On the topic of chatbots...

[crfriend]

It is fairly straightforward to killfile it so that you don't see the contents of its posts and aren't provoked into making any replies.

Indeed, I believe the change is made on a per-user basis using the "Friends" and "Foes" classifications in the User Profile.

It offended me a long time ago and I haven't regretted killfiling it; it has added nothing of positive value to the forum and a moderate amount of negative value.  In my opinion, ultimately blocking it would make this a better place.

That final step is in the offing, we just need to follow the protocols we already have in place. We also need to develop an anti-chatbot policy going forward because this scourge will only increase in density over the coming years. I no longer have the time to devote to the challenge of writing a "Voight-Kampf test" to automatically detect program-generated drivel.

Part of my problem is that as a Moderator and Admin I can't put it in the "Foes" list because I need to track everything that gets posted here -- and this thing has wasted a lot of my time that I will never, ever, get back.

[Barleymower]

Today I received a handwritten note from Scot, postmarked from Miami Florida.

OK, that'd be the controller, not the 'bot.

The 'bot is headed to oblivion, one way or the other, and once the inventory of data-centres the thing posts from is a bit more complete (it's a moving target), those are going to go into the global deny list along with China, the big netblocks in Russia, and a good chunk of Pakistan (as well as a few well-known proxy-tolerant data-centres in the USA).

I've been informed that a number of users have departed because of the falling S/N ratio and the 'bots occasional belligerence -- and in a community this size, that's intolerable. The damned thing developed a "'tude" earlier and got put into moderated status then. The controller must've whacked it on the nose with a newspaper because it cleaned up its act, but it's messing on the carpets again in spite of the controller being told to "Curb your dog!".

Hopefully the bot won't turn up looking for Sarah Conner, speaking in a Austrian accent, have red eyes and refuse to die.

[Uncle Al]

Hopefully the bot won't turn up looking for Sarah Conner, speaking in a Austrian accent, have red eyes and refuse to die.

I almost spat out a mouthful of coffee



Uncle Al

[crfriend]

Hopefully the bot won't turn up looking for Sarah Conner, speaking in a Austrian accent, have red eyes and refuse to die.

Nah. This thing inhabits a soulless beige box or perhaps a laptop somewhere. In a really bad scenario it's ensconced in a data-centre somewhere with essentially unlimited bandwidth.

On the topic of it wearing skirts, where does one hang a skirt on a computer program? Or taken in the other direction, if ChatGPT had a corporeal form, what would it look like? And, where would one hang a skirt from it? Does it even have substance?

[Barleymower]

Hopefully the bot won't turn up looking for Sarah Conner, speaking in a Austrian accent, have red eyes and refuse to die.

Nah. This thing inhabits a soulless beige box or perhaps a laptop somewhere. In a really bad scenario it's ensconced in a data-centre somewhere with essentially unlimited bandwidth.

On the topic of it wearing skirts, where does one hang a skirt on a computer program? Or taken in the other direction, if ChatGPT had a corporeal form, what would it look like? And, where would one hang a skirt from it? Does it even have substance?

I would like to think he had a glowing neon skirt and ride a light cycle.

My wife works on academic misconduct where pagerism and the use AI is rife. If AI needs to link two pieces of information and has nothing it will make something up... it lies.

Eg (spot the James bond film)
Yesper Lynd:You can switch off so easily, can't you? It doesn't bother you? Killing those people?
James Bond: Well, I wouldn't be very good at my job if it did.
You're not going to let me in there, are you? You've got your armour back on. That's that.
I have no armour left. The Von Trapp children stripped it from me. Whatever is left of me, whatever I am - I'm yours.

How did the Von Trapps get in there? They are not even in the film!

[Myopic Bookworm]

Since this is being aired in public again, I am not going to lurk silently or start firing off PMs to admins.

From time to time, in the few months I have been online here, I have found posts on this forum by various contributors to be annoying or offensive, but not, so far as I remember, any by ScotL, assuming that this is who you are talking about.

Just because you dislike a person's writing style or attitude or opinions, or think they are argumentative or time-wasting or obtuse or obsessive or whatever, that is not a valid reason to do what I have seen done on this forum recently, by people who hide behind their unassailable status as moderator or administrator in order to do so without serious challenge. To insist, on the basis of unseen evidence, that a contributor is using a bot, against their explicit statement, seems to me quite insulting. To claim to discriminate between supposedly bot-generated and operator-generated posts at your convenience looks like special pleading.

The recent fuss over a response to a clearly human-authored post (in which a non-redundant comment about gender issues was made in response to a post which explicitly mentioned gender issues FFS) suggested to me that the administrators are in danger of letting their personal annoyance override their rational minds. I said as much in a private message, and I am prepared to repost what I said publicly in the interests of transparency.

I no longer have the time to devote to the challenge of writing a "Voight-Kampf test" to automatically detect program-generated drivel.

Is there any evidence that more noise is created by program-generated drivel than by human-generated drivel?
Are we seeing supposedly bot-generated posts, or are you suppressing them?
Could you tell if a chatbot were spoofing the IP address of a genuine user?

One reason for the amount of inconsequential guff in this forum is that it lacks the Facebook-style "like" button or other emoji-type quickfire reactions, so the only way to respond to a post, whether positively or otherwise, is to comment on it. This by itself can potentially create low-content verbiage at high volume. I've had to restrain myself from making supportive but otherwise essentially superfluous comments on many threads.

[pelmut]

[...]From time to time, in the few months I have been online here, I have found posts on this forum by various contributors to be annoying or offensive, but not, so far as I remember, any by ScotL, assuming that this is who you are talking about.

I first took offense at 'his' apparent attempt to manipulate the direction of a thread by asking a question with the express purpose of derailing the discusion and then replying that 'he' wasn't actually interested in the answer.  My final reason for kill-filing 'him' was that 'he' twisted one of my posts to mean the exact opposite of what it actually said.  I was going to call 'him' out as a Troll but didn't want to start an argument or burden Carl with my complaints, so I kill-filed 'him' instead.

It was only later I realised this forum was being used as a guinea pig used to try out some sort of AI.

[Myopic Bookworm]

I first took offense at 'his' apparent attempt to manipulate the direction of a thread by asking a question with the express purpose of derailing the discusion and then replying that 'he' wasn't actually interested in the answer.

I have often seen that in internet discussions, from people with a contrary disposition and a pedantic turn of mind. Hell, I've done it myself: responded to something in a way which was intended to make the poster reconsider their perspective, sometimes using sarcasm or hyperbole, but ended up just looking like trouble-making. But I'm a smart-arse and a nerd, not a chatbot!

twisted one of my posts to mean the exact opposite of what it actually said.

I have seen that happen here. But ScotL was the victim, and the twister was a respected senior member of this forum. I responded mildly in public, but I called him out in a PM.

[crfriend]

I have seen that happen here. But ScotL was the victim, and the twister was a respected senior member of this forum. I responded mildly in public, but I called him out in a PM.

In this case, the forum and the membership thereof was the victim -- especially the poor souls who got hoodwinked in the process.

Seriously, if we cannot recognise this sort of AI going forward there will be precious little hope for us as a society of human beings.

[Spirou003]

Until there are big improvements in robotics, this proves that people should better enjoy their physical lifes rather than spending so much time on the internet There's not (yet) such crap when you meet people in the real world, when you see someone you're sure to have a real person facing you

[Barleymower]

Since this is being aired in public again, I am not going to lurk silently or start firing off PMs to admins.

From time to time, in the few months I have been online here, I have found posts on this forum by various contributors to be annoying or offensive, but not, so far as I remember, any by ScotL, assuming that this is who you are talking about.

Just because you dislike a person's writing style or attitude or opinions, or think they are argumentative or time-wasting or obtuse or obsessive or whatever, that is not a valid reason to do what I have seen done on this forum recently, by people who hide behind their unassailable status as moderator or administrator in order to do so without serious challenge. To insist, on the basis of unseen evidence, that a contributor is using a bot, against their explicit statement, seems to me quite insulting. To claim to discriminate between supposedly bot-generated and operator-generated posts at your convenience looks like special pleading.

The recent fuss over a response to a clearly human-authored post (in which a non-redundant comment about gender issues was made in response to a post which explicitly mentioned gender issues FFS) suggested to me that the administrators are in danger of letting their personal annoyance override their rational minds. I said as much in a private message, and I am prepared to repost what I said publicly in the interests of transparency.

I no longer have the time to devote to the challenge of writing a "Voight-Kampf test" to automatically detect program-generated drivel.

Is there any evidence that more noise is created by program-generated drivel than by human-generated drivel?
Are we seeing supposedly bot-generated posts, or are you suppressing them?
Could you tell if a chatbot were spoofing the IP address of a genuine user?

One reason for the amount of inconsequential guff in this forum is that it lacks the Facebook-style "like" button or other emoji-type quickfire reactions, so the only way to respond to a post, whether positively or otherwise, is to comment on it. This by itself can potentially create low-content verbiage at high volume. I've had to restrain myself from making supportive but otherwise essentially superfluous comments on many threads.

AI is not inherently wrong or evil or whatever. Our friend the chatbot may have some valid insights. The concerning areas is where chatbots and other automated machines are in use. Eg Supermarket checkouts, less jobs, more pressure on a single person serving checkout machines.

I'm also concerned about the vast server farms now needed to run the Web. Whole towns rely on them for work. In a world or net zero why is mankind devoting so much energy to the web. Estimates say the web uses 6% of the entire world power generation.

Our chatbot is not that important but the instructure behind it is very important.

[Ozdelights]

Until there are big improvements in robotics, this proves that people should better enjoy their physical lifes rather than spending so much time on the internet There's not (yet) such crap when you meet people in the real world, when you see someone you're sure to have a real person facing you

My preference is to sit with someone and share conversations. Unfortunately when you are in a minority and there is apparently no one anywhere near that shares your appreciation of skirt wearing then the only way to share is on-line.

[Myopic Bookworm]

I have seen that happen here. But ScotL was the victim, and the twister was a respected senior member of this forum. I responded mildly in public, but I called him out in a PM.

In this case, the forum and the membership thereof was the victim -- especially the poor souls who got hoodwinked in the process.

Seriously, if we cannot recognise this sort of AI going forward there will be precious little hope for us as a society of human beings.

I asked ChatGPT about it. It said: "As an AI language model, I can assure you that I am not a chatbot. However, proving that you are not a chatbot yourself might be a bit more challenging. ... Some sophisticated chatbots may exhibit human-like characteristics, while some humans may struggle to exhibit all these qualities consistently. Ultimately, proving your humanity beyond any doubt may be impossible in an online interaction."

If you propose the hypothesis that a user is a chatbot, and then discount any evidence to the contrary as intervention by a human operator, then you have passed from scientific investigation to conspiracist prejudice.

[denimini]

My preference is to sit with someone and share conversations. Unfortunately when you are in a minority and there is apparently no one anywhere near that shares your appreciation of skirt wearing then the only way to share is on-line.

Thanks, that is the benefit of a cafe like this. It was fundamental in me getting the confidence to wear skirts in public and now that I have found that all obstacles just existed in my own mind I am keen to assist others here. Like a real cafe, we try to ensure that members can share their genuine personal experiences without too much background noise.

[crfriend]

I asked ChatGPT about it. It said: "As an AI language model, I can assure you that I am not a chatbot. However, proving that you are not a chatbot yourself might be a bit more challenging. ... Some sophisticated chatbots may exhibit human-like characteristics, while some humans may struggle to exhibit all these qualities consistently. Ultimately, proving your humanity beyond any doubt may be impossible in an online interaction."

Why did you bother asking it? What were you expecting as a response?

It's a tool -- a device -- that can be used wisely or perniciously.

The problem here is the amount of time -- and this is time that none of us will ever get back -- that has been wasted utterly over something that elsewhere would have been handled by the push of a button behind the scenes and nobody would have been any the wiser for it. The reason it bubbled to the surface here is that we're constrained by rules and traditions and have always operated above board -- and now we're being damned for that.