AI or incompetence?

[Phil H]

My daughter is doing battle with a car manufacturer over a warranty claim; her car, which currently has a manufacturer's warranty, is leaking coolant and the local main dealer has declared it 'fair wear and tear' without even seeing the car. The manufacturer agrees with that assessment as in an email they state;

"Regarding the main issue, please note that of course you can diagnose the car first to make sure if it can be covered under the warranty or not, but take into account that diagnostic charges cannot be covered.

From our end we need to inform you that such issue always comes due to wear and tear conditions." (sic)

So firstly, the brand in question confirms that they do not need any diagnosis to declare a fault as "wear and tear".
Secondly, the email says the warranty expired some time ago (it didn't).
Thirdly, they addressed the email to "Mr..." with a completely different surname.

With the current concerns around deployment of AI it might be that they are using the email equivalent of a chatbot. If that's not the case, then AI must be a better alternative to the halfwit who replied on behalf of the main UK customer services division.

The battle continues......

 

[CatmanV2]

If that's an AI (and speaking with some authority) it's a bloody awful one.

I'd go with incompetence. That doesn't even read like good English. What brand is it?

<edit: Actually I'd go with Google translate>

C

 

[zagatoes30]

If that's an AI (and speaking with some authority) it's a bloody awful one.

I'd go with incompetence. That doesn't even read like good English. What brand is it?

C

Exactly my thoughts, AI Chatbots have come along way in the last few years

 

[Oneball]

Poor English yes but not necessarily a halfwit and a chatbot isn’t going to get your name wrong.

What exactly is wrong with your daughter’s car and what is covered by the warranty?

For example;

Car owner rings up dealer “my car has a split radiator hose, I want to book it in to be repaired under warranty”

Dealer looks at warranty which says coolant hoses are excluded and says “that’s classed as wear and tear we don’t need to see the car to tell you it’s not covered by your warranty”

 

[CatmanV2]

Exactly my thoughts, AI Chatbots have come along way in the last few years

AI chatbots can be stunningly good. Which begs the question why so many businesses are still using utter shite

C

 

[drellis]

My bmw is out of warranty today, I had forgotten as the renewal was about £1000 which I thought a bit steep, anyway, they phoned me, got it down to £700, but..... I then asked is the electric motor covered, the young lad on the other end said no! I maybe think he got it wrong but no way I was taking out a warranty if it doesn't cover it.
The bmw battery is guaranteed for 8 years anyway

 

[Grantoonriusmo]

My daughter is doing battle with a car manufacturer over a warranty claim; her car, which currently has a manufacturer's warranty, is leaking coolant and the local main dealer has declared it 'fair wear and tear' without even seeing the car. The manufacturer agrees with that assessment as in an email they state;

"Regarding the main issue, please note that of course you can diagnose the car first to make sure if it can be covered under the warranty or not, but take into account that diagnostic charges cannot be covered.

From our end we need to inform you that such issue always comes due to wear and tear conditions." (sic)

So firstly, the brand in question confirms that they do not need any diagnosis to declare a fault as "wear and tear".
Secondly, the email says the warranty expired some time ago (it didn't).
Thirdly, they addressed the email to "Mr..." with a completely different surname.

With the current concerns around deployment of AI it might be that they are using the email equivalent of a chatbot. If that's not the case, then AI must be a better alternative to the halfwit who replied on behalf of the main UK customer services division.

The battle continues......

More than likely incompetence. The level of incompetence I have to deal with internally and externally on a daily basis makes me question how I haven't burried someone under the patio.

With the incorrect surname, "Mr" and warranty not expired, are you sure the email wasn't for a different customer?

 

[Grantoonriusmo]

AI chatbots can be stunningly good. Which begs the question why so many businesses are still using utter shite

C

Have you tried ChatGPT yet? If not, give it a go. It's mind blowing good.

 

[Phil H]

Poor English yes but not necessarily a halfwit and a chatbot isn’t going to get your name wrong.

What exactly is wrong with your daughter’s car and what is covered by the warranty?

For example;

Car owner rings up dealer “my car has a split radiator hose, I want to book it in to be repaired under warranty”

Dealer looks at warranty which says coolant hoses are excluded and says “that’s classed as wear and tear we don’t need to see the car to tell you it’s not covered by your warranty”

I understand that, but all she knows is there's a pool of coolant under the car and no clear source or obvious leak. Could be a hose, pump, rad, heater, t/stat assembly, reservoir, who knows, but the dealer says it needs diagnosis at £125/hr +VAT then claims (supported by the manufacturer) that 'fair wear and tear' is going to be the verdict anyway.

More than likely incompetence. The level of incompetence I have to deal with internally and externally on a daily basis makes me question how I haven't burried someone under the patio.

With the incorrect surname, "Mr" and warranty not expired, are you sure the email wasn't for a different customer?

If it was for a different customer, and it seems possible, that adds insult to injury!

Have you tried ChatGPT yet? If not, give it a go. It's mind blowing good.

I'm an IT luddite, and such incarnations are strange to me, but thank you - I will take your advice and give it a go.
In the meantime, please provide the location of said patio, as I may have more business for you

 

[Oneball]

If she’s not happy go to a different dealer.

 

[lozcb]

Well this is the modern generation you younger guys voted for , more computers saying no , and chat bots as you call them having no understanding or empathy which from a customer services mentality is bizzar in my book , Not so long ago , keeping customers happy was mandatory , " the customer was always right" was the adage , which needs and takes the personal touch and a human mind to understand the empathy required , but for quite some years now the MFI / British telecom win some lose some customer
ideology ,,,they will soon be back eventually seems to be winning through and exacerbated with AI ................i know which i prefer ..................so we should all really be voting with out feet

 

[CatmanV2]

Have you tried ChatGPT yet? If not, give it a go. It's mind blowing good.

Oh very much so. However, not necessarily accurate. I can't imagine it having too much trouble passing the Turing Test *unless* you were looking for it. Sometimes its responses can be somewhat formulaic.

What I am seeing is a rise in what I think are AI generated spam, Facebook messages and posts on forums such as this.

C

 

[Swedish Paul]

Goodness. This whole rubbish about AI. The technology has not changed for 40 years. Neural networks, genetic algorithms, I was programming this stuff in the 1990’s.
The thing that has changed is storage sizes and processing power. Gives the illusion. Self awareness ********. However, I can see why replacing a person that costs more than electricity and silicon seems a great idea to the powers that be deciding with an Excel spreadsheet. Which apparently hasn’t yet been replaced by AI. Go figure.

 

[DLax69]

Ah. When I saw reference to artificial intelligence I assumed we were talking about the US Congress. Sorry. As you were.

 

[skiman]

Goodness. This whole rubbish about AI. The technology has not changed for 40 years. Neural networks, genetic algorithms, I was programming this stuff in the 1990’s.
The thing that has changed is storage sizes and processing power. Gives the illusion. Self awareness ********. However, I can see why replacing a person that costs more than electricity and silicon seems a great idea to the powers that be deciding with an Excel spreadsheet. Which apparently hasn’t yet been replaced by AI. Go figure.

Before "Artificial Intelligence" can be realised, there is the huge, dangerous and destructive obstacle posed by Artificial Stupidity. Something that those Excel-guided "strategic decision makers" seem unable to recognise.

 

[Gooner]

Goodness. This whole rubbish about AI. The technology has not changed for 40 years. Neural networks, genetic algorithms, I was programming this stuff in the 1990’s.
The thing that has changed is storage sizes and processing power. Gives the illusion. Self awareness ********. However, I can see why replacing a person that costs more than electricity and silicon seems a great idea to the powers that be deciding with an Excel spreadsheet. Which apparently hasn’t yet been replaced by AI. Go figure.

That’s exactly what an advanced AI secretly taking over the world would write

Hasn’t the 40 years of learning made a difference? E.g. using genetic algorithms to improve neural networks? I agree ChatGPT isn’t sentient, it’s kind of a word predictor and it’s not hard to make it make a mistake. But some medical diagnostic AI tools are repeatedly more accurate than even expert humans now so ChatGPT etc. will be soon (IMHO).

I’m asking because not many people have your experience, interested to know what you think.

 

[CatmanV2]

People are confusing AI and the singularity of consciousness.

C

 

[Harry]

I’m more worried about an Alien: Covenant scenario happening with this new AI tech. Say we do manage to invent space ships, find new habitable planets and then a cheeky android who thinks it knows better infects us with facehuggers.

I don’t think it will happen in the next 15 years, but perhaps in the next 30?

 

[Swedish Paul]

That’s exactly what an advanced AI secretly taking over the world would write

Hasn’t the 40 years of learning made a difference? E.g. using genetic algorithms to improve neural networks? I agree ChatGPT isn’t sentient, it’s kind of a word predictor and it’s not hard to make it make a mistake. But some medical diagnostic AI tools are repeatedly more accurate than even expert humans now so ChatGPT etc. will be soon (IMHO).

I’m asking because not many people have your experience, interested to know what you think.

Yes, the pattern recognition part of it is wonderful. To be able to quickly narrow things down in medical diagnostics is a great tool. It takes years to train a medical professional, and should always be used to improve analysis.
Natural language processing has long been well understood and in combination with speech recognition will put many a low paid worker out of a job on helplines and support scenarios. Already is.
Self awareness as in science fiction won’t happen with current technology so I wouldn’t worry about terminators but what is really concerning is that with a growing, often poorly educated population, the jobs will get less and less. I worry that science fiction of the near future with huge unemployment, pollution, global warming and extreme weather is speeding quickly towards us.

 

[MarkMas]

Have you tried ChatGPT yet? If not, give it a go. It's mind blowing good.

ChatGPT is a really good predictive text engine, that can create very plausible content.

The problem is that ChatGPT is not a knowledge engine, but a ******** engine. (Someone described it very well as 'a mansplaining engine'.)

So if you ask it a relatively commonplace and general question, like “How do you change a flat tyre?”, it will come up with ‘good ********’ based on a large number of predominantly sensible sources. For example, when I had a crow trapped in my fireplace, ChatGPT provided VERY sensible advice immediately, avoiding laborious searching of the blog posts that were, presumably, its sources.

The problems arise when asking it for more obscure and specific factual information; it will make up a plausible-sounding answer based on any quasi-relevant information it can find. This will include making up 'information' by assembling 'factoids' into 'facts'. For example, it will make up scientific papers and books (including assembling plausible-sounding titles and using well-known author names), and will also make up legislation and 'legal precedents', citing non-existent legislation and cases, and adding the names of actual legislators, courts and judges.

And as you know, 83% of all statistics are made up on the spur of the moment, but ChatGPT will make up 'statistical information' and 'data' in a very plausible way.

Finally, in contentious topics, or areas of popular conspiracy theories, where there is a lot of false information around, it will give too much weight to obvious nonsense.

People are beginning to realise that this is worse than useless.

It has been rewarded for being plausible and positive, with no consequences for just saying, very inaccurately, whatever seems vaguely credible at the time (like many politicians, CEOs and pundits).