increase in full-time employment during the time period. And in Hudson, New York, the same monthly stipend paid to unemployed participants increased employment from 29% to 63%, over a one-year time period.
Challenges with UBI
There are several challenges associated with implementing a UBI policy, some of which include:
1. Cost: One of the main challenges with UBI is the high cost of providing a guaranteed income to all citizens. Depending on the level of income proposed and the number of people who would be eligible, the cost of UBI could be quite substantial and would likely require significant increases in taxes.
2. Incentives: UBI could also disincentivize people to work, as they would be receiving a guaranteed income regardless of their employment status. This could lead to a reduction in economic productivity and could make it more difficult for businesses to find workers.
3. Administration: The implementation of UBI would also require a complex administrative infrastructure (most probably IRS), with a need for systems to verify eligibility and distribute funds. Ensuring that funds are distributed fairly and efficiently could be a significant challenge although easier since everyone would get the same amount.
4. Cultural resistance: Some people might resist UBI if they believe it would be a handout or charity, which could lead to public opposition to the policy.
However, like anything else, there are ways to address these challenges through experimentation and adaptation.
Conclusion
The prospect of AI taking away jobs has gone from a challenge for the next generation, to a challenge within this decade in a matter of 1-2 years. As a business leader and AI expert, I believe it is necessary for other leaders and elected officials to start thinking of how to support people to adapt to the new workforce requirements. That support will likely require a solution such as a basic level of economic security offered to the population as a whole.
