You can check out any time you like…
In 1986 I joined Shell Oil, working at a refinery in Stamford-le-Hope in Essex. During the induction process we were shown a chart of the price of oil over time and were reliably informed that Oil would increase in per barrel cost to $100 per barrel by the year 2000 and that we were joining a well defined and well formed business model. I joined in September 1986 when the price of oil had plummeted to below $10 per barrel. At our refinery that meant that we were not making money.
A number of us new graduates were sat down with the refinery manager and it was spelled out to us that we needed to implement process automation technology quickly to change the cost structure of the refinery from 2400+ people to a level this price of oil could sustain. A small team of us started working on this mission, using technology to eliminate labor and automate plants that were relatively old.
We collected data from the plant equipment using Psion organizers and downloaded it every hour into the LP and QP algorithms that optimized plant settings to drive product output. It started to work and it became clear that we would be able to drive the automation we set out to achieve. There was a strike, 2ndary pickets and a good dose of use of the word scab and the locked arm stronghold around the refinery as we tried to get into work each day. It is the most visual experience I have had of the impact of technology on people and by the end of the process we had less than 700 people running the refinery. After a few months of being called in to “fix” the technology in the middle of the night, and being told by an angry shift supervisor (I won’t name you Keith Benton), that we were losing over 600,000 GBP per hour when the automation was not working, I realized that it was impossible to go back. The new operating model, the net margin of the business, the quantity of product/day/head had changed and become the new normal; now we had to make it more reliable, but we could not go back.
As we sit here today and contemplate the impact of AI on knowledge workers I am struck by the similarity to the Shell example. Sure, I have seen this countless times in my later career in telecomms and financial services where the next level of efficiency drove the next round of RIFs (Reductions in Force), usually followed by a bull market a couple of years later and more hiring. I have a feeling this time will be different.
Manufacturing never bounced back from a manpower perspective; in fact it became more automated and more efficient. A friend shared with me today the impact of an AI solution on common tasks that knowledge workers perform every day, and the improvements in productivity were more than triple what we achieved at Shell (which ended up being a 3x reduction in number of staff). As we see more automation and the number of layoffs increasing as AI automation drives efficiency we need to be thinking about how we re-skill those people to allow them to compete in the new (current) marketplace. The change will be irreversible and will require a corporate overhaul to replace legacy thinking with that of modern AI centric thinkers (the same way that 22–25 year olds managed by 29–30 year olds) drove the change at Shell.
The Future is Cognitive!
