I called my normal HVAC company for my rental home because the tenant reported the AC wasn't cooling the house. When I called, I got one of the latest AI voice assistants to help me, and it was an awful experience and I ended up not hearing back after the assistant told me the office would call me back.
So, I went over to the house and used ChatGPT to help me diagnose the issue by taking some photos of the compressor panel outside. It walked me through what to check, I provided some diagnostic codes I witnessed... and it walked me through the very simple repair of replacing the $25 capacitor. It was going to cost me almost 4x that just for the service call to diagnose what was wrong in the first place.
So, the weird experience was: Gen AI made me lose trust in my normal HVAC company, and more Gen AI basically allowed me to replace my HVAC company and do the repair myself all in one day.
Like the time I had one of the bigger shops in town come by to get a quote for replacing a dual stage fan motor on an AC. The tech asked me if I'd like them to replace the contactor while they were in there because it is a part that often fails. I asked what a contactor was and he explained it. "Oh, like a relay?" I asked. I told him to quote the cost for "replacing the contactor, while they're already in there."
He quoted me $400 for the contactor, $750 for the fan. The contactor itself I later found out was was $7. I literally laughed in his face when he said that.
So, like I said, you just aren't going to get professional level assistance from an AI. Thankfully.
To end the story: one of the other guys I called for a quote on fixing this unit repaired it for free; the unit was still under warranty and it was fully covered. The original installer of this $12K unit was refusing to return my calls. Another "Not gonna get pro level service from an AI" story.
With an AI, I can say "I don't understand that part, can you explain more?" Or "what about this concern I just thought of", or "I already know almost enough about this, I just need this one gap filled in." It's an objectively better experience.
However, neither AI nor (most of) the videos can help with finagling frail snap-fit assemblies one encounters in appliances. A lot of appliance repair work is very simple but requires significant practice and figure-it-out time, in addition to waiting for sketchy parts from Ebay. Half the battle is just finding the damn parts.
As words in a dictionary are sorted alphabetically rather than phonetically, this is unhelpful.
YouTube videos have the same kind of problem, in that you can only easily find the video explaining which dielectric unions suit your problem when already know what those are (to use an example that I had to ask ChatGPT for because I have no plumbing experience even if I did know about galvanic corrosion and therefore immediately understood why they're important once I saw the name).
In 2009 or so a projector at some event that needed one wouldn’t start, and I noticed it was flashing a pattern, so I found a computer and internet connection (both very slow), painfully found and downloaded the manual for that model, and identified that it was saying the fan wasn’t starting. Lo and behold, a strut was broken and obstructing the fan blades, and bending it out of the way fixed it, and the event was able to begin.
I’ve found manuals for a drawbar organ, multiple digital pianos of different ages and brands, AC split systems, and more. Manuals are good stuff. They don’t cover everything, but they’re very useful.
For these sorts of things, AI is doing approximately nothing for you: you would do better (and learn more!) finding the actual manual, or you’ll want to see someone doing the thing in a video.
Tbh, I think people feel more comfortable asking an AI. Even though I “know” it’s all smoke and mirrors, I still prefer the human-like interaction to the grind of watching video after video and building my own understanding.
OOPS… there you see how it’s going to end. I’m the meatspace button-pusher.
I guess I'm seeing similar benefits to a novice programmer. Professionals would scoff at my work but they are expensive and difficult to work with. Meanwhile I'm getting the job done.
On the other hand I'm not touching AI for any development work. I'm too worried about my skills atrophying or not properly learning anything new.
It feels like there is precisely enough information to deduce each step. But only just enough miss one clue and you have something on upside down on step 7 that you won't notice until step 37.
I feel whoever makes them could probably make a wicked NY Times Crossword puzzle.
They’re also actually good if you know to follow them exactly: double check every side, every hole, every screw and you won’t go wrong.
Kind of a superpower to turn anyone with a bit of tech inclination and problem solving skills into an HVAC tech - not a very good one, but one with enough motivation to get the results you need
(Though that's also the kind of hands-on troubleshooting step/fix that a person could just google for and find pretty easily back before the internet got all fucked up.)
And that was awesome. Thanks, Google! :)
I don't know where the change happened. It certainly wasn't overnight.
Where Google used to be magical and other search engines quickly improved, it all kind of turned into shit.
It really seems that I was getting better, more-direct results from Altavista 30 years ago than I do with top-flight search engines today. (That's a deliberately low bar, chosen because Altavista wasn't even intended to be "good" back then. I mean, it started as just as a side project at DEC to demonstrate that their Alpha hardware was able to index the entire World Wide Web.)
So lately, I've been doing the same thing as you: I'm increasingly using ChatGPT to do this basic fact-finding stuff. In this way, it mostly operates the search engine for me, but it lets me drill down through a sea of terrible search results to find something useful fairly quickly.
It's still not great -- I still have to reject mountains of bullshit. But it's better than alternatives, and I can reject the bullshit with conceptual descriptions instead of trying to get Google to do what I need it to do (what it used to do).
It feels all wrong using an LLM to do this stuff, but whatever. I'm still getting stuff done.
Did you attempt to prompt it further into figuring out the actual problem, or know what they did to actually fix it? My bet is on a bad starting capacitor for the motor --- something that's a relatively cheap and quick repair.
It walked me through measuring refrigerant, subcool and superheat, pulling the vacuum, brazing the lines, exactly what tools to buy, I even input the numbers from the meter and it told me how much to add and so on. And this was with GPT4 or something far less intelligent.
In the past I tried to learn this stuff but the HVAC community are massive gatekeepers and try to hide information behind paywalls or spread FUD even though anyone could do it with the right tools and a little bit of knowledge.
I assume recorded videos and uploaded them in the Gemini phone on their app; and then probably said "what's wrong?"
Gemini is very good at those kinds of things. I recently got some ratcheting straps and needed to use them, but at the time I didn't know what they were called, so I didn't know what to search for on Google. I opened the Gemini app, pushed the button to take a picture (just like in text messages,) and included a message that was similar to "what is this and how do I use it?"
The exhaust blower not working triggered a safety that prevented the furnace from firing.
Spinning it bypassed the safety.
You likely inhaled a lot more carbon monoxide than you know.
Helping something start is not likely to ruin your day (unless you get caught in a rotating part)
"Spinning it to bypass the safety" is not a thing.
Please don't spread FUD.
If the exhaust fan couldn't maintain that negative pressure after the user stopped spinning it, the furnace would turn off again.
Their hack worked because the fan couldn't get the initial inertia up to speed (bad capacitor, dusty bearings, etc), but could maintain speed once it gets there. Have you never had an old home fan that would just hum when you turn it on but then work fine if you gave it the original crank? Same premise.
There was no risk here. If the fan didn't spin up to speed after that initial manipulation, and didn't constantly maintain the necessary flow, the furnace would have turned off again.
High hundreds of thousands feels like the upper limit before it would show up in statistically noticeable changes in patterns of deaths in some demographic.
High hundreds of individuals would still be "one in a million fatal errors over a few years", which seems better than I'd expect given I've personally had ChatGPT tell me that Solanum nigrum berries were "black tomatoes" (they're not usually fatal, but are a bit toxic, and no I did not eat them).
Imagine one of the models that has "accidental-deaths-via-bad-advice" just slightly turned up, with the model-provider's intent being to kill 5% more people per year.
“At its core, it's a small motor with a fan attached that has one primary job: to vent harmful exhaust gases out of your home before the burners ever kick on. This is the very first step in the heating sequence, and it's non-negotiable for a safe startup.“
As exhausting the combustion products is a critical safety feature, I would be surprised if any furnace was designed such that it could possibly keep running if the draft inducer motor stopped. It seems like it would be trivially easy to make a circuit such that gas valves could only open if the draft inducer motor + fan wasn't spinning.