ChatGPT, the powerful AI tool that’s making headlines around the world and has already changed the way some businesses accomplish work tasks, has made an insightful revelation about the best way to catch errors on utility bills.
But first, some historical context.
Utility bills have been a source of customer angst and confusion for almost 150 years. The historic Pearl Street Generating Station began operating in Manhattan in 1882. Thomas Edison is credited with building and operating the first electric utility grid, but seldom honored as the father of the modern utility bill. (My guess is that he preferred that Nikola Tesla take the heat for that.) The first payment received by Edison Electric Light Company is of historic value—a $50.40 payment from Ansonia Brass and Copper Company to Edison which included an annotated $1 bill memorializing the historic payment, a bill sold at auction to a collector for $15,000 in 2014.
Fast forward 100 years and billions of utility bills, I observed in 1980 as an energy management engineer that utility bills were fraught with errors: improperly applied taxes, suboptimum rate schedules, duplicate bills, bills for facilities that had been sold, unrealized opportunities for demand and usage charge reductions and a host more. My conclusion was that great potential value existed in auditing and validating utility bills, but to do so required non-existent energy management software and professional bill processing expertise.
So, I set about developing energy management software on my new Radio Shack microcomputer with a blindingly fast 4 MHz processor, 64K of RAM and two 8” floppy disk drives. Within a year the first generation of EnergyCAP utility management software was released and the customer stories of utility bill savings started to roll in.
Fast forward again, 40 years this time, to the 2020’s AI revolution and the mysterious intelligence of ChatGPT. I wondered what ChatGPT’s insightful revelation would be today on the best way to spot errors on utility bills. ChatGPT said: