Improving ASR as a member/moderator

[Thomas Savage]

Ohhhhh, Like a 1-800 Audiophile help line ? !

I'd rather be on the end of a inmate 1-800 GoodTime , than a 1-800 audiophile helpline.

One could argue id be better suited to the former too .. .

 

[Thomas Savage]

Everything here is real, Thomas, unless an AI or a bot responds. Even in earlier times, pen friendships could and in many cases did develop into wider friendships.
It is only the form of communication that is somewhat limited, but it can also be an opportunity for understanding, because perhaps you meet mentally on a level that would be very difficult in direct communication.

I could hardly share these thoughts in real time with a native English or American speaker, because my English skills are too limited for that.

Nevertheless, my thoughts are of course the thoughts of a real person, even if thoughts per se are fleeting.

Some acquaintances that began virtually in forums have grown into "real" friendships, including mine, which is hardly surprising as we share the same interests.

Im sensitive to the fact you feel that way , I also am genuine in the way I communicate online .

ASR has given me real world benefits also , so theres a value . Before I started posting on audio forums I could barely write in my native language! , Let alone communicate in a way that could meet the demands of being a moderator here for years , on my own too boot .

I think i encapsulate all , in what was in my view a amusing retor to rick,, ' yes id take the call , no i wouldn't post the bail ' .

Sums up online life ..,

Let's not conflate the trust and faith needed between the time periods , unknowns and extention of self that pen pals required with simply posting and getting a immediate response online . If we all needed to goto the trouble of actually writing, posting and waiting to express our thoughts here , they'd be no ASR . Alot of what's happens online relies on addiction related behavioural drivers .

 

[RayDunzl]

Ohhhhh, Like a 1-800 Audiophile help line ?

I called one of those 1-800 Psychic Hotlines late one Saturday Night while watching Professional Wrestling at Audio Buddy's house.

Whoever answered didn't know why I called, so I figured it was a scam and hung up.

 

[Chester]

Im sensitive to the fact you feel that way , I also am genuine in the way I communicate online .

ASR has given me real world benefits also , so theres a value . Before I started posting on audio forums I could barely write in my native language! , Let alone communicate in a way that could meet the demands of being a moderator here for years , on my own too boot .

I think i encapsulate all , in what was in my view a amusing retor to rick,, ' yes id take the call , no i wouldn't post the bail ' .

Sums up online life ..,

Let's not conflate the trust and faith needed between the time periods , unknowns and extention of self that pen pals required with simply posting and getting a immediate response online . If we all needed to goto the trouble of actually writing, posting and waiting to express our thoughts here , they'd be no ASR . Alot of what's happens online relies on addiction related behavioural drivers .

Is that your subtle way of saying you don’t want to be my maid of honour?

 

[Thomas Savage]

Is that your subtle way of saying you don’t want to be my maid of honour?

Hay in the cell block theres plenty of maids but very little honor, so just be sure one knows what ones ordering .

 

[Curvature]

Can we create a sticky about using AI?

For example, rules could include:

• Disclose which AI service was used.
• Say if you verified the information or not.
• No text dumps.

I found this particularly irritating: https://www.audiosciencereview.com/forum/index.php?threads/ancient-loudspeakers.64117/#post-2352748

 

[RickS]

Can we create a sticky about using AI?

For example, rules could include:

• Disclose which AI service was used.
• Say if you verified the information or not.
• No text dumps.

I found this particularly irritating: https://www.audiosciencereview.com/forum/index.php?threads/ancient-loudspeakers.64117/#post-2352748

Is really @amirm's call. While I understand some frustration, there is a range of nuance here and is not going away. With many members now using AI as part of translation apps, what disclosure would we have them make? Should anyone using translation be obligated to disclose?

As I allude in the OP, some of the best work is done when there is fellow member scrutiny. In any case, without some automation, moderators are left to relying on members reports or casual encounters anyway. Flagged Amir above and will get his feedback.

 

[respice finem]

Perhaps at least long and "multi-source" citations should be "marked" as AI generated when it is (or may) be the case.
I say "may", because one cannot be sure simply googling it, if it is not already falsified by some language model.
This is a growing problem IMHO, the more this spreads, the harder will the "cleanup" become (if possible at all).
Sorry for the pessimism.

 

[Mnyb]

There’s also permanent thread bans as I have found out/experienced for pointing out/reporting hateful/misogynistic comments regarding greta thunberg on the humour thread a couple of years ago, I merely pointed out that I found “insert names of 3 political figures here” much more damaging to democracy

Maybe one should simply report and do no more , I’ve also pointed out this misogynistic behavior and especially attacking young people that actually bother to form and hold an opinion consistently even if you don't agree with it .
I’ve politely pointed out in the humour thread that it was not ok . There was a whole swath of ”memes” who locked more like political propaganda.

 

[RickS]

Expect will see some AI-enhanced moderation eventually. Will we get AI to objectively police AI posts?

 

[Timcognito]

Expect will see some AI-enhanced moderation eventually. Will we get AI to objectively police AI posts?

Seriously, what are all those Robots doing on this site? I guess it's tribute to ASR that we get so many. Wonder if Stereophile and Absolute Sound get any? Please list the number with members and guests.

 

[respice finem]

Expect will see some AI-enhanced moderation eventually. Will we get AI to objectively police AI posts?

Even better, set two AIs against each other, will save some users the bickering
(but this would flood the server in minutes, probably)

 

[Thomas Savage]

Is really @amirm's call. While I understand some frustration, there is a range of nuance here and is not going away. With many members now using it as part of translation apps, what disclosure would we have them make. Should anyone using translation be obligated to disclose?

As I allude in the OP, some of the best work is done when there is fellow member scrutiny. In any case, without some automation, moderators are left to relying on members reports or casual encounters anyway. Flagged Amir above and will get his feedback.

Tricky, dont want to be putting walls up to potential members , contributing members at that .

Members are free to report posts , then its up to some poor guy to digest that nuance and decide whats ok .

There will be mistakes , im sure we can all be patient, understanding and maintain perspective. As you rightly say , this ain't going away , although improvement in AI might make the issue moot , AI might also make ASR moot along the way ..

 

[amirm]

Can we create a sticky about using AI?

For example, rules could include:

• Disclose which AI service was used.
• Say if you verified the information or not.
• No text dumps.

I found this particularly irritating: https://www.audiosciencereview.com/forum/index.php?threads/ancient-loudspeakers.64117/#post-2352748

We probably need to do something. For now though, I ran that post by multiple AI detectors and they all said it is written by a human.

 

[Curvature]

We probably need to do something. For now though, I ran that post by multiple AI detectors and they all said it is written by a human.

The poster came back and said he wrote it himself. There were half a dozen errors in it, though.

 

[tmtomh]

The poster came back and said he wrote it himself. There were half a dozen errors in it, though.

Okay, but that’s a different issue.

 

[respice finem]

We probably need to do something. For now though, I ran that post by multiple AI detectors and they all said it is written by a human.

Good. Unless the detectors are already taken over

 

[sigbergaudio]

Is really @amirm's call. While I understand some frustration, there is a range of nuance here and is not going away. With many members now using AI as part of translation apps, what disclosure would we have them make? Should anyone using translation be obligated to disclose?

Perhaps those who consistently use AI and/or other ways to translate can put it in their signature or something. Can be helpful for people interacting with them, as they can then be prepared for some mistakes in the text, and show good faith rather than start a debate because of poor translations.

 

[Somafunk]

A.I. formulated posts are an issue on another forum I am a member of but I placed those posters on ignore, I have zero desire to interact with such.

 

[Oddball]

I simply leave or don’t contribute the thread where it appears obvious or probable that AI is used repeatedly, even for translation reasons. You can still translate via Google Translate and it will not add the apparent AI behaviour.

Don’t have the time or energy to argue with the bot kind that will get you into spending more time for rebuttal then actually worth it.

Even fire dies if no oxygen to fuel it. So don’t find the AI so distressful unless they start invading the threads at large - which I did not notice thus far.