My colleagues found a way to raise my blood pressure this morning. I woke up and saw that our resident car designer, Adrian, had thrown a link into Slack going to an article about some random woman on TikTok who somehow found her car dead after changing a sun visor. She says her husband used AI to try to fix the car and probably wasted a bunch of time. Everything about this article and the whole situation is insane, and I have to say something. You can’t trust AI to fix your car without constantly fact-checking it, and you’re better off just logging into an old-school car forum or a Facebook group for your car.
This story comes to us from Motor1. The publication recently debuted a “Trending” section, where some of the site’s writers will comment about car-related stories from TikTok. Many of these stories involve people not knowing that U-Haul charges fees for truck mileage, people not knowing how car financing works, or people not knowing how car dealerships work. If you want to consume TikTok without actually watching TikTok, these posts seem to do the trick.
Normally, these posts aren’t really our style, and that’s fine. Every site has its own vibe! But this piece caught our attention because of the fantastically terrible way that AI ruined this poor woman’s day, and how seemingly nobody noticed the obvious path of diagnosis.
The Conundrum
@sarahnovakwallace Ladies if your hubby is a mechanic or works for @Chevrolet or @General Motors HELP. I can’t get my car to start #chevrolet #gm #generalmotors #mechanic #tahoe
First, I’ll give you the TikTok. If you cannot watch the embed above because you’re at work or have TikTok blocked, click this link. You should be able to view the video without creating a TikTok account.
In the video, TikToker Sarah Novak (@sarahnovakwallace) talks about how the driver-side sun visor in her 2020 Chevrolet Tahoe broke. She ordered a new OEM part, and her husband installed it. The next day, Sarah attempted to drive her Tahoe, only to hop in the driver seat, attempt to start the vehicle, and find only a Christmas tree of lights and no start.
Sarah and her husband were immediately stumped and turned to AI, which told her that if you mess with any wiring, the vehicle’s various modules might stop working. So, her husband disconnected the battery for some time, hoping to reset whatever computer was angry. This didn’t work, and the next time Sarah attempted to start the SUV, she got exactly nothing out of the vehicle. Sarah continues by saying: “We’re ChatGPTing it, but I don’t know if we’re doing it right or what the problem is.” Then she says she only recently spent $400 on a brand-new battery.
Sarah’s car still wasn’t working when she published the video on December 29, and she was so desperate that she reached out for help on TikTok. Alright, so the stage is set. Here’s what happened next.
AI Fails At Diagnosis
Motor1‘s article doesn’t start off great, as whoever is in charge of graphics at Motor1 used a photo of a third-generation Tahoe, when Sarah’s Tahoe is a fourth-generation model. I’ve noticed this issue in a lot of TikTok articles on various websites. Anyway, the writer’s response was this set of paragraphs, from Motor1:
To many drivers, a sun visor feels like a purely cosmetic part. In modern vehicles, especially full-size SUVs like the Tahoe, that assumption can be misleading.
According to General Motors service documentation and industry explanations from the Society of Automotive Engineers, late-model GM vehicles rely on multiplexed electrical networks, such as CAN and LIN, to enable dozens of control modules to communicate. Components in the headliner area can share wiring paths with airbag systems, interior lighting, and the body control module, or BCM, which manages functions like starting authorization, door locks, and dashboard displays.
If wiring is pinched, grounded, or left partially disconnected during an interior repair, the BCM can detect a fault and disable vehicle startup as a protective measure. Automotive electrical experts frequently note that once a BCM fault is logged, simply disconnecting the battery may not restore functionality, as many modules retain error states until they are cleared with a diagnostic scan tool.
The writer then talks about how battery disconnects could make the no-start condition worse before dropping this nugget:
That distinction helps explain why Novak’s nearly new battery was unlikely to be the root cause, even though battery failure is a common suspect in no-start situations.
But here’s the thing: there’s no reason to just rule the battery out by default. Just because a battery is new does not mean that it cannot be drained. If Sarah’s husband messed around for too long with accessory power on or accidentally let the visor’s light stay on all night, those alone could drain a battery regardless of its age. Replacing a visor in one of these SUVs isn’t rocket surgery, either. Watch this:
Likewise, the only electrical connection that the mirror has with the rest of the car is for its light. If you watch the video above, you’ll see just two wires and a plastic connector for them.
One comment in the video reflects this:
The terrifying thing is that, depending on how you word the question, AI will give you some entirely unhelpful answers. Here’s what Google AI gave me after typing in “2020 chevy tahoe won’t start after sun visor replacement.” It went straight to assuming something is wrong with the vehicle’s Body Control Module:
Your 2020 Chevy Tahoe likely won’t start after a sun visor replacement due to a disconnected or blown fuse for the Body Control Module (BCM) or ignition, a loose battery connection, or potentially disrupting the ignition switch wiring, common after interior electrical work; check the main fuse box (under the hood and driver’s side) for relevant fuses like BCM or ignition, ensure battery terminals are tight, and if issues persist, consider resetting the BCM by disconnecting the battery for a few minutes.
Possible Causes & Solutions
Blown Fuse: The wires for the sun visor lights might have shorted or been pulled loose, blowing a fuse.
Action: Check the fuse box under the hood and the one on the driver’s side (left of the steering wheel) for blown fuses related to the BCM, ignition, or interior lights.Loose Battery Connection: You might have jostled the battery or terminals during the process.
Action: Ensure battery terminals are clean and very tight.Disconnected/Damaged Wiring: The wires for the visor might be connected to a circuit that also affects starting or dash power.
Action: Re-examine the visor’s wiring and the connection points in the headliner. Ensure no wires were pulled or damaged.BCM Malfunction: A brief electrical issue can sometimes put the Body Control Module into a protective state.
Action: Try disconnecting the battery negative terminal for 10-15 minutes to reset the BCM, then reconnect and try starting.
ChatGPT was equally unhelpful:
A no-start right after replacing a sun visor on a 2020 Chevy Tahoe strongly suggests something electrical was disturbed. The visor area is tied into several sensitive systems (airbag, BCM, interior power), so this is a common scenario. Here’s a clear way to narrow it down.
I won’t bore you with the details, but ChatGPT gave me a list that started with checking fuses related to the airbag system, the BCM, and interior lighting. Then it said that the visor replacement could have disrupted the anti-theft system or the airbag system. To its credit, ChatGPT did tell me that a drained battery could be the cause, but rated it as unlikely. If I logged out, logged back in, and gave ChatGPT the same question, it gave me a slightly different answer. One of the AI’s suggestions was to replace a fuse even if it is only slightly blown. Is a slightly blown fuse like being slightly pregnant?
I would not be surprised if AI gave Sarah’s husband an answer closer to what Google AI gave me, considering he tried to reset the vehicle’s computers by disconnecting the battery. The problem I had with this is that AI has no idea how to diagnose or how to troubleshoot. It can only take what it can search and spit it back out. Thus, poor Sarah and her hubby are chasing around computer issues without even having checked the easy stuff.
To be fair to the AI, it didn’t give the most insane response; that would go to some of the commenters on TikTok, who told Sarah that she fried the SUV’s Powertrain Control Module (PCM), told her to reflash the anti-theft system, or told her to reprogram the key. What’s with the suggestions for nuclear options after basically no troubleshooting?
Try The Easy Stuff First
Here’s what I would have done. First, I would have checked the battery. If it’s below 12 volts, or maybe the clamps aren’t tight, that’s probably the problem. A multimeter is only $8 at Harbor Freight, and anyone who works on their own car or their own home should own one, anyway. A dead battery would cause Sarah’s issues with flashing dash lights, no start, and then, later, a totally dead car.
If the battery tests good, uninstall the new visor. Use a flashlight to see if you somehow penetrated a wiring harness when you installed the visor. This is unlikely if you used the OEM screws, but it is still possible. It’s also worth checking fuses.
It’s possible that the no-start condition is not even related to the visor at all, and it’s just a wild coincidence that the Tahoe is not starting a day after the visor’s replacement.
I wouldn’t be sweating about the PCM, the BCM, or the anti-theft system until after I tried the easy stuff, like checking the battery. It wouldn’t be my first place to look. It would be like jumping straight to the assumption that you’re going to die just because you have a new cough. But this is one reason why AI cannot be trusted for wrenching unless you double and triple-check it. At that point, you’re sort of better off without it. AI has no nuance or context. It has no idea what it’s doing.
Just to be clear here, I’m not blaming Sarah or her husband for anything. They did the best they could with what they thought was a reliable resource. Not everyone spins wrenches or gets into car diagnosis, and that’s okay! I reached out to Sarah and will update if I hear back. Sarah has not published a follow-up, so maybe she and her husband figured it out.
If you have a problem with your car, don’t ask AI, and be careful soliciting opinions on just regular social media. Go to a make or model-specific Facebook group or an old-school car forum and just pitch your question there. You’ll almost certainly get better advice than you’d find on TikTok, and certainly better than AI.
Top graphic images: GM; DepositPhotos.com
I use chatGPT for work and for helping me organize my writing, but don’t trust anything that it creates. I always ask for it to ask me questions and then organize my thoughts and proofread. Occasionally for fun I ask it random questions that I know the answer to, like how many splines on a 5th Gen Camaro axle. It gave me the wrong answer, I corrected it, ChatGPT apologized and then gave me another wrong answer just after telling it the right one. It can be a powerful tool some times…
I’m pretty sure that Motor1 articles are mostly AI generated from TikTok videos.
According to https://www.motor1.com/news/772160/safely-remove-damaged-rim/ you can change a rim without jacking up your vehicle.
You shouldn’t trust or use AI for anything, actually
Couldn’t agree more
I have already dealt with people trying to argue with my advisors over stuff a LLM said it would be. One lady even tried to use ChatGPT to educate us on how to debug and code the audio unit in her car to fix her issue. It took my bosses several hours over a few visits to get her to leave us alone.
Yesterday I asked Google if it was possible to encode data on salami. The AI answer was, yes, that it had been done by Jason Torchinsky of The Autopian. AI sucks and The Autopian is definitely The Bent.
Also, why is the visor mirror light controlled by a computer and not just a switch? (not interested in an answer, just yelling at clouds)
It is actually just a switch, ChatGPT is a nonsense device.
It’s a space heater with extra steps.
I’m curious on what ai prompt she used to buy a $400 battery for a Chevy Tahoe?
No, no, no. The true problem here is why a 5 year old sun visor broke in a $45K+++ vehicle.
Here’s Why You Shouldn’t Blindly Trust AI
For Car Repair, Even For Seemingly Easy IssuesFIFY
I came here to write exactly this
My favorite is when you catch chatgpt giving bad answers and ask it for clarification, as in “shouldn’t I just start by checking the battery?”
The completely obsequious response is always something like “that’s a great idea. you’re smart to start with the simplest and most obvious solutions before jumping into more difficult ones.”
Uhh..okay…if it’s such a great idea and I’m so smart, then why didn’t you suggest it in the first freaking place??
Just last week I googled how to change the wiper blades for my model of Mercedes-Benz.
(Because the dealer put crap blades on my car during the last service in Nov – but thats another story…)
When I scanned the AI-slop answer, I immediately knew it was wrong.
And my husband wonders why I don’t trust AI.
Wait you mean to tell me this half baked crap being shoved down everyone’s throats against our will doesn’t work right?
Many billions are being invested in it. Grab some popcorn and watch the tech stocks in the stock market…
In the context of what I said, what exactly do you mean?
Those stocks will crash, fortunes will be lost, and it might be entertaining to see the inevitable as it unfolds.
Gotcha. I couldn’t tell if it was this, or “lots of money is being invested therefore it is good” – which, sadly, is a perspective I actually see sometimes.
Hell, not long ago, I asked an AI for information about a school group I was in back in high school (and it was supposedly one of the better, more accurate ones). It could not confirm that it had ever even existed. Meanwhile, I’m like, “But we were in the yearbook every year, and even on the evening news a few times.”
So yeah, AI is unaware of an actual documented experience I had in my own life, info on which has been publicly available since 1988.
Meanwhile, AI will invent a celebrity because someone wrote a post about their cat that did not specify in words that the pictured cat was not, in fact, a human celebrity, but a feline who was looking fancy enough for a red carpet.
I worked on a car once where the 3rd row power folding seats were stuck down. Long story short, the driver’s visor shorted the wires for the vanity light in the telescoping part of the hinge and blew the fuse for the second interior illumination circuit, the “smart” switch for the seats all the way in the third row ran itself off that illumination power despite having other good power and ground to it, and the other seat switches interact with it through LIN so no dice there either. Everything else related to the power folding seats was fine. Very frustrating.
Most “AI” now are Large Language Models. LLMs are LANGUAGE models, not information sources. They are a tool… for language.
With good, detailed, prompts:
If you want to draft or edit writing based on information you give it, it can do a great job on the grunt-work of writing, with human review.
If you want it to summarize other writing or large sets of text, it will do a decent job and be able to answer your questions about specific content.
It can even do some basic programming and debugging, given that programming languages have defined vocabulary and syntax, and typically follow patterns (again, with human review!)
Any so-called “facts” from an LLM are just based on statistical likelihood from the underlying training data. It doesn’t “know” anything. With a really good prompt, it might point you in a statistically likely direction.
Asking an LLM blindly for information is severely misusing the tool?
Don’t get mad at your hammer just because your using it to brush your teeth.
PS I hate how ubiquitous, and how stupidly LLMs are being used.
this!
But it’s being marketed as an ‘everything’ tool to the Executives of corporations and they in turn expect the employees to use it to be more productive. My Son is an experienced programmer. The tool is of little value to them, but need to show they are actively using it.
I think we’re in agreement. Your son’s experience definitely falls under my post-script of it being used stupidly and ubiquitously.
“Don’t get mad at your hammer just because you’re using it to brush your teeth” is something I’m going to be using from now on. Exactly right.
This guy LLMs 😉
Seriously, an excellent explanation of how LLMs actually work, and why careful prompting is so important to get any genuinely useful information out of one.
LLMs are also a fine example of another classic expression that applies to computers in general — Garbage in, Garbage out. An AI LLM is only as accurate as its input dataset and the prompt used to direct how it interacts with that.
They can be useful, when used appropriately and the end-user has some understanding and skill in prompting. But as a replacement for internet searching and as an all-purpose “assistant” by untrained end-users, they’re underwhelming and can easily go off-track.
Don’t forget that the input dataset includes all the false information on the internet.
And it’s able to use that to come to even more bizarre and false conclusions, complete with entirely made-up source material!
Spotting and fact-checking for AI “hallucinations” is essential.
Right on! I’ve been teaching college for nearly 30 years and trying to help my students learn how to productively use “AI” is an ongoing battle. As we used to say in the military: you have to be smarter than your tool.
The AI shows up as the first answer to the question no matter what you google, and it gives its answers as if they are a summery of real research on your question.
The blame should not go to the user for failing to realize the ever present and all powerful Google AI is putting out garbage. It should go to Google for replacing the search tool that people have been using for decades with an AI answer and endless Utube links.
It appears in the same space where for the past several years there had been a short summary of (usually) the Wikipedia page, which was generally trustworthy information
For years I saw people mocking those who used Wikipedia to get information on a subject. Many of those same people quote AI as the gospel truth.
If you add “-AI” to Google search it (mostly) does not provide AI responses.
That’s useful information – I will try it.
I keep trying to tell people that LLMs are an interesting and useful tool, but being pushed by some of the the worst people on the planet for tasks they’re not suited.
Sort of like voting for the orange felon…
These things are like fancy games of telephone. Think of all the bad information out there, it’s pulling its answer from much of that as well as anything valid and, as more wrong information gets put out, even by LLMs, it only compounds. Why not search directly for the answer from a primary or otherwise specialized, trustworthy source? That extra 5-30 seconds too long to assure far greater likelihood of accuracy? However, I think it’s less ridiculous that people don’t know what it is than that they seem to think it’s got some kind of awareness, which if it did, why would they trust it to not give the wrong information on purpose?
AI aside, this woman achieved exactly what she set out to do in changing her visor: exploit the scenario for maximum number of clicks and views. Well played.
Sadly, she might not have even gotten the clicks and views for her trouble. Content scrapers like the article and accounts that repost Tik Toks to Instagram or Reddit might have gotten the bulk. Once your viral video mistake escapes containment, you may well get the shame without the ability to monetize the fame.
AI gives me great answers on subjects I know nothing about and garbage answers on any subject about which I have extensive knowledge.
I trust social media comments only slightly more than AI. Like trusting gas station sushi slightly more than Bill Cosby’s mixed drink.
I might breeze the AI overview because it’s right there at the top, as mentioned, but I take it wish a grain of salt. I replaced a home light switch over the holidays and thanks to the good people at Grassroots Motorsports forum I learned the difference between a single pole and 3 way switch. I’m betting AI would have left that out.
Depending on which sites the LLMs search for information, “single pole” and “3-way switch” could result in some interesting answers, but it wouldn’t help you change an electrical component.
I recently enlisted the aid of ChatGPT to help me rebuild a Ford 8.8 differential. It walked me through all of the steps and verified information when I questioned it. It helped me look up data and measurements, as well as calculate shimming adjustments for me.
For someone who has only ever changed the fluid in a solid axle, this was like having another tech in the shop.
However, despite knowing the car it came out of, the hypoid gear type, the clutch type LSD, and the very well-published fluid requirements for ONLY 75/140 with friction modifier… it still told me to break in the all-new rear end with 3 quarts of 75/80. I would have burned up the LSD clutches at the very least doing what it suggested.
It’s NOT perfect.
You still need to do some thinking, brain is still a requirement.
AI only knows words – you HAVE to ask the right questions. It also helps A LOT if your questions are conversational instead of “differential carrier bolt torque Ford 8.8”.
That last part is something I still struggle with, having been trained so much on Internet 1.0.
The game back then was entering as many key words as possible to narrow down the results (or even a Boolean syntax). These days that gets you almost nothing, meanwhile people who were raised on natural language querying are just typing a stream of consciousness questions into Google — or ChatGPT — and getting amazing results.
It blows my mind because if feels like it shouldn’t work this way, but it’s calibrated to the masses and trained on their tactics. I still don’t know if this is a good thing.
My Google-Fu was top notch on Internet 1.0. In today’s age, it’s useless.
I have found that AI reacts and give much better answers using natural language than simple queries. I’m a very new user of AI and it has all but replaced using Google for me.
I haven’t gotten into AI yet because the slippery slope aspect — ie, I don’t want to become too dependent on it (for comparison, I’ve backed off on using Gmaps so I can force myself to regain my sense of direction, and I hesitate on web searches to give myself a little longer to try to remember things I used to know)
Secondary to that, I still have major concerns about power usage and water rights, so the “low-grade activist” in me doesn’t like the moral imperative/Golden Rule part of using it, much in the same way I don’t like corn ethanol. But like corn ethanol, it’s not going anywhere…
I’m the same way re internet 1.0 search training.
I’d argue it was better, as it forced you to think through the problem yourself and extract the sctual salient points, as compared to today’s just ramble your question and basically gamble if you get a correct answer.
100% agree, and secondarily it forced people to think logically and concisely. I still use those tactics when searching for old emails, for example.
It was more aligned to how general logic and coding worked, so I would argue it was designed to dovetail more directly into the broader computing world (naturally, since it was an offshoot of database querying).
I appreciate how far search prompts have come, but I don’t love it. The “old way” seemed faster, more logical, easier to hone, and more friendly to non-native speakers, too.
I’ll yell at clouds all day, but I try to provide a rationale, at least 🙂 The more we rely on computers, the less we know about computers.
I remember being taught BASIC, COBOL and Fortran, and later Pascal – yeah, I’m old – not to write some giant program (though I’m sure Torch has), but rather to learn how computers thought. Now, it seems the opposite…AI is being taught about us.
I used AI to find a part number for a tire in my car’s size on a manufacturer website but not listed on any sites that sell tires. That was really helpful. I also used AI to estimate lap times of tracks I’ve never driven on based on actual lap times from tracks I have driven. It was actually pretty fun trying to understand its level of intelligence. I gave it my modifications and tires and it got one track very close to what I expected and one that was very off. I asked it to account for the gearing in my transmission and it recalculated. Then I asked if it was accounting for braking zones which it said it wasn’t. So we recalculate again. It was doing very heavy calculations and at one point had to wait over 3 minutes for it to finish. I also hit my calculations limit and I had to finish the next month. Long story, short… it can do amazing things if you understand its limitations and can ask your questions accordingly.
I’ve done that and it works pretty well. Sometimes it can take a while to find it the old fashioned way. Of course, I confirm it’s correct, but it can definitely cut down time for those kinds of things.
You inspired me to ask Google AI, as an experiment, what the Nürburgring lap time would be for a 1958 Beetle driven in reverse. Among its stupid bullet points:
Yep, the old air-cooled VWs had rather delicate cooling systems…
Ha! I guess to be fair, high RPMs and low speeds would be worse case scenario for keeping the engine cool. A “cooling system”? Yeah… a big stretch. I used chatgpt, for whatever it’s worth. If you broke down the question a bit further, such as first asking what the top speed of said Beetle is while going in reverse and then asking it to use that speed to calculate a lapt time, it probably would have given a reasonable answer.
Some people have become too reliant on AI for anything. I gave a lecture to a college class a few months back and the professor told me he had been frustrated about papers being written by AI and submitted as original work. I laughed and thought how could it be…
I receive the extra credit reflections on my lecture. It was bad, it was very very bad. There were quotes from me that I never said, the grammar was horrible, and it honestly is what I expected from someone who only received the notes from the lecture and tried writing a 3 paragraph summary with no other context or information.
AI has its place, but currently fixing a car or writing a paper is not it.
Now you kids get off my lawn!
it’s actively making people dumber and less able to do any critical thinking at all. It’s shocking.
I see it’s uses, but people are generally using it in a manner it is not yet capable of doing. My wife keeps telling me to use it for more professional email responses, and that is more an ideal use, but still not one I am willing to do until I am reasonably confident it won’t make a mess of what I wrote.
*insert meme of Napoleon Dynamite zapping himself when trying out the time travel machine*
“Hold on, I forgot to code the sun visor before installing it”
I’m going to take some Black Ice Little Trees air fresheners to my local dealership and ask if they can code them to my car so I don’t get a check engine light.
Dang, when did Harbor Freight multimeters get to be $8? My first one was free with a coupon, and I think $2 retail. Damn inflation. *grumbling old man noises*
I’ve had so many of those free one’s leads break on me at the exact wrong time for them to decide to break.
I guess they’re okay for 12v, but I wouldn’t trust them with 120/220 ac!
They’ll have to pry my trusty Fluke 77 multimeter out of my cold dead hands 🙂
[Insert juvenile quip about having one’s fuse only partially blown a few times.]
OMG, this is a re-opened wound for me… I had an e-argument with someone once about the ‘3rd state of fuse’, i.e. partially blown.
Yeah, sure buddy. I’ve got a bridge in New York to sell ya.
The problem I’ve been having is, even though I know it’s likely crap, I can’t help but read the AI Overview since it’s right at the top and just too tempting. You inspired me to do a little digging and I just added an extension to Chrome that blocks it from showing. Hallelujah.
I love it when the AI Overview is blatantly wrong! Robots wont be taking over anytime soon.
AI responses are basically what the response you would get from the BS’ing Salesguy vs an Experienced Tech.
Sometimes, swearing results in Google not being able to provide an AI overview. I’ve taken to adding to each Google search “you fucking clanker.”
Be worried when it responds with Roger Roger. We are not advanced enough in cloning to counter that
Ok, I can;’t believe I’m standing up for AI here but..how should I say this….I don’t think the AI is the problem in this situation. Personally I’ve found AI really helpful for car troubleshooting as long as one has enough enough knowledge to recognize what information you need to give it and recognize when it’s confidently barreling toward a wrong conclusion. Better prompts also help. When I fed this one into my favorite LLM, it gave me what seemed to me to be a pretty sensible plan of attack, starting with the battery and connections, “You’re an expert Chevy mechanic who gives great advice to other drivers in online Tahoe forums. Help me troubleshoot a friend’s 2020 Chevy Tahoe. The driver’s side sunvisor recently broke and they replaced it with a new OEM part. The next morning they got a number of error lights and no start. Then they tried disconnecting the battery for a period for a reset, after that they car appeared to be completely dead. Give them a simple plan for troubleshooting, taking into account both the likelihood of the fault and the ease of testing/resolution. They are very basic DIYers.”
Lots of truth here… most people really don’t understand that the quality of output depends a lot on how the question is asked.
100%. If you put in “Why did installing a new sun visor kill my Chevy Tahoe” it goes straight for the fuses and electrical system. I think there’s a ton of reasons to be very worried about AI and what using it will do to our society but a lot of people’s bad experiences come from the companies encouraging people to use LLMs in ways that don’t really work very well.
Yup. And their incentive is to push how easy and “smart” it is, when they have a moral (and ought to have a legal) obligation to educate and warn their users about the risks.
AI is too naive to include in its reply, “Poor GM engineering” as part of the answer.
LOL. I prompted it with the same prompt but added “You’re an Autopian commenter” and asked it to diagnose, it said this: “You’ve angered the spirit of Bob Lutz and your Tahoe is now staging a work stoppage until you install the correct OEM-specification visor with the proper RPO code.” People are still much funnier thank goodness….
Did the AI give a helpful answer to that prompt? And yeah, I can see how it can be useful if you either a) already have some base knowledge or b) use it as a starting point. Ie: if you have absolutely no idea where to begin, ask it and when it tells you it’s the body control module, google “sunvisor replacement body control module.” You’ll either get a lot of results, indicating that there may be some truth there, or 0 results, indicating that it’s nonsense.
Problem is, there is no mandated AI School where we can learn how to use these tools effectively. Just a bunch of tech companies assuring us how easy and great they all are. So…on the whole we’re screwed.
I thought it did! I don’t really know computer-era cars but here’s the start (won’t post the whole thing as too long): “BOTTOM LINE: Most likely the battery terminals weren’t reconnected properly when they tried the reset. Start there – it’s the easiest fix and explains why the car went from “error lights” to “completely dead.” Here’s the troubleshooting plan:
STEP 1: Check Battery Connections (5 minutes)
STEP 2: Test Battery Voltage (5 minutes if they have a multimeter)
……………………it continues to go on to fuses and etc……
that all sounds reasonable and also low effort. As in, even if it’s dead wrong, you’ve lost 5 minutes and 0 dollars so who cares. I guess it also takes some common sense (“maybe I should do a little more research before I buy this $800 replacement part it says I need”).
No one who is uses search tools (but not AI) is going to know to word the search in the form of a prompts like “You’re an expert Chevy mechanic who gives great advice to other drivers in online Tahoe forums. Help me troubleshoot a friend’s 2020 Chevy Tahoe.” That may work, but it is completely non-intuitive for most people.
AI is the problem if it is going to be used to generate the primary search result without any added asterisk or explanation on why you need to question the result or add more detail to the question.
Totally agree with you. Caveat is I also think it’s a helpful tool when used properly and would be a shame if we didn’t learn to use it better to fix our cars. It’s kind of like if some company suddenly mailed blowtorches to all of us with a note that said “use for every household activity, no instructions needed!”–that’s a terrible, irresponsible act, but it doesn’t mean that blowtorch wouldn’t be really helpful getting a rusty part off my car if I learn when and how to use it.