You may be thinking that the CS response was a little odd - this might explain why.

[Deleted User]

Octopus Energy is now using AI to respond to customer queries:

https://www.businessinsider.com/ai-work-automation-octopus-energy-greg-jackson-2023-5?r=US&IR=T

molerat

Likely no worse than the average off shore call centre replies though.

[Deleted User]

If I was part of a CS management team, I would be very concerned if AI-generated responses are getting better feedback than responses written by a member of my team. Low human feedback ratings suggest that the problem has not been properly understood and/or properly answered.

FWIW, I had annual chronic illness review today by phone. It could all have been done perfectly safely via an App with me filling in the answers without the need for a practice nurse on the end of the phone. We need to embrace AI technology where it makes sense to do so.

TheElectricCow

That would actually explain a recent interaction I had with Octopus where their reply struck me as strangely worded at the time and written as if they had no real understanding of the question I had actually asked.

I had just assumed I was dealing with a regular human, either with poor comprehension skills or a non-native English speaker, who had completely failed to provide any useful information to further the conversation. Seems like it didn’t really benefit me or Octopus in my case, but if it was a machine I’ll concede that they sure enough had me fooled. If it wasn’t a machine perhaps they should have tried that instead I might have got somewhere quicker that way.

badger09

I look forward to reading the response to my email earlier today, questioning a gas charge held pending checking since 3 May. 

Perhaps the AI will be able to  press the release button🤞🏻

Gerry1

More likely that the AI will think there's a discharge of gas and will operate your smart meter's Shut Off Valve...

badger09

Gerry1 said:

More likely that the AI will think there's a discharge of gas and will operate your smart meter's Shut Off Valve...

Wrong. 

Response, from human being received late last night saying system was confirming CV & now through, so statement will be issued shortly. 

Statement arrived an hour ago. 

I thanked human being, & received almost immediate response, probably from AI. Fine by me. 

PS Gas is still doing its job😘

Gerry1

Dear badger09,

I can confirm that your CV has now been approved.  You will start your new job when Gas stops next Monday.

Love All Power,

AI.DA

InvertedVee

[Deleted User] said:

If I was part of a CS management team, I would be very concerned if AI-generated responses are getting better feedback than responses written by a member of my team. Low human feedback ratings suggest that the problem has not been properly understood and/or properly answered.

FWIW, I had annual chronic illness review today by phone. It could all have been done perfectly safely via an App with me filling in the answers without the need for a practice nurse on the end of the phone. We need to embrace AI technology where it makes sense to do so.

I think Martin should give an AI-bot an account on the MSE Forum and see what happens. Or perhaps he already has!

Spoonie_Turtle

InvertedVee said:

[Deleted User] said:

If I was part of a CS management team, I would be very concerned if AI-generated responses are getting better feedback than responses written by a member of my team. Low human feedback ratings suggest that the problem has not been properly understood and/or properly answered.

FWIW, I had annual chronic illness review today by phone. It could all have been done perfectly safely via an App with me filling in the answers without the need for a practice nurse on the end of the phone. We need to embrace AI technology where it makes sense to do so.

I think Martin should give an AI-bot an account on the MSE Forum and see what happens. Or perhaps he already has!

We already get bots popping up trying to spam.  (They tend to fail due to the safeguard of being unable to post links until X number of posts, by which time most of them have been recognised for what they are and banned.)

Spoonie_Turtle

Has anyone had a response from the AI, and did it sign the e-mail with a name?

I have, I suspect, and it did use a name beginning with T - although I don't wish to share the full name just yet unless someone else confirms it, just in case it was a human, although I can only imagine they were operating with debilitating levels of brainfog.