Anybody using A.I. to ask automotive questions?

[stealthee]

stupid rants of EXPERTS, not "experts"?

The only one ranting is you. You asked a question. I and many others answered your follow ups.

You're the only one upset because you don't like our answers. It's that simple.

 

[sacrileger]

I've gone back and forth on the idea of implementing AI to answer questions here in the forum.

thank you for your thoughtful and constructive reply... it is exactly the type high quality discussion i came here for.. what you said is spot on... all of it!

i would only add that there are various a.i. engine 'plug-ins' you would train yourself by feeding it with shop manuals and various other technical resources you know that they are reliable and you trust as being accurate... basically you would control the a.i. content yourself and make its output conservative, moderate, risky etc.... furthermore, if the answer was generated by a.i., you could put in a disclaimer that the answer was generated by a.i., 'proceed at your own risk or wait for some competent member of the forum to approve it'... you do have extremely knowledgeable guys here who are by far the best in this field and can spot b.s. answer from a mile away... your guys here would debug your automotive a.i. expert system in record time and your a.i. engine, because of the expertise you have on this forum, would be THE 'go to' automotive a.i. platform... you would probably be the market leader in this field... it's something to consider...

as an example of such an a.i. engine which you would train and control, i would use mike adams' a.k.a. 'health ranger' who trained his own a.i. (https://brightu.ai/) with resources he trusted... i believe he used 'DeepSeek' (dont quote me on it though.. it just behaves like deepSeek to my estimate)... i use his a.i. engine a lot... is it perfect? no! i have asked questions and the answers i got were patently inaccurate but still somehow correct but it's the best there is as far as holistic medicine and wellness goes.... i would say the accuracy of his a.i. engine is about 80% and i am sure that actual holistic professionals would rate it even lower.. but it is what it is for now... mike is making it better and better every day by feeding it with various data he vetted himself and finds accurate

thank you again for your knowledgeable, objective and analytical reply to my OP... all i wanted was to ask if members here would know what a.i. search engine they use, trust and recommend so i could test the a.i. they recommended for myself... from your answer, i gather there is not such an a.i. engine you would trust or recommend at this time yet.. on the other hand, there might be one if you decide to implement something similar to brightu.ai which you control and manage..

A lot more to consider than just the simple idea of whether or not AI has good answers or would be useful.

one more thing... if you would be interested, here's a link to a 'presumably' independent review of various a.i. platforms... DeepSeek is as competitive as openAi, googleDialog etc. but is the most affordable, according to the review... they offer various pricing plans and very likely reasonably priced 'startup' and/or trial/evaluation package to get businesses going:

DeepSeek: A Comprehensive Overview of the AI-Powered Conversational Platform

Introduction In the rapidly evolving world of artificial intelligence (AI), DeepSeek has emerged as a powerful and cost-effective solution for businesses and developers looking to build advanced conversational AI systems. Whether it's for customer support, virtual assistants, or personalized user ex

www.linkedin.com

p.s... i am not shilling for deepSeek and i have no stake in it... i am just trying to be helpful and based on mike adams' brightu.ai, i think he is also using deepSeek... i am judging it based on the a.i. engine's behaviour and output; it's pretty good imo

 

[1990TSIAWDTALON]

Look, I will not put up work your disrespect on this forum. There I said it. Respect is earned not DESERVED. I, and MANY OTHERS got our knowledge from practical experience. You got answers you didn't agree with or like and that's YOUR DAM OPINION.
I TOLD YOU QUIT STIRRING THE POT.
AI THAT!
Topic was discussed!
END.

 

[Ludachris]

that's the type of analysis and response i expected and came here for... not the other stupid rants of local "experts"

It's nice to hear you held this community in such high regard when you asked the question. It's too bad it didn't live up to your expectations. This is just one relatively small online community, but I suspect it probably is pretty representative of what you'll find pretty much anywhere else. A range of differing opinions that may or may not align one way or another on any given subject, fluctuating from group to group. It's the same everywhere - because it's full of humans, and humans tend to have different opinions no matter what community they're a part of.

To degrade the majority of the people here and scoff at those who didn't react the way you expected seems a bit silly, especially around the topic of AI, which a LOT of people are still pretty skeptical of - here and likely in endless communities everywhere. I tend to think it has a lot of potential to be useful, but it's still pretty early. If the subject matter of this forum were AI, then maybe the conversation would be different, maybe not. That doesn't mean people here are stupid. If you think they are based on their responses, then that just means people everywhere are stupid.

I see people get pissed off by their interactions in this forum and other forums/groups and they start making derogatory statements about the people in the group as a whole ("that forum is full of XYZ"), which always blows my mind. They take shots at the forum and the people in it (as if this forum is a singular hivemind or something) and then end up going through the same repetitive situation everywhere else they participate online. When you make fun of a forum like this using generalized statements, you're essentially just making fun of all groups of people everywhere, all society, because this group no different than others online and offline everywhere else. We come from all different backgrounds and walks of life, different education levels, different talents, different specialties, etc.

A wise man once said when you diss Dre you diss yourself.

 

[curt-s]

I've recently tested two major AIs by asking them to design a simple electrical circuit: a p-channel mosfet based soft start for charging some large caps to avoid large inrush current.

I used chatgpt and Google. Not only did they have to be corrected constantly but when they were they acted in the usual "you're right, that totally won't work". WHY would you even say it in the first place then?

Then later, chatgpt actually tried to lie to me saying that what I was asking was:

1) not possible, even though it created the design and told me it would work
and
2) the design was never meant to do the thing it itself designed to do it.

Even though it was the design intent from the start, told me what's neede to do it, and it chose the components specifically that would do the damn thing.
Using the exact words that it later said was not possible.

At one point, it told me that I should use a logic level PMOS, said "this one will work given the specifications", and gave me a part number. I looked it up and it was not a logic level PMOS -- RDS(on) was still 10V. I told it that, and it's all "you're right, that's not a logic level pmos according to the datasheet."

.. why would you say it is? Did you not look it up at all first? No, it said it was the correct part given the specs but did not actually look up the specs. So it lied.

It also asked, shortly before I literally told it that AI needs to die, if I wanted it to design a circuit using facts rather than feelings and hunches, or something of the sort. I can't remember what exactly it said but it was along those lines.

If I didn't understand what the components were already and be able to validate real-time, I would have been screwed.


So, really, the moral of the story here is the "AI" LLMs don't even believe their own story moments later. If LLMS can't trust what they themselves say, why should we?


No, I will not trust these things because they gather information from the millions of idiots out there and spew it as if it were fact. It just doesn't know those people feeding its database are idiots. There doesn't exist yet a large enough wealth of PROPER, ACCURATE information out there for a particular subject for LLMs to short-cut your way through. So it's not just about "choosing the right LLM". It's about choosing the right topic where little to no misinformation exists, because it can't correct it nor does it know what CORRECT information is. Automotive work where the database is likely built on public forums is definitely not a winner chicken dinner subject to get correct information on.

 

[1990TSIAWDTALON]

Good test! KiCad is a cool circuit design program that ISN'T AI.

 

[curt-s]

Good test! KiCad is a cool circuit design program that ISN'T AI.

Yep just started using that the last couple weeks. I was nearly going to buy a desktop CNC to try and mill PCBs with, so I started learning KiCad.
Instead, I just went back to proto board. Note: stay away from the green shit. I was getting 11.5V leakage around holes fed by 12V and was confused why my PMOS were turning on -- it was leaking to adjacent grounds. Through the paint!
The blue ones seem to be fine.

Anyway

 

[steve]

I'm thinking about converting one of my older 3D Printers into a (relatively) low power laser etcher so I can make PCBs by burning off a paint mask and then etch the board.

 

[1990TSIAWDTALON]

I was etching DVDs on my computer. I had to use special "gold" colored disks and flip them upside down to do it. I haven't done it in a long time, as my tower got hit by lightning and I never replaced it.

 

[steve]

Lightscribe was a feature on a lot of CD/DVD and Bluray writers. Sadly it didn't make me any better at labeling what was on one...

 

[curt-s]

I'm thinking about converting one of my older 3D Printers into a (relatively) low power laser etcher so I can make PCBs by burning off a paint mask and then etch the board.

I miss having access to this guy, an LPKF prototyping workhorse. Had the laser and the CNC mill.
It cut extremely fine lines into the copper to isolate non-conductive areas then heated the remaining copper enough to delaminate. A fan just blew away the bits.
Etchant will still have its limitations on thin traces.. not sure what the absolute width you could go is

You must be logged in to view this image or video.

Used it to create my 1G LED tail lights and controller/power supply PCBs like 13 years ago.

 

[miliman13]

I've recently tested two major AIs by asking them to design a simple electrical circuit: a p-channel mosfet based soft start for charging some large caps to avoid large inrush current.

I used chatgpt and Google. Not only did they have to be corrected constantly but when they were they acted in the usual "you're right, that totally won't work". WHY would you even say it in the first place then?

Then later, chatgpt actually tried to lie to me saying that what I was asking was:

1) not possible, even though it created the design and told me it would work
and
2) the design was never meant to do the thing it itself designed to do it.

Even though it was the design intent from the start, told me what's neede to do it, and it chose the components specifically that would do the damn thing.
Using the exact words that it later said was not possible.

At one point, it told me that I should use a logic level PMOS, said "this one will work given the specifications", and gave me a part number. I looked it up and it was not a logic level PMOS -- RDS(on) was still 10V. I told it that, and it's all "you're right, that's not a logic level pmos according to the datasheet."

.. why would you say it is? Did you not look it up at all first? No, it said it was the correct part given the specs but did not actually look up the specs. So it lied.

It also asked, shortly before I literally told it that AI needs to die, if I wanted it to design a circuit using facts rather than feelings and hunches, or something of the sort. I can't remember what exactly it said but it was along those lines.

If I didn't understand what the components were already and be able to validate real-time, I would have been screwed.


So, really, the moral of the story here is the "AI" LLMs don't even believe their own story moments later. If LLMS can't trust what they themselves say, why should we?


No, I will not trust these things because they gather information from the millions of idiots out there and spew it as if it were fact. It just doesn't know those people feeding its database are idiots. There doesn't exist yet a large enough wealth of PROPER, ACCURATE information out there for a particular subject for LLMs to short-cut your way through. So it's not just about "choosing the right LLM". It's about choosing the right topic where little to no misinformation exists, because it can't correct it nor does it know what CORRECT information is. Automotive work where the database is likely built on public forums is definitely not a winner chicken dinner subject to get correct information on.

I went to Lowe’s to look for a bathroom exhaust fan…. When I asked the first employee “ hey were is….”

They pulled out their phone and regurgitated the same words to the phone, great I know the location.

When I got the correct isle the employee there asked how can I help, I said

“ I’m looking for this type of device with this type of fixture”

This employee pulls out her phone and just ask the same thing..

It’s making people severely mentally incapable..

The lady said to me no this is the one the AI said so, I’m like lady look at the box with your own eyes !!!!!!

Again she insist she and the AI are right because it said it’s right.

This lady choose not to trust her own eyes , over what some perceived greater being told her.

This has been my experience with many other services, they argue with me about how the AI told them “ this “ and “that “……….

The worlds human population is going to approach new levels of Un-consciousness.

What sucks is when I was in graduate school a decent portion of my courses were in optimization, machine learning and general AI . I know how these things work, yet the masses treat this glorified
“if statements” as GOD.

Truly disgusting
Also they consistently lie and just say things to satisfy enquiry.

One of the machine learning projects I did was to create a smart obstacle avoidance using Q- learning, I used the space invaders game modified to avoid obstacles rather than shoot them, it did exactly that - by cheating it would move to the corners of the game were no obstacles are generated…… people who think these things are smart , usually aren’t themselves.

 

[Ludachris]

I went to Lowe’s to look for a bathroom exhaust fan…. When I asked the first employee “ hey were is….”

They pulled out their phone and regurgitated the same words to the phone, great I know the location.

When I got the correct isle the employee there asked how can I help, I said

“ I’m looking for this type of device with this type of fixture”

This employee pulls out her phone and just ask the same thing..

It’s making people severely mentally incapable..

The lady said to me no this is the one the AI said so, I’m like lady look at the box with your own eyes !!!!!!

Again she insist she and the AI are right because it said it’s right.

This lady choose not to trust her own eyes , over what some perceived greater being told her.

This has been my experience with many other services, they argue with me about how the AI told them “ this “ and “that “……….

The worlds human population is going to approach new levels of Un-consciousness.

What sucks is when I was in graduate school a decent portion of my courses were in optimization, machine learning and general AI . I know how these things work, yet the masses treat this glorified
“if statements” as GOD.

Truly disgusting
Also they consistently lie and just say things to satisfy enquiry.

One of the machine learning projects I did was to create a smart obstacle avoidance using Q- learning, I used the space invaders game modified to avoid obstacles rather than shoot them, it did exactly that - by cheating it would move to the corners of the game were no obstacles are generated…… people who think these things are smart , usually aren’t themselves.

Oof. AI gives Lowes and Home Depot more reasons to pay people less. The employees in the stores can now know even less about their own department, and it's okay with management.

I still remember the days when those stores used to hire retired tradesman. But when you have to keep increasing profits for shareholders, you have to cut costs, so now it's a bunch of people who know nothing and can't help you too much more than the employee at Walmart.

 

[miliman13]

Oof. AI gives Lowes and Home Depot more reasons to pay people less. The employees in the stores can now know even less about their own department, and it's okay with management.

I still remember the days when those stores used to hire retired tradesman. But when you have to keep increasing profits for shareholders, you have to cut costs, so now it's a bunch of people who know nothing and can't help you too much more than the employee at Walmart.

You know, now that you mentioned the guys- tradesmen who I got to know personally over the years, when receiving guidance from them in one of my many many home repairs, projects, ect.

All seem to be absent, making what you just mentioned seem more real and disturbing . It wasn’t until now that I thought about them……

another thing I noticed while finishing grad school, I spent a very long time in school, I’m a lill learning disabled so it took me longer.. about a decade plus.

In that time I worked with many people on lots of projects and my peers became less and less capable with time. AI didn’t hit all at once, years ago it was this app them that app. The point is towards the end you could tell who was using AI, and it was everyone.

I was very late to using it, I was told by my instructors to use AI to learn AI.

They told me at this point if you don’t use it your the only one that will fail, one instructor said if “3-5 people use AI, I can fail them, if only 3-5 people don’t use AI then I can’t the entire class.”

I make lill things for funsies, if you pick 10 random engineers, scientist ect I bet 9 of them can’t do anything by hand or by thought alone. The few people you’d think who would be able to create.

 

[DogWhistle]

If you think it’s hard for a man to say, “I don’t know”, it is apparently impossible for ai to admit it.

One interesting part about the OP’s original post is that he used ai because it would provide answers faster than waiting for a forum response. That’s the same argument the FacePlace DSM group members used to use. They finally got so tired of the WA guesses that were the responses, the DSM forums I see now barely ask for repair help any more.

Was on a forum for another flavor of car I own recently at 6am. It showed 6 members online and 989 guests. When I DM’d the Moderator, he said these were mining bots.

If a member doesn’t have the patience to wait a couple days for an experienced, helpful answer, a DSM is not the car for them.

 

[BLACK'98DSM]

Who remembers the days of a couple thousand site visitors here in one day
It wasn't even that long ago...

 

[SixBolt_16G]

It’s making people severely mentally incapable..

One of the Sr programmers noticed this after our "Hack week" a few weeks ago, ( In the tech field, It is where you have a week to collaborate with others to accomplish building out a small project, something other than work. This time, no AI allowed.) that AI has really helped some of the Jr programmers to come up to speed. They use it as a "learning tool". But for the Sr. programmers who are already experienced in their craft and are using it to increase production, its noticeably dumbing them down. I walked in on a conversation after the presentions on a Friday where several engineers were questioning why it was so hard to get anything done, with one of the engineering managers pointing out that its because they are no longer using their brain and that was his whole point of not allowing AI to be used in this round. Its happeniing on the design side too. Its killing the human creativity because everyone is relying on the AI to come up with the design and they, too arent exercising their brain. Increaed $$$ for corporate, detrimental to the human experience.

Concerns were brought up in a meeting that we are relying on this too much and are using it as a reason for laying off people who have the right knowledge to hire mindless human drones who dont know what they are doing, to manage things that they dont fully understand at a quarter of the price. Talking to upper management is impossible because they assume that everything AI does is right based on hearsay. They hadnt noticed in one of the images that they generated themselves, that the word wholesome was spell

You must be logged in to view this image or video.

Not saying that AI isnt useful or isnt great at a lot of things, but the way people look at it like its God is very concerning. What happens when its not a spelling mistake, but a programming mistake that could cost the company or clients millions in damages, because no one is checking and their are assumptions made that what it did is correct. Just like EVs, there hasnt been enough time to understand where and how it should be used as a tool, along with its implication, without everyone just hopping on the bandwagon to be "a leader in technological advancement". So it raises the question, can i always trust how it is telling me to change something like the timinig belt on my car? There is a lot to lose if it is wrong while assuming that its right, but not having the knowledge to know the true answer. Especially now that Ive gone through several instances of it leading me in the wrong direction for hours to days and then finally telling me that something cant be done the way that Im trying to do it. It lies to you more than you think.

 

[miliman13]

One of the Sr programmers noticed this after our "Hack week" a few weeks ago, ( In the tech field, It is where you have a week to collaborate with others to accomplish building out a small project, something other than work. This time, no AI allowed.) that AI has really helped some of the Jr programmers to come up to speed. They use it as a "learning tool". But for the Sr. programmers who are already experienced in their craft and are using it to increase production, its noticeably dumbing them down. I walked in on a conversation after the presentions on a Friday where several engineers were questioning why it was so hard to get anything done, with one of the engineering managers pointing out that its because they are no longer using their brain and that was his whole point of not allowing AI to be used in this round. Its happeniing on the design side too. Its killing the human creativity because everyone is relying on the AI to come up with the design and they, too arent exercising their brain. Increaed $$$ for corporate, detrimental to the human experience.

Concerns were brought up in a meeting that we are relying on this too much and are using it as a reason for laying off people who have the right knowledge to hire mindless human drones who dont know what they are doing, to manage things that they dont fully understand at a quarter of the price. Talking to upper management is impossible because they assume that everything AI does is right based on hearsay. They hadnt noticed in one of the images that they generated themselves, that the word wholesome was spell

You must be logged in to view this image or video.

Not saying that AI isnt useful or isnt great at a lot of things, but the way people look at it like it’s God is very concerning. What happens when its not a spelling mistake, but a programming mistake that could cost the company or clients millions in damages, because no one is checking and their are assumptions made that what it did is correct. Just like EVs, there hasnt been enough time to understand where and how it should be used as a tool, along with its implication, without everyone just hopping on the bandwagon to be "a leader in technological advancement". So it raises the question, can i always trust how it is telling me to change something like the timinig belt on my car? There is a lot to lose if it is wrong while assuming that it’s right, but not having the knowledge to know the true answer. Especially now that Ive gone through several instances of it leading me in the wrong direction for hours to days and then finally telling me that something cant be done the way that Im trying to do it. It lies to you more than you think.

It’s unfortunate that it’s occurring in that kind of setting, as you know engineering ( academia ) is very demanding and unforgiving a minor mistake can cost you the whole semester on a 1 problem final using only mental knowledge .
Now look at us .


It’s useful, but yeah - more time - more dumb - less creativity. AI feeds on human knowledge, I suppose this cycle is unsustainable, and will decay.

Or skynet idk

 

[Ludachris]

So it raises the question, can i always trust how it is telling me to change something like the timing belt on my car?

Well, people have been trusting others in forums for years, and we've dealt with misinformation through it all too. Maybe it's more risky with AI, but by how much? In some industries, where there is much more documentation in online databases, it will be less risky. Other situations it will be more. The nice thing is, you can always check the sources it's using to see if you trust those sources.

I just asked ChatGPT how to change a timing belt on a 1g... it gave me an overview from the factory service manual:

What You’re Working With​

You must be logged in to view this image or video.

You must be logged in to view this image or video.

You must be logged in to view this image or video.

4

Engine: 2.0L turbo 4G63
Drivetrain: AWD
Timing components:

• Timing belt
• Balance shaft belt
• Hydraulic tensioner
• Tensioner pulley
• Idler pulley
• Water pump (replace it — don’t cheap out)

---

High-Level Game Plan​

1. Tear Down (2–3 hrs)​

• Disconnect battery
• Remove passenger wheel + splash shield
• Remove accessory belts
• Support engine from below
• Remove passenger motor mount
• Remove crank pulley (impact gun helps big time)
• Remove timing covers

Now you can see everything.

---

2. Set Engine to Top Dead Center (Critical)​

Rotate crank clockwise until:

• Cam gear timing marks line up horizontally
• Cam dowel pins are straight up
• Crank timing mark lines up with case mark
• Oil pump sprocket timing mark aligned

If these aren’t aligned before removal, you’re gambling.

---

3. Remove Old Belt System​

• Compress hydraulic tensioner slowly in a vise (very slowly)
• Insert retaining pin
• Remove tensioner
• Remove timing belt
• Remove balance shaft belt
• Remove pulleys + water pump

If oil has leaked on the belt, fix the leak now.

---

4. Install New Components (This Is Where People Screw Up)​

Replace:

• Water pump
• Idler pulley
• Tensioner pulley
• Hydraulic tensioner
• Balance shaft belt

Install balance shaft belt first.
Then install timing belt in this order:


Crank → oil pump → idler → intake cam → exhaust cam → tensioner pulley


Keep tension on the belt the entire time on the non-tensioned side.

---

5. Set Tension Properly​

Install hydraulic tensioner.
Pull retaining pin.


Correct tension spec:

• Tensioner rod protrusion should be ~3.8–4.5 mm
• Let it sit 15 minutes
• Rotate engine by hand 2 full turns
• Recheck timing marks

If timing marks are off even one tooth — redo it.


No ego here.

---

6. Reassemble Everything​

Reverse disassembly.
Torque crank bolt properly (this is important).

---

Common Mistakes on 1G GSX​

• Ignoring balance shaft timing
• Not checking oil pump sprocket phasing
• Not replacing tensioner
• Using cheap eBay belt kits
• Not rotating engine by hand before starting

---

What You Should Replace While You're There​

• Front crank seal
• Cam seals
• Thermostat
• Accessory belts

These engines are 35 years old. Don’t shortcut.

---

Time & Difficulty​

DIY first time:
6–10 hours if careful


Difficulty:
7.5 / 10


If you’ve rebuilt suspension and done real mechanical work, you can handle it.


If not — this isn’t the job to learn on.

When I asked it for its sources for that tutorial, it said it :

I didn’t pull that write-up from a single live source or manual. It was generated from my general mechanical knowledge of the 1990–1994 4G63 DSM platform, based on:

• Factory Service Manual (FSM) procedures for the 1G 4G63
• Common DSM timing belt walkthrough standards
• Known specs and procedures used by the DSM community for decades

If you want actual primary sources, here are the authoritative ones you’d use to verify every step:

---

Factory Service Manual (FSM)​

Gold standard.

• 1991 Mitsubishi Eclipse / Talon / Laser FSM
• Section: Engine → Timing Belt → Balance Shaft → Tension Adjustment

Available via:

• Factory PDF archives
• Mitsubishi service publications
• DSM technical forums

---

Community Technical References​

These are long-standing DSM resources:

• VFAQ timing belt guide (archived technical walkthrough)
• DSMtuners technical articles
• Mitsubishi 4G63 rebuild manuals
• Gates Racing 4G63 installation guides

---

Component Manufacturer Install Guides​

For torque specs and tension specs:

• Gates timing belt install manual (for 4G63)
• OEM Mitsubishi tensioner spec sheet
• Aisin water pump install documentation

---

If You Want Exact Citations​

Tell me what you specifically want verified:

• Tensioner rod protrusion spec?
• Torque specs?
• Order of belt routing?
• Balance shaft phasing method?

So maybe for common repairs and maintenance, it could be really useful for the DIYer, so long as you ask it to reference factory service manuals, and you trust that it actually does for your year/make/model.

 

[curt-s]

After my experiences with "AI" over the past couple weeks, I wouldn't trust it to tell me water was wet.

 

[1990TSIAWDTALON]

Super good example of AI giving the exact OPPOSITE "answer".
THIS IS WHY I DON'T TRUST AI EVER.
This came up in a push notification about this thread -
Post in thread 'I can not reach 4000 RPM' https://www.dsmtuners.com/threads/i-can-not-reach-4000-rpm.549303/post-153962404

 

[CrackedDSM]

Not gonna lie I clicked into this thread and saw the angry responses and immediately thought "shit, did past me piss the wrong guys off this time?"


Glad to see it wasn't me, this time.

 

[1990TSIAWDTALON]

Cory, we're never pissed at you! You are very entertaining,.whitty and smart!

 

[miliman13]

I know what you mean about those AI email summaries.

I had my insurance provide me with a new windshield at no cost and when I got one of those AI emails I almost had a attack.

The summary said you have a due bill for $$ xxx.xx
And that’s not at all what the email said , it was just informing me how much the company paid to the service provider on my behalf.

But the lill AI helper forced on us ,almost killed me

It’s only gonna get better , ima grab pop corn while the world burns !!!!

 

[1990TSIAWDTALON]