Dubrovnik why do you keep dumping wikipedia articles and chatgpt essays in here? Is that your way of discussing a topic
It's definitely not the Wikipaedia article. Why would you think it's authored by ChatGPT? In the absence of any attribution, I assume it is just his views on the topic of how "perfect information" should be interpreted.
Dubrovnik why do you keep dumping wikipedia articles and chatgpt essays in here? Is that your way of discussing a topic
Hi @ Octopus !
And Dub didn't troll in that last one.
And his post isn't a long series of copied nested quotes with a lot of interactive colored text within the quotes.
He didn't act like Opto and do namecalling there.
Gradually (but slowly) - Dub has been improving his posts.
Doesn't mean he'll be nice or not. That's unknown.
Maybe he'll 'revert'. No worries.
Idea: how appropriate quotes from Wiki/other sources and snippets of AI sessions are depends on quality versus length.
How to evaluate: Nobody can decide on that for somebody else (an idea) except when the moderators decide on what is actionable spamming/trolling or not. (a fact)
Best to assume if it's the unfiltered product of Superficial Intelligence it's probably not worth posting.
Wiki is OK so long as it's relevant, but people reading should be aware it may be BS.
@MARattigan
Your posts through the years are very good posts.
Just pointing out that some of them are coming out in very pale text these days and are harder to read.
They may not appear the same to you depending on how you have your colors and brightness set on your keyboard or Windows or screen or whatever.
Just mentioning.
As for long 'back and forths' with many nested quotes several people do those.
Yourself included.
And I don't do the namecalling. Opto does that rulebreaking. Not I.
Good people interfere with his posts while he tries to troll their posts.
--------------------
I tend to make sure the nesting of quotes doesn't increase too much or maintain in my postings. Most people do.
There's an idea of making one's own quotes using the quotation marks with shorter copied snippets - as opposed to truncating quotes.
Also - the method of using colored text 'interactively' within quotes often doesn't work well because it becomes too ambiguous which person is saying what.
Especially when its a long series of nested quotes.
Yes I talked back to you.
And the world continues.
-----------------------------
Dubrovnik why do you keep dumping wikipedia articles and chatgpt essays in here? Is that your way of discussing a topic
It's definitely not the Wikipaedia article. Why would you think it's authored by ChatGPT? In the absence of any attribution, I assume it is just his views on the topic of how "perfect information" should be interpreted.
In his past couple posts its like 2 wikipedia articles and the last one was AI, probably chatgpt. I can recognize AI written things and usually check with a detection tool to confirm. So thats why I think so.
Yes, there is the problem with nested quotes of determining who said what. I would have thought colour coding was the easiest way to distinguish. So long as different colours represent different times, there's usually only one sensible order. Can you suggest a practical way that is easier? As for problems with contrast I get that also with some posts, but it's no great problem - just select the text you want to read and it will probably come out in high contrast.
Dubrovnik why do you keep dumping wikipedia articles and chatgpt essays in here? Is that your way of discussing a topic
It's definitely not the Wikipaedia article. Why would you think it's authored by ChatGPT? In the absence of any attribution, I assume it is just his views on the topic of how "perfect information" should be interpreted.
In his past couple posts its like 2 wikipedia articles and the last one was AI, probably chatgpt. I can recognize AI written things and usually check with a detection tool to confirm. So thats why I think so.
Ah OK.
I don't use AI chat too much. I can usually recognise BS, but not all BS is produced by AI.
Yes, there is the problem with nested quotes of determining who said what. I would have thought colour coding was the easiest way to distinguish - can you suggest a practical way that is easier? As for problems with contrast I get that also wirh some posts, but it's no great problem - just select the text you want to read and it will probably come out in high contrast.
Hi Martin !
I mentioned about the pale text because its been happening a lot on your posts lately. It might look much better on your screen than in the forum.
Idea about the color coding - it works much better when there's only one exchange of quotes at a time.
With one extra color per poster.
The combination of the quotes increasing (or just maintaining) with more colors ...
plus 'necessary' rereading through the quotes every time to try to detect what's changed and then identify who said what and when ...
To each his own though.
-------------------------------
As for AI - it could make whoever 'unpopular' by posting AI-assisted text (as opposed to AI -generated).
A lot of things aren't perfect.
I asked AI when John Tromp published his famous number 5 x 10^44
It told me in the year 2000. I posted that
Dio kindly pointed out my post was wrong. That the number was published in 2021.
It then turned out that John Tromp posted upper bounds for numbers of possible chess positions in both 1998 and 2021.
Wiki would make less mistakes than AI I guess.
What colour is giving the problem? Have you tried my suggested bypass?
"Wiki would make less mistakes than AI I guess." - that would seem to be a reasonable guess. You're not forced to use either.
I would guess the same. But I would not bet on the same being true in a year.
Dubrovnik why do you keep dumping wikipedia articles and chatgpt essays in here? Is that your way of discussing a topic
It's definitely not the Wikipaedia article. Why would you think it's authored by ChatGPT? In the absence of any attribution, I assume it is just his views on the topic of how "perfect information" should be interpreted.
To me a recent long post looked like it could well be close to unedited output from an AI. I then ran a AI output detector on it.
Let me say I have no objection to well-prompted AI output. Generally speaking it is above average quality opinion/reasoning. But I think it always needs to be identified as such. Can you confirm the origin, @Dubrovnik?
I would guess the same. But I would not bet on the same being true in a year.
Sometimes one can spot that the AI's response is suspect.
Also its ability to answer accurately is affected by how the question is asked and what question.
Also if questions are asked ambiguously or instructions not phrased properly the AI shouldn't be blamed.
Sometimes it will qualify its answers on its own.
Often it volunteers information. But often its better to make it answer only the questions you asked and respond to all your points first. Insist that its not to make suggestions until you tell it to.
But on the other hand it often volunteers good information.
Sometimes it cannot correct its mistake. Especially formatting mistakes.
Chess doesn't need to be solved. It's a game played between two human opponents. Neither of those opponents come to the game to solve chess, they come to the game solely to pit their knowledge and experience of the game against the knowledge and experience of their opponent. Playing against a computer is nothing more than 'practice'
Chess doesn't need to be solved. It's a game played between two human opponents. Neither of those opponents come to the game to solve chess, they come to the game solely to pit their knowledge and experience of the game against the knowledge and experience of their opponent. Playing against a computer is nothing more than 'practice'
I agree chess doesn't need to be solved.
Needs and wants are in some contrast.
Regarding reasons and causes people do whatever there's a lot of variation.
I would guess the same. But I would not bet on the same being true in a year.
Sometimes one can spot that the AI's response is suspect.
True, but the sentence you quoted referred to the belief that AIs may be more reliable even than collaborative human sources like wikipedia within the next year. it's quite close already and the rate of progress is very high.Also its ability to answer accurately is affected by how the question is asked and what question.
Also if questions are asked ambiguously or instructions not phrased properly the AI shouldn't be blamed.
Sometimes it will qualify its answers on its own.
Often it volunteers information. But often its better to make it answer only the questions you asked and respond to all your points first. Insist that its not to make suggestions until you tell it to.
But on the other hand it often volunteers good information.
Sometimes it cannot correct its mistake. Especially formatting mistakes.
What colour is giving the problem? Have you tried my suggested bypass?
"Wiki would make less mistakes than AI I guess." - that would seem to be a reasonable guess. You're not forced to use either.
My posting wasn't a complaint. It was a comment.
Correct - we're not forced to use AI nor Wiki.
Nor post at all! 
AI is already improving.
And in a year from now it will be quite a bit better.
But many people will be quite a bit better at using it too.
And there'll be a lot more people using it.
How do you guys write comments 2 pages long lol?
Dubrovnik why do you keep dumping wikipedia articles and chatgpt essays in here? Is that your way of discussing a topic