Using AI on Askaboutmoney

Brendan Burgess

Brendan Burgess

Founder
Messages
58,253
I have extracted these posts from an earlier discussion.

It's important to keep this under review.

Fortune said:
I'd be in favour of the phrase ChatGPT being added to the content filter so it can't be posted here
Click to expand...


Will we change the Posting Guidelines as follows:

AI generated answers to tax and other financial questions are usually incorrect. Please do not post them here.
This includes questions/answers such as:
  • Answering questions asked by others with an AI generated response
  • Starting a question such as "Chat GPT told me that there were no estate taxes in Ireland. Is this correct?"
  • Correcting another poster on the grounds that "Chat GPT told me that there are no estate taxes in Ireland."

Should it be banned as above?
Should it be completely banned from askaboutmoney?


Is there any role for AI on askaboutmoney? Does it answer questions like "How can I get rid of weeds on my gravel driveway?" better than the contributors to askaboutmoney?
 
Last edited:
One of the characteristics of a good advisor is that they will always be prepared to say "Sorry, I don't know enough to answer that question".

One of the characteristics of a bad advisor is that they will rarely if ever be prepared to say that.

AI programs can be great but as least as of yet, they're firmly in the latter camp.
 
ChatGPTs own 'opinion'...

Q: Is ChatGPT reliable for financial matters?

"ChatGPT can provide general information and insights related to financial matters, such as investment principles, budgeting tips, or explanations of financial concepts. However, it is not a substitute for professional financial advice. For personalized financial guidance, it's best to consult a certified financial advisor or expert who can consider your specific circumstances. Always verify information from multiple credible sources before making financial decisions."
 
ClubMan said:
Isn't it the case that most or all such publicly accessible AIs/LLMs are being trained on the complete visible (i.e. non deep/dark) web contents?
Click to expand...

Yeah, pretty much. LLMs are trained on massive datasets derived from publicly available online sources like websites, books, and code repositories. Which includes a huge amount of websites but not all websites. Reddit, the FT and some other large websites get paid for inclusion in the LLM datasets.
 
Nevertheless, if you're given a wrong or incomplete answer on AAM your chances of having it queried or corrected by another poster are a lot higher than if you got the same wrong/incomplete answer on ChatGPT.

So I think the results in the reliable advice steeplechase are:
  1. Advice from a qualified and competent professional
  2. Advice you get on Askaboutmoney
  3. Advice you get from technologies yet to be invented
Also ran: ChatGPT.
 
On specific questions, it is wrong often enough not to be reliable.

It should not be used to challenge the professional advisors like Liam Ferguson or Steven Barrett as in "Chat GPT says you are incorrect". My money would be on Liam or Steven every time.

But I have been impressed with torblednam's use of it. He seems to use it to summarise an issue and then he checks it to see that it is correct.

"Childless couples are being discriminated against when it comes to CAT"

So the first [€400k] of any gifts or bequests are exempt. The rest are taxed at 33%. What is the "first" €400k of bequests though - I can see how that alone could lead to a lot of legal wrangling, over whose inheritance is tax free and whose isn't. How, and by whom, is the tax actually...
www.askaboutmoney.com

But is this an appropriate use of AI?

Why Property Is a Fundamentally Broken Investment In Ireland

I would like to illustrate to a few local TDs what is wrong with the Private Rental Accommodation market, and why its only logical no one should provide Private Rental Accommodation in Ireland. In this post I try to highlight how a landlord with existing tenants essentially has a liability on...
www.askaboutmoney.com

I don't know the extent to which Stormy edited it?

And as it's an opinion piece, does it matter?
 
Brendan Burgess said:
But I have been impressed with torblednam's use of it. He seems to use it to summarise an issue and then he checks it to see that it is correct.
Click to expand...
This is similar to saying Wikipedia should never be used. Wikipedia can be extremely useful for funding the sources cited (for me, Wikipedia is trustworthy regardless). The same can be said for AI; it always gives sources. The big issue with AI is that it quite often misinterprets the sources it cites, or just completely hallucinates what they say
 
Brendan Burgess said:
I don't know the extent to which Stormy edited it?
Click to expand...
I did use AI, I genuinely thought it made it clearer. Maybe it comes down to peoples command of the English language (its never been a strong point of mine). However, I would say that I edited it heavily, it certainly wasn't a post that I simply prompted, copied and pasted into AAM. It took me a while to go through the numbers etc. I don't think there are any factual inaccuracies in it.
 
Brendan Burgess said:
Whatever we decide, people should definitely disclose that they have used AI as in "I used AI to summarise the arguments".
Click to expand...
I agree with this 100%. It seems like good etiquette for any use of AI in creating/drafting a post to be clearly divulged up front.
 
Folks

This is not a general shooting the breeze and cracking jokes about AI. Feel free to do this in other threads.

This is about its use on Asakboutmoney.
 
Brendan Burgess said:
And, as often happens on askaboutmoney, the answer is "You are asking the wrong question. You should be more concerned with..."
Click to expand...
I think this applies to the AI question posed here.
Brendan Burgess said:
It should not be used to challenge the professional advisors like Liam Ferguson or Steven Barrett as in "Chat GPT says you are incorrect". My money would be on Liam or Steven every time.
Click to expand...
What is the difference between 'ChatGPT says you are incorrect' and 'Cremeegg says you are incorrect'.

If a reliable source can be quoted, or a strong argument made (very under appreciated imo) then what matter where it comes from.

If a post should be banned it should be banned, irrespective of the use of AI.
 
cremeegg said:
What is the difference between 'ChatGPT says you are incorrect' and 'Cremeegg says you are incorrect'.
Click to expand...

If Liam or Steven answer a question on pensions, it's unlikely that you would say "You are wrong Liam".

You might say "I am not a qualified financial advisor, but I thought that there was no tax on pensions?"

Brendan
 
More and more lately I see content posted (including from established and respected posters) that strongly hints (e.g. the telltale "em dash" and other stylistic quirks) at having been AI generated or at least drafted but without any attribution. I find this very disappointing and bad etiquette to be honest. I have reported some of these suspected posts to the moderators but am loath to keep bothering them.
 
You must log in or register to reply here.
Back
Top