May 9, 2023

AI: Threat or the Next Productivity Transformation Tool?

I was listening to the local morning radio station yesterday as I was getting ready for work. I heard the news reader say that IBM was going to pause hiring approximately 7,800 new employees in back-office positions – as the CEO thought the jobs could be replaced by AI. To read more, Bloomberg covered the story: IBM to Pause Hiring for Jobs That AI Could Do, Bloomberg, May 2, 2023 (

I started to smile, as it brought me back to my early days at IBM. At the time, mainframe computers were all we had, the PC and server computers had yet to appear, Bill Gates was in high school, and Yahoo was a not so kind name to apply to certain individuals. I was quickly transferred into a large research and manufacturing facility in Poughkeepsie, New York. You know Poughkeepsie, original home to Smith Brothers cough drops and a dubious line in The French Connection delivered by Gene Hackman. Anyway, there I was in the home of the supplier of the largest mainframe computers. We were constantly conducting long-range capacity planning to support the introduction of future families of computers. Capacity in this case defined as how many manufacturing direct employees and how much production and support floor space would be required. With my nifty MBA, I was deemed worthy to jump the seniority line and work with senior executives to develop long-range capacity forecasts along with my other responsibilities.

It was almost a given that a brand-new manufacturing facility was required for expansion within the next five years. I learned in a subsequent position in HQ’s that this new expansion location need was not unique to Poughkeepsie, in fact almost every IBM manufacturing location had a similar need for aa growth location to meet anticipated demand.

Most fascinating, I was told that an earlier IBM corporate strategic projected that within five years IBM would need to hire every high school graduate in the United States to meet their manufacturing workload requirements. Really, though I am not sure this was fact or legend!

So, with hindsight we know IBM did not become the universal employer. What happened? A very powerful factor intervened – productivity. By moving manufacturing schedules to three and even four shifts (yes, you can create four-shift schedules as there are 21 shifts in a week). Next, they factored in worker productivity improvement, in both the direct manufacturing positions as well as most indirect employee positions. The introduction of the PC became a massive productivity tool. I was the CFO of a large research and manufacturing facility with 10,000 employees in the mid 1980s. As we introduced PCs into our location, we required each department to reduce their staffing levels by one headcount for every three PCs added to their department. Further productivity occurred with new enterprise resource planning systems and extended further with advanced cloud applications such as SalesForce.

One productivity area IBM completely missed was the reuse of recently obsolete computers returned from customers. When the next faster, better generation of computers was introduced, the customers would upgrade to new equipment and return their leased computers to IBM. Approximately 80% of the high-end computers were leased rather than purchased. The company had very clever manufacturing engineers, and they started to introduce the returned computers to control manufacturing equipment previously manually operated by direct employees. My favorite example was in Poughkeepsie, and I was able to proudly show off a custom-engineered tool to visitors at the site. Why me? I was a cost accountant, and we were expected to know something about everything. Yes, our knowledge was thin, but we covered the entire manufacturing floor. This custom equipment was unique to Poughkeepsie, and using current terminology we would probably describe it as a manufacturing robot. The machine’s task was to plug hundreds of connector wires to the back of a mainframe’s motherboard. The motherboards were quite large, rectangular, and approximately three feet high and four feet wide.

The machine would take one end of the connector cable and plugged it into a specific socket on the motherboard. It would then take the other end and make the connection at another location on the board. It was a wonder to watch, and the arms rapidly moved up and down across the motherboard. Time to complete the motherboard was reduced dramatically, reducing direct headcount requirements, and increasing capacity. Productivity.

So, back to the beginning, why did I smile when I heard the IBM news on the radio? During my working career, IBM moved from potentially requiring every high school graduate as employees to pausing the hiring of the next 7,800 employees – due to the application of new technology for further productivity.

Now, should we be concerned that the application of new technology will result in a shock to employment levels, resulting in massive unemployment? In 1970, the annual total employment in the US was 100M individuals. According to the Bureau of Labor Statistics, total annual employment in 2020 was 200M – essentially, we doubled the number of individuals employed in 50 years. The average annual growth rate was 1.55% for the period. Also, the unemployment rate decreased from 6.1% to 3.9% in 2021. So, we’ve seen massive application of technology while experiencing absolute growth in the workforce and unemployment did not materially change.

What will AI bring? I’m still exploring the big picture, but I think we all should consider how AI will impact our organizations to ensure we remain competitive and productive.

In this article:
This article discusses the role of technology in productivity trends.
Share on social media: