The enormous potential of AI cannot be dismissed
In an op-ed recently in the New York Times, the Nobel-Prize winning economist Paul Krugman wrote, "Artificial intelligence (AI) is already having a significant impact on the economy, and its influence is expected to grow significantly in the coming years… Overall, the effects of AI on the economy will depend on a variety of factors, including the rate of technological advancement, government policies and the ability of workers to adapt to new technologies."
A very reasonable and sound opinion, coming from Krugman, who is very well-regarded in the profession as well as in the American political establishment. Unfortunately, he then immediately goes on to disown the opening paragraph and offers a disclaimer: "What I did was ask ChatGPT to describe the economic effects of artificial intelligence; that's just an excerpt."
Krugman was referring to the new chatbot software ChatGPT – released in November 2022 by OpenAI – which has taken the world by storm with its flashy abilities. ChatGPT can quickly compile information and write essays, spit out computer codes, do complicated mathematics, hold conversations, generate short stories, and whatnot.
The availability of ChatGPT unleashed a veritable tsunami of predictions, reviews, and dire warnings in newspapers across the globe and social media on the effects of the AI revolution. Comments and reactions are pouring in incessantly from the IT industry and from various politicians, intellectuals, and, of course, economists.
A few of the related thought-provoking questions are: Can computers really think and feel? Will machines finally replace humans and take away jobs? And how soon will that happen? These are all debatable issues, but allow me to offer my answer before I change my mind next year.
I am aligned with Krugman, who does not foresee significant changes immediately. "Artificial intelligence likely won't transform the US economy in the next decade," he said in NYT, adding that it could boost labour productivity by doing some jobs better than humans. My caveat. He may be wrong since AI race is heating up, with tech giants Microsoft, Google, and Baidu among those ramping up their efforts to launch their versions of advanced chatbots.
AI will have a major impact in four areas: productivity or output per capita, jobs loss, education, and healthcare and medical research. There will be substantial gains in productivity when AI can help us with repetitive tasks. The impact on the economy might take a little while, but it is inevitable. Whether it takes five years or 10 years for productivity gains to show up in each sector. It is already visible in e-commerce, banking, and transportation.
We all know that the AI revolution is a work in progress. There is hardly any discipline where we do not see some form of AI adoption and adaptation. Business, education, journalism, medicine, government, law enforcement, arts and culture, transportation; you name any field, and AI has made inroads there with mostly positive outcomes.
Let us take the case of healthcare. Until a decade ago, physicians based their diagnoses entirely on knowledge stored in their memory. Things have changed today thanks to the rapid development of AI and its use in clinical decision support systems (DSS). A recent Johns Hopkins study reported that more than 250,000 people in the US die every year from medical errors and other reports claim the number to be as high as 440,000. Medical errors are the third-leading cause of death after heart disease and cancer, and these errors can be reduced with the help of AI.
AI will have a major impact in four areas: productivity or output per capita, jobs loss, education, and healthcare and medical research.
In their book, "The Age of Scientific Wellness" (Harvard University Press, 2023), Dr Lee Hood and Dr Nathan Price provide some interesting facts about the benefits of AI-enhanced medical care in the areas of diagnosis, pharmaceuticals, and treatment of diseases.
Studies have revealed that each patient generates over 100 million megabytes of data, from basic testing results, MRI, information on patient outcomes, and more. "These (DSS) systems leverage what computers are naturally good at – storing, recalling and correlating vast amounts of information virtually instantaneously – and link it to the ability of a human expert to reason intuitively and think creatively," Hood and Price write.
AI-enabled healthcare has already saved innumerable lives in many different ways. MedAware, an AI programme, is used to help doctors avoid prescribing the wrong prescription due to unintended mistakes. MedAware sends an alert to the doctor, who is frequently overworked and sometimes exhausted, when the medication does not match the patient's need or could have an adverse reaction.
As for job loss, there is a good possibility that AI will not only increase productivity per worker but also replace some repetitive or routine tasks currently done by humans. On the other hand, jobs with a strong human element – such as being a therapist – are especially unlikely to be taken over by technology, according to Dimitris Papanikloaou, a professor of finance at the Kellogg School of Management at Northwestern University. "Jobs that emphasise interpersonal skills are much harder to be replaced by an AI," he says.
The advance of AI will be facilitated if governments and/or international organisations can create guidelines to address privacy, security, and equal access issues. The company OpenAI is already facing roadblocks in European countries. A week ago, Italian authorities ordered OpenAI to stop processing Italian users' personal information because it may have exposed some users' messages and payment information.
As AI is used to automate existing jobs, it can lead to discrimination against workers and exacerbate inequality. This debate centres in particular on algorithmic bias and the potential for algorithms to produce unlawful or undesired discrimination in the decisions to which the algorithms relate. These are major concerns for civil rights and consumer organisations representing populations suffering undue discrimination.
An official US-EU study recommends investing in training and job transition services for displaced employees, encouragement of development and adoption of AI that is beneficial for labour markets, and investing in the capacity of regulatory agencies to ensure that AI systems are transparent and fair for workers.
In academia, the pushback has been swift. Boston University, my alma mater, has proposed some standards for using AI tools. Students will be advised to inform the professor which AI tools were used, even if to generate ideas, to not use these tools during examinations, and to agree to "employ AI detection tools and originality checks" to avoid any potential plagiarism issues.
Similarly, instructors would have to follow these guidelines: i) Learn how AI tools work to enhance student learning; ii) Grade work created by AI lower than non-AI works; iii) Employ AI detection tools to evaluate the degree to which AI tools have likely been employed; and iv) Impose penalty for mindless use of AI generated materials.
Dr Abdullah Shibli is an economist and works for Change Healthcare, Inc., an information technology company. He also serves as senior research fellow at the US-based International Sustainable Development Institute (ISDI).