On Feb. 22, 2024, the entire network for AT&T crashed. After it recovered, there was some kind of glitch between my AT&T cell phone and United Airlines. For more than 50 flights, messages sent to me from United Airlines’ system worked in some cities and not in others. During three trips out of the United States, it worked perfectly. Something, somewhere inside certain portions of the AT&T network failed to completely recover. When I discussed this with computer specialists, they thought it was concerning. What worries me is, could this same thing happen to supply chain management systems? Maybe they would work in one city and not in others.
After the Microsoft crash, it appeared that the scheduling system at one major airline only partially recovered. For several months, the signs at the gates stopped updating when a plane was late. On several flights, we were told to board when there was no plane at the gate. Delta Airlines was a big mess after the Microsoft crash and had to cancel flights for several days. Another glitch I have experienced is that when I receive calls on my AT&T cell phone from a company that uses Microsoft Teams for phones, it shows up as a scam call. Calls I have received from university phones are now tagged as scams because the university uses Teams for phones.
The final frustrating glitch was when I ordered Christmas fruit boxes from Harry and David. I have successfully ordered from them for over 25 years. When I first called an agent, she could see my gift list, but she was unable to transfer it into the new computer system. The order was not completed. Several days later I called a different 800 number that probably went to a different call center and everything worked perfectly. My gift list appeared in a form the agent could easily use.
Root cause possibilities
The original Microsoft Windows programs up through the 1990’s were all based on the old DOS system. It is likely that some of this ancient code is still in modern systems that are being used today. New interfaces and artificial intelligence systems may be getting layered on top of antique code that young computer programmers may not understand. The old programmers who understood it, are retiring.
The problems I have discussed were not caused by cyber-attacks. Both of the big crashes were caused by internal errors by programmers. I am concerned that these errors may increase as the old programmers from the 1980’s retire.
You need to think about it. What is your back-up plan if these systems go down? Could your plant operate? One simple approach for trucking logistics is to send your regular customers the same order they had last week. Tell them to not contact you unless last week’s order is totally wrong. You have to be ready in case the old DOS system causes problems. If an artificial intelligence system is layered on top of it, it may get really confused if it gets conflicting information from computers located in different cities.
Some people questioned me several years before the COVID outbreak, when I said, “big is fragile.” COVID proved me right. This article is a warning: If one of your computer systems crashes, it may not fully recover. It may work in some locations and not in others. Another problem I have observed was after a system crashed, it would work on a new laptop, and it failed to work on an older desktop. How will artificial intelligence react to this? That is the million dollar question.
