last week, New York Times We have published a controversial article that attempts to address one of the questions everyone is asking. How do you use AI?
NYT A detailed and specific way in which people use artificial intelligence in their work. It's a good list and provides readers with a service to categorize technology. More importantly, thinking about existing (expansion period) specific use cases (e.g., selecting wines from restaurant menus, creating lesson plans, and entering medical notes can help fire the imagination of managers and executives and come up with something new.
However, the use of AI technology is already wider than many people think. Here's a recent summary of Openai:
Over 5 billion people around the world are actively using Openai's AI tools, especially free-to-use ChatGpt. They send over 2.5 billion messages per day to the platform. This includes more than 330 million people per day in the US. This width of use provides a unique window into the economic impact of AI. ChatGpt saved teachers nearly six hours a week.
Openai, where I work in some ways, last month first released how the tool is used. The pace of adoption was amazing:
ChatGPT is the most sophisticated consumer technology in history, reaching 1 million users in five days, 100 million users in two months and more than 500 million users now. Since its launch, we have seen a dramatic increase in the use of ChatGpt for work. Today, 28% of US adults employed report using ChatGpt in the workplace, compared to just 8% in 2023.
The complete OpenAI report, passing through the attributes of speed adoption, breaks down the way Americans use ChatGPT (the most common uses are learning and luxury skiing), states with fastest user growth (including the Kansas state in my hometown!), and the role of technology in advance of entrepreneurship.
When considering how AI could reshape the labor market, it is important to remember that many of the professions that use AI have not been invented. From my recent National Issues article:
The AI revolution creates many opportunities that we cannot imagine today. Standing in 2024, trying to predict future jobs is no easier than standing in 1944, and will try to predict that the future labor market will include systems analysts, circuit layout designers, fiber scientists and social media managers. In fact, about 60% of the employment workers had in 2018 had not been invented as of 1940. New occupations are largely emerging as technology advances and new goods and services need human workers to engage in new occupational tasks. Furthermore, technological advances make society richer and increase demand for goods and services – especially new Products and Services – This increases the demand for workers' skills, talent and effort.
To explain, imagine trying to explain to 19th-century classical economist David Riccardo the work of all those who support Bruce Springsteen's records and tour: sound engineer, digital editor, graphic designer, photographer, videographer, art director, musical instrument engineer, social media director, marketing expert, craftsman, craftsman, commercial craftsman, lighting engine, commerce engine, commerce engine, commerce craftsman, Mighty E Street band and Springsteen's own members. The tasks these occupations and workers performed for them did not exist in the time of Ricardo. They also did not exist because the wealth created by today's technology was not being produced. Society on Ricardo's Day could not give rock bands.
For existing jobs, we can speculate on three examples.
First, let's take the case of lawyers and paralegals. These individuals need to spend time creating briefs and categorizing documents. These are two tasks that allow a larger language model to run than it is now. This will increase the time spent interviewing witnesses and developing legal strategies. AI tools help lawyers complete these tasks by supporting a broader strategy by suggesting potential questions and asking the line of discussion with the witness. However, AI cannot effectively interview witnesses or set the strategy itself. Some law firms experimenting with AI tools today have found that it is very efficient to conduct basic legal research, which will help junior associates move faster.
Secondly, consider the case of a doctor. AI systems can read and interpret scans and test results more effectively and cheaper than humans can, so doctors need to spend much less time performing these tasks. AI tools also allow doctors to record and update patient information on medical charts and records by listening when they examine patients. This allows physicians to spend more time communicating with patients, thereby improving the quality and effectiveness of those conversations. In case of advanced illness, doctors will give them time to coordinate with other doctors to comprehensively manage their care.
Retailer managers (third example) need to reduce the time spent managing employee schedules and inventory cycles. AI tools can complete those tasks. This allows managers to spend more time overseeing and coaching workers, solving problems and creating a positive shopping experience for their customers. AI also helps managers by making suggestions for optimizing the shopping experience in the store and suggesting potential management strategies based on employee career history and other factors.
Of course, this technology also has its own risks. (For a good discussion, see the Academy of Science, Engineering and Medicine, “The Future of Artificial Intelligence and the Work,” which I was co-author.)
But we must be optimistic about AI. And the early use cases that are emerging are another good reason for optimism.
