Pretty good but I still see it as achievable using the same approach as ChatGPT uses.
This thread will remain because it is a great topic but please do not post any material from ChatGPT in the main forum – this is simply because it always sounds great but actually it could be complete nonsense.
ChatGPT content posted in the main forum will be deleted, which is exactly what every forum mod should be doing, but most don’t because they like post volume.
Peter wrote:
but please do not post any material from ChatGPT in the main forum – this is simply because it always sounds great but actually it could be complete nonsense.
LOL now that’s dumb. What if you’re capable of determining if it’s correct and you verify that it is correct?
hazek wrote:
LOL now that’s dumb. What if you’re capable of determining if it’s correct and you verify that it is correct?
Sorry, hazek, posting AI crap in every thread is like posting porn everywhere. I was very annoyed, and was about to ask Peter to delete, but he reacted first. :)
This was literally the 2nd time I ever did that if not the first and it was actually relevant as it clarified another user’s post for someone who is less verse on that topic. So it actually added value.
But sure ok, it’s not my forum. I still think this is a dumb policy.
And I can assure you I’m not wrong
That’s a bold statement. Reminds me on my statements about AI 35 years ago when I was developing neural networks based software for voice recognition. I was pretty sure everything would be different in few years. Then I grew up.
I often question “how can something be intelligent if it is artificial”
Is an artificial Picasso, a Picasso or a forgery or a copy?
What does Chat whatsit say about the words “awful” or “wicked” or lots of other words which have different meanings now than they had a few decades ago.
If Peter had to give his details to one of the French telephone responders which ask “what is your postcode” even if he spoke fluent French, would it understand him?
Does being able to answer all the questions on Mastermind make you intelligent or just you have a good memory for facts?
In other words "can a computer really think or can it just copy what has been done before only quicker.?
Finally, do we really want to live in a world where computers do all our “thinking” for us?
I have been watching some things on You Tube where young people in the street are being asked the simplest of questions or calculations and the vast majority do not have the answer. You have to see them to believe it.
Ok, I’m now gonna write something demonstrating how old fashioned my thinking is: do we really need this AI BS?
PS
Please nobody replies to the above, thanks 🤣
Peter wrote:
People who start a post with “You guys”… followed by some pompous comment tend to get short thrift here
That’s a little harsh. 
I don’t think any offence was meant by “you guy”. It’s just casual speak.
He was being challenged and I thought he didn’t rise to it and deal with it politely! I appreciate different people see things differently, but I really don’t think any offence was intended. I always think that if no offence was intended, then we shouldn’t take any even if the words could be considered offensive in some way.
I’ve not used Chat GPT 4, just the free version 3.5. My first impression was “wow!” But when you start to use it a bit, you quickly realise that it’s just telling you what an average person would expect the answer to be. Often it’s completely made up.
I tried a number of times to get it to help me with something on an answer that I couldn’t find by googling. In those cases everything it came up with was false. When I pointed that out, it just apologised and then gave me another false answer. That’s the danger of Chat GPT. If you don’t understand the subject already, then you don’t realise that the answer may be rubbish.
What is does provide is a human type interface instead of a machine type response that you get from Google.
It might well have a use in some sort of customer service roles. Imagine for a moment a hotel. Many of the questions they get are probably pretty routine. What time is breakfast served? What is checkout time? Do you have parking? What’s the cost of parking? Could you get someone to look at the shower in my room etc etc. These answers could probably feed to something like ChatGPT and it could probably easily answer them, pretty reliably, without human involvement.
As for translation, translation engines were already getting very good before AI became a fashionable term. To me that’s just a program not AI 
