Unlock Editor's Digest totally free
FT editor Roula Khalaf selects her favourite stories on this weekly newsletter.
When will artificial intelligence replace human employees on a bigger scale? This query has turn out to be the topic of much speculation within the wake of the AI boom. But long before now we have to fret about it, a shortage of human employees could prove to be the largest obstacle to the AI industry.
Revenue at Taiwan Semiconductor Manufacturing Company, the world's largest chipmaker and maker of the chips powering the AI revolution, rose 45 percent to $7.9 billion in July, reinforcing rapid growth within the second quarter. Demand for AI chips stays strong, with the high-performance computing business accounting for greater than half of TSMC's revenue last quarter.
Despite these stellar numbers, shares of AI corporations have been volatile in recent months, making investors increasingly wary of the risks at corporations like TSMC – from the earthquakes in Taiwan to China-Taiwan tensions to general geopolitical tensions. Less discussed, but just as vital, is a looming crisis on account of the shortage of engineers and technicians.
Until now, the prevailing view was that increasing chip production capability was simply an issue of cash. The global chip shortage that began in early 2020 was combated by governments throwing billions at chipmakers to extend their capability, preferably of their backyard. TSMC has expanded its semiconductor factories within the US, Germany and Japan.
The US is certainly one of the countries most aggressively increasing capability, with investment within the chip industry expected to achieve greater than $250 billion over the subsequent five years. But because it seems, money is of limited use on the subject of chip manufacturing.
The problem is that constructing a chip factory isn’t as easy as establishing a brand new smartphone assembly factory overseas, where you possibly can quickly hire and train local employees. Chip factories require highly expert employees with master's and doctorates in science and engineering to run them. Even constructing a chip factory itself requires specialized labor.
The major investments and subsequent expansion of the US chip sector mean greater than 160,000 recent job opportunities in engineering and technical support, in addition to additional job opportunities in related construction trades, in response to McKinsey evaluationYet only about 1,500 engineers enter the chip industry every year. For chip technicians, the number is even lower: only about 1,000 recent technicians join every year. Over the subsequent five years, the necessity for these employees is anticipated to rise to 75,000.
Meanwhile, employment within the U.S. chip industry has fallen 43 percent since its peak in 2000, McKinsey reports. If things proceed at this rate, the shortage of engineers and technicians could reach as many as 146,000 employees by 2029. In South Korea, home of chipmaker Samsung Electronics, the chip industry has been battling a labor shortage since 2022 and is anticipated to face a labor shortage of 56,000 people by 2031, in response to industry estimates.
Another problem is demographic trends. Both Taiwan and South Korea, where most of TSMC and Samsung's employees are based, are fighting declining populations. The number of scholars enrolling in colleges has fallen yearly since 2012. These two countries account for over 80 percent of the world's contract chip production. A shortage of employees has already pushed back the beginning date of the TSMC plant in Arizona and is claimed to have flown in about half of the plant's 2,200 employees from Taiwan. Cultural differences make hiring even tougher.
Since each recent plant costs nearly $30 billion to construct, the factories must operate 24 hours a day, seven days every week to justify this cost economically. TSMC founder Morris Chang has identified If a machine breaks down at 1 a.m. within the U.S., it is going to be repaired the subsequent morning, but in Taiwan it is going to not be repaired until 2 a.m. Replicating this Taiwanese work culture in other countries could possibly be difficult.
Can't AI then just start making the chips? Indeed, AI helps design, test and confirm recent designs and accelerates the event of recent chips. However, making the physical chips from those designs stays an entire different story. The need for expert engineers to operate the machines is something that AI is unlikely to fill any time soon.
It's only natural that corporations struggle to fill positions that require high levels of skills and qualifications, but within the chip industry, the gap between jobs and employees is becoming dangerously large.
