>asked 4 different mechanics to fix a problem>they charged me an arm and a leg>didnt fix shit>ask chat gpt what the problem is>fix what it told me to fix>problem instantly solvedFuck mechanics. Love AI. Simple as
>ask chatgpt/google what the problem is>determine if I can do it or not>buy parts off rockauto>give quality parts to mechanic and cash, tell mechanic any vehicle specific tips that make it easier (ie: turn wheel left to get at it easy)>great job done every time for last 10+ years
>>29014298Wrenchlet
>>29014298>not getting your mechanical knowledge from cryptic imagery in your dreamsYNGMI
>>29014309Watching bad YT videos on an oil-smeared screen while losing your mind is a very similar experience.
>>29014301Interesting, do the mechanics not tell you to fuck clean off and call you a wrenchlet? I'd be a little miffed if some cuck came in with some cheap chink shit off the Internet telling me how to install it. Or is this a larp?
Decided to try this just for kicks. I asked what's causing excessive inner tire wear on my 8th gen civic's front tires (I already know it's the tie-rods). It told me that I either have excessive negative camber and need to get it realigned, or have low tire pressure. I've never heard of low tire pressure causing uneven wear, and this generation civic has no camber adjustment at all on the front. I think I'll stick with shop manuals and forum boomers.
>>29014313Installed an oil cooler recently, dude literally had it go 'fast forward' on the parts that were actually important, threw on fucking plumber pipe tape on AN fittings, while he talked at length about unrelated shit specific to his build between him and Tomioka.
>>29014330People who think they need to explain the entire world before getting to the fucking point are the worst.
>>29014298Relying on AI for solutions is like going to an ice cream shop for a vanilla cone with sprinkles and instead you get a vanilla cone with bacon bits.
>>29014324A very specialized AI system might even be helpful here, but those text bots basically work by associating and scoring sequences of words based on their training data, and spitting out words that score the highest (within the context of the sequence of words/tokens already present). With a large enough neural network to analyze the context sequence and score new words, the result can be surprisingly convincing English.The thing is, if it doesn't "know" or "understand" well enough, it just spits out the most correct looking words that might not add up to presenting correct information. The architecture doesn't really know if it knows something or not.These also mix up causes and effects, objects and subjects, etc. all the time.Maybe it somehow associated tires with random tire-related problems in your case.
>>29014324Low tire pressure wears both inside and outside the tire so it could be correct if you simply didn't know it was also wearing the inside. And camber can still be out on cars where it's not adjustable. It's still adjustable but requires more work that's not worth it and sometimes could cause liability issues.
>>29014375>The architecture doesn't really know if it knows something or not.So like most people