История внедрения компьютеров в HVACR предсказывает современные дебаты об искусственном интеллекте

How HVACR’s First Computer Revolution Foreshadowed the Age of AI Long before generative AI, HVACR contractors were being told to question the hype, clean up their processes, calculate the true cost of technology In the fall of 1986, The ACHR NEWS ran an editorial that would sound oddly familiar to anyone watching the current debate about artificial intelligence. The question then wasn’t “Should contractors use AI?” but something much more basic: “Do contractors even need computers?” Back then, computers were still exotic machines – hulking, climate-controlled beasts attended by engineers wielding stacks of punched cards. One major mechanical contractor told The ACHR NEWS there were “only two contractors in the nation who could use a computer profitably.” But before the interview reached print, he called back after spending three days at IBM. He had ordered the company’s first computer and wanted to know whether there was still time to change what he had said. That’s how quickly the ground was shifting. As the archive shows, the 1980s were a decade when the HVACR industry confronted a surge of new technology with a mix of skepticism, excitement, and business calculation – much like the conversations happening now around AI. “Only Two Contractors Could Use One Profitably” In the early ’80s, computers were still largely the domain of big corporations, locked away in glass-walled rooms and operated by specialists. The idea that a small or mid-sized HVACR contractor should own one – let alone use it every day – was still far from conventional. But by 1986, falling hardware costs and the rise of off-the-shelf software had started to change the equation. An editorial from October of that year listed the new possibilities: bid preparation, takeoffs, detailed costing, engineering design, product selection, accounting, job costing, service operations – “the truth is that many of the nation’s smaller and mid-sized contractors now are using computers for some part of their operation.” “The computer is, after all, just another tool that every day is becoming easier to master,” the editors wrote. “It probably won’t take long before they’re hooked.” “Computer Does the Work of 3 People” Some contractors were already there. In Salt Lake City, Greg Pritchett of CSI Heating & Air Conditioning reduced his office staff from five to two, thanks to computerized engineering and bookkeeping. But what really stood out was how CSI used data – not just to keep records, but to spot patterns and solve problems. RESULTS: As computer adoption accelerated, The ACHR NEWS urged contractors to look past “fancy language or blinking lights” and evaluate technology based on business needs, implementation costs, and results. (Courtesy of Chris Gray/BNP Media) The company’s system tracked callbacks, warranty issues, employee training, and even which circuit boards were failing most often (twenty service calls on one part made it “highly suspect”). Sales leads, technician performance, customer histories, inventory, fleet efficiency – even utility bills – were all logged and analyzed. When one month’s utility costs spiked, CSI found a faulty meter. When propane expenses crept higher, they discovered a leaking truck. Looking for quick answers on air conditioning, heating and refrigeration topics? Try Ask ACHR NEWS, our new smart AI search tool. Ask ACHR NEWS Viewed from today, CSI’s system performed functions that would later become familiar components of customer relationship management and field-service software. But in 1985, it was something new: using operational data to identify problems and make management decisions. That basic proposition remains at the heart of contractor technology today: collect better information, find patterns that would otherwise be difficult to see, and use them to make better decisions. HVAC Gets Its Own Computer By the mid-1980s, computers weren’t just for the back office. They were becoming HVAC-specific tools. One 1986 article spotlighted Hvac-Calc, a handheld, battery-operated computer with a four-color printer-plotter, designed for contractors. Built around ACCA Manual J, it let users run whole-house or room-by-room load calculations in minutes, graph out temperature distributions, and compare energy usage for different systems. This wasn’t just about speed. Contractors could bring calculations, operating-cost comparisons, and projected energy savings directly into the sales process. The computer was no longer just a business tool; it was becoming a selling tool. Should You Even Buy One? But even as the industry was swept up in the computer boom, a note of caution persisted. “A computer is only as good as what you put in it,” Craig Bernhardt of The Lee Co. warned. “Garbage in, garbage out.” Before automating, contractors were urged to get their house in order – fixing processes, clarifying goals, tightening up management. Technology wouldn’t fix a broken business. That warning translates easily to the AI era. A faster tool does not fix bad data, unclear processes, or a business that hasn’t decided what problem it wants technology to solve. Some experts were even blunter: “In fact, computers really aren’t for everyone,” said Donald A. Turner of Management Computer Services. Contractors were asked to consider their annual receipts, business size, and satisfaction with current systems – sometimes the answer was, “then he doesn’t need a computer right now.” A 1986 package in The ACHR NEWS laid out “The 10 commandments of computer buying”: “Know thy business,” “Do not hold false hopes,” “Thy patience shall be rewarded,” “A bargain beckons only fools,” “Weigh thy options,” “Sustain thy computer,” “Embrace the future,” “Fear not tradition,” “Thou shalt take charge,” and, finally, “Thou shalt make thine own blunders.” The explanation under the last commandment was even more direct: “Computers do not make mistakes. People do.” Don’t Get Hung Up on Blinking Lights Buying technology, contractors were told, was a business decision, not a technological one. Charles Mendenhall of Digital Communications Associates put it plainly: “Don’t get hung up on fancy language or blinking lights. Ask yourself, ‘Is this relative to my business? Will it solve my problems?’” Replace “blinking lights” with today’s AI terminology and the advice barely needs updating. Contractors are once again being asked to evaluate unfamiliar technologies promising major gains in productivity. Mendenhall’s question remains the useful one: Will it solve my problems? The ACHR NEWS archive is full of practical advice for evaluating vendors, calculating hidden costs – personnel, training, supplies, maintenance – and planning for the long haul. “Computers don’t run on air,” one story quipped. A $25,000 system could easily end up costing twice that when all was said and done. Contractors were warned not to expect everything to work immediately; one guide cautioned that it could take “a year or more.” And yet, the real success stories emerged from contractors who rolled up their sleeves and made the technology work for them. Marcus “Butch” Metoyer Jr., in Lansing, Michigan, bought a small PC after seeing a demonstration, learned BASIC, and wrote software for heat-pump savings calculations. “Now we wouldn’t know what to do without a computer,” he said – a far cry from the skepticism of the early ’80s. Artificial Intelligence Enters the Conversation By the late ’80s, the conversation was starting to shift once again. William G. Beazley, writing about computer-aided design, warned firms not to be dazzled by “attractive media and flashy demos.” Instead, he urged them to focus on true needs, costs, and outcomes. Beazley pointed to “expert systems,” described in the article as artificial-intelligence technology for imitating decisions usually made by experts. These were rule-based systems built around knowledge bases, not today’s generative AI. Still, the promise was significant: “Expert systems offer an order of magnitude productivity increase in automating engineering tasks,” Beazley wrote. The technology was different, but the ambition should sound familiar: capture expertise, automate portions of expert work, and allow engineers to accomplish more. Beazley predicted these systems would soon be used for heat exchanger design, piping, and HVAC. In other words, the idea of using AI to extend engineering expertise was already appearing in the pages of The ACHR NEWS nearly four decades ago. The More Things Change... The questions that surrounded computer adoption in the 1980s sound remarkably familiar in today’s AI debate. Does the technology solve a real business problem? Is the company organized well enough to use it? What will implementation really cost? How much training will it require? What happens when bad information goes in? And who remains responsible for what comes out? The machines have changed. The buying principle hasn’t. In 1986, the smartest advice was to ignore the blinking lights and focus on the business need. Forty years later, that’s still the best place to start. Looking for a reprint of this article? From high-res PDFs to custom plaques, order your copy today!