Siva Balasubramanian serves as the Harold L. Stuart Endowed Chair in Business, professor of marketing, and associate dean at Illinois Tech’s Stuart School of Business. His research includes how artificial intelligence models have been, are being, and could be adopted in the modern business world.
What jobs do you feel are currently most vulnerable to elimination due to AI deployment?
A: My view is that no job is safe. Operational pressures are such that if layoffs improve the bottom line, businesses will act first and ask questions later. That is the hard truth.
Five years ago, industry offered assurances that no one would lose their jobs because of AI. If AI displaced workers in future, the message was that industry would retrain them. A recent Wall Street Journal article “The Disappearing White-Collar Job” questioned that premise. It appears people may lose jobs because it’s difficult to retrain everyone on AI. That hit a nerve for many who believe it reflects what most CEOs think.
Five years ago, industry offered assurances that no one would lose their jobs because of AI. If AI displaced workers in future, the message was that industry would retrain them. A recent Wall Street Journal article “The Disappearing White-Collar Job” questioned that premise. It appears people may lose jobs because it’s difficult to retrain everyone on AI. That hit a nerve for many who believe it reflects what most CEOs think.
What are the current risks associated with businesses who could turn to AI such as ChatGPT to replace white-collar jobs?
A: ChatGPT performs many white-collar tasks. But let us compare AI and human intelligence. Much of how AI works is modeled on how the human brain works. However, does AI actually think like humans, or is it some other unique type of intelligence that doesn’t come close to human intelligence? My answer is the latter.
Machines do not think like humans; they just pretend to think like humans while they demonstrably do not. Even the creators of large language models (LLMs) do not fully understand how they work. That’s why we encounter ChatGPT’s hallucinations. Human thinking is fundamentally rich and different—there’s a lot of organic thought underlying human intelligence. Human employees can also be loyal. Businesses can therefore lose a lot by outsourcing everything to AI.
Machines do not think like humans; they just pretend to think like humans while they demonstrably do not. Even the creators of large language models (LLMs) do not fully understand how they work. That’s why we encounter ChatGPT’s hallucinations. Human thinking is fundamentally rich and different—there’s a lot of organic thought underlying human intelligence. Human employees can also be loyal. Businesses can therefore lose a lot by outsourcing everything to AI.
What current AI deployment trends have you noticed taking place in the business world?
A: Recent layoffs were characterized as a response to over-hiring during COVID-19. That disguises the real reason: the pressure to decrease costs by leveraging AI.
Businesses are under the gun on AI. FOMO (fear of missing out) drives their actions. Their anxiety revolves around “if I do not embrace AI, my competitors will race ahead.” We will reach a point soon where every business has adopted AI, but the degree [to which a specific business adopts] may vary.
Businesses are under the gun on AI. FOMO (fear of missing out) drives their actions. Their anxiety revolves around “if I do not embrace AI, my competitors will race ahead.” We will reach a point soon where every business has adopted AI, but the degree [to which a specific business adopts] may vary.
Where do you see AI’s expansion into the workforce eventually ending up?
A: The ultimate AI innovation will likely occur in the knowledge domain. This will be momentous, because it will unleash, and democratize, knowledge to levels previously unseen.
Consider Wikipedia’s outlines of academic disciplines that lists thousands of disciplines. Academics recognize a scholar as accomplished if they master the bulk of the literature in one discipline. The total accumulated knowledge from all disciplines basically undergirds everything that humanity knows. However, millions of publications authored by knowledgeable scientists are not widely read.
AI can change that. It possesses the remarkable ability to scan, absorb, and summarize all that knowledge in short order. It can unleash innovation by synthesizing that knowledge base in novel ways. This holds immense potential for creative multidisciplinary work, by allowing researchers to rapidly develop and cross-pollinate ideas from disparate fields. That’s the awesome promise of AI for giant leaps in innovation and human progress.
Consider Wikipedia’s outlines of academic disciplines that lists thousands of disciplines. Academics recognize a scholar as accomplished if they master the bulk of the literature in one discipline. The total accumulated knowledge from all disciplines basically undergirds everything that humanity knows. However, millions of publications authored by knowledgeable scientists are not widely read.
AI can change that. It possesses the remarkable ability to scan, absorb, and summarize all that knowledge in short order. It can unleash innovation by synthesizing that knowledge base in novel ways. This holds immense potential for creative multidisciplinary work, by allowing researchers to rapidly develop and cross-pollinate ideas from disparate fields. That’s the awesome promise of AI for giant leaps in innovation and human progress.
In what way might a worker “upskill” or “reskill” to make them more resilient to an AI replacing their job?
A: The most promising path is to learn about AI as much and as quickly as possible, to develop an informed understanding of its risks and rewards. Read all AI reports. A recent one from IBM underscores that “AI won’t replace people—but people who use AI will replace people who don’t.”