MIT’s Yossi Sheffi Talks AI, the Future of Work

“Every industrial revolution had its case of people resisting the application of technology,” Yossi Sheffi said, noting that as early as 1750 when UK textile factories moved to machines to make their goods, people have been worried about being replaced by technology.

He said the assembly line threatened jobs during the Second Industrial Revolution. Then came the Third Industrial Revolution, WiFi, laptops and smartphones, and now we have connected devices.

“When we talk about it in general [terms], we talk about conversational computing, we talk about AI - people and machines are getting integrated,” he said.

Sheffi, who recently published a book on the topic, “The Magic Conveyor Belt: Supply Chains, AI, and the Future of Work,” offered his insights during his keynote address at the recent Association for Supply Chain Management (ASCM) conference in Louisville earlier this month.

The Massachusetts Institute of Technology professor and director of the MIT Center for Transportation and Logistics, enthralled the audience at the ASCM conference with his views of technology, AI, and the future of work. He spent time discussing where artificial intelligence fits in the equation, and noted that even though it seems like AI will eliminate jobs, it won’t. Technology didn’t do that during any of the previous Industrial Revolutions, instead, it altered the type of jobs needed.

Sheffi cited the examples of the first workers that built the automobiles, telegraph operators, and even farmers.

“The point that you should realize is basically all of them found other jobs,” he said. “Other jobs came up. So, people are worried about technology … and new industries developing and new jobs coming up. It’s always happened, but it’s hard to see [in real time] and for some jobs, they absolutely don’t exist anymore.”

Generative AI is going to destroy jobs – unless it doesn’t

The hot topic today is Generative AI. Sheffi will be expanding upon this in the upcoming Logistics Management and Supply Chain Management Review Virtual Summit in December.
But at the ASCM conference, Sheffi noted that Generative AI has its limits.

“Traditional AI systems can analyze data very well and make predictions,” he said. “The stuff that most of the people do [as] forecasting. It’s not generating AI – it’s basically what’s called machine learning. … It’s actually just like any other forecasting technique. You look in the past and you try to project the future.”

Sheffi then added that Generative AI takes that many steps further, learning in much the same way that humans do, “through observation.” In this case, it is millions and even trillions of data points. But, it is not going to eliminate jobs.

“[A] previous prediction of the University of Oxford in 2013 said that by 2023 … 37% of U.S. jobs will be eliminated. Well, we are at 3.5% unemployment. It didn’t happen. Just did not happen. That’s why I say forecasting is tough, especially in the future. Job destruction is a very slow process.”

To best illustrate this, Sheffi pointed to AT&T’s invention of the automated telephone exchange in 1892. In 1950 – some 58 years after people said the telegraph would eliminate the telephone operator – there were still 350,000 people working at operators. It wasn’t until nearly 100 years later, in 1989 that the job of telephone operator finally disappeared.

Be sure to come back to Supply Chain Management Review in early December to hear more from Sheffi discussing artificial intelligence and the future of work.

SC
MR