Well, here's hoping it can cope with Scunthorpe and someone who is in the habit of typing who're instead of who are.
Current, so called, ai isn't reliable enough for people to leave it to make decisions, it would be like giving a 6 year old a checklist and letting it get on with it
Exactly, see post #28 for an example of what can go wrong. https://www.diyaudio.com/community/threads/site-policy-on-ai-generated-posts.414207/post-7716882
Moderating posts in the way you describe is too demanding on server processing. Moderation should take place the instant a post is sent to the server. Before accepting it, AI can moderate it and accept it or send it back to the poster together with a describtion why it was not accepted. With such a system, there will be no need of temporary banning. The latter can be reserved for exceptional cases.
Please note, my posting is only describing a way how AI can be used. The owner and moderators are free to set the forum they way they deem fit.
Interesting point of view. I’d also be remiss if I didn’t acknowledge we’re all different.
One thing I fear some would miss out on is developing the patience for it. I could be understating what I see are the benefits of that.
until know I did not see any proof of real intelligence in the so called AI.
Even read that tests showed its politically influenced, unable to make citations from literature and its continuously lying if nailed several times to its obvious faults.
I am sure we only get to use a small fraction of its power.
Even read that tests showed its politically influenced, unable to make citations from literature and its continuously lying if nailed several times to its obvious faults.
I am sure we only get to use a small fraction of its power.
AI is a powerful tool. But still at the toddler stage.
With great power comes great responsibility and there are lots of players who will abiue that responsibility.
AI (better regarded as an Idioy Savant) can churn thru huge amounts of data and make connections/correlations. It is having a very positive effect in small slices of medicine, chemistry, physics, astronomy and nuch much more. ChatGTP shows us the range but is at this point not a whole lot better than a sideshow, the important stuff (right now) is happening in smaller niches.
dave
I already said what I had to say, so I shouldn’t add this, but…
The merits of AI can be discussed at length, but I don’t think this will provide practical implications for the management of the site.
From a concrete standpoint, automated posting should be prohibited and it is already most likely.
For the rest, telling people “don’t post assisted by AI or under the influence of AI, or don’t post text AI gave you” is akin to saying “don’t post while wearing red socks, or when drunk, or when you’re not confident your information is correct”.
If not today, it’ll soon be hard to tell if any of these things are happening. (I have seen things you people wouldn’t believe…)
The merits of AI can be discussed at length, but I don’t think this will provide practical implications for the management of the site.
From a concrete standpoint, automated posting should be prohibited and it is already most likely.
For the rest, telling people “don’t post assisted by AI or under the influence of AI, or don’t post text AI gave you” is akin to saying “don’t post while wearing red socks, or when drunk, or when you’re not confident your information is correct”.
If not today, it’ll soon be hard to tell if any of these things are happening. (I have seen things you people wouldn’t believe…)
I think most are on the same page here, so maybe I'm not adding anything new. This is mostly an English language forum, but from time to time, I think it would be great if someone could post in their native tongue and provide a forum button to translate. This would allow others who can help improve the translation if needed. This is much like you see on other sites that translates comments, it is not perfect but a good step. So if AI can improve the existing translation tools, this seems beneficial.
Like others, I don't want to see AI passed off as someone's contributions, nor would I want AI in the search results. Maybe it is acceptable to upload a PDF or image of the AI results for discussion if it's warranted assuming those are not searchable (i.e. block in robots.txt).
Additionally, a lot of search prompts have been replaced with AI prompts, I'm not sure if that is desirable either. I can see both a search and AI prompts coexisting though, and also have AI summarize a thread.
Like others, I don't want to see AI passed off as someone's contributions, nor would I want AI in the search results. Maybe it is acceptable to upload a PDF or image of the AI results for discussion if it's warranted assuming those are not searchable (i.e. block in robots.txt).
Additionally, a lot of search prompts have been replaced with AI prompts, I'm not sure if that is desirable either. I can see both a search and AI prompts coexisting though, and also have AI summarize a thread.
It was trivial to generate this example which shows the polished looks are effectively a layer of deception. Not that I think ai is trying to deceive, that’s a human thing.
When it comes to research, I’ve learned not to go far beyond my current understanding before consolidating the position. I would apply this even more with ai.
So it would be good to see humans appreciate how much you need to hold ai’s hand. Present company excluded I’m not confident I’ve seen that level of understanding from a majority of users. There are still blatant examples of pushing things too far, using ai to patch gaping holes in knowledge shared on various sites.
When it comes to research, I’ve learned not to go far beyond my current understanding before consolidating the position. I would apply this even more with ai.
So it would be good to see humans appreciate how much you need to hold ai’s hand. Present company excluded I’m not confident I’ve seen that level of understanding from a majority of users. There are still blatant examples of pushing things too far, using ai to patch gaping holes in knowledge shared on various sites.
Last edited:
This is a nice way to solve the problem of banning AIGC while allowing it for translations.
👍
@ElArte
I'd like to see people (us?) taking their chance to influence society if desired.
So for a political statement we can use our website to promote what we think is right.
We cannot control the world but we can make a statement.
Could be something like
"AI free zone"
Oh, I see someone did already a design on it.
I'd like to see people (us?) taking their chance to influence society if desired.
So for a political statement we can use our website to promote what we think is right.
We cannot control the world but we can make a statement.
Could be something like
"AI free zone"
Oh, I see someone did already a design on it.
I believe that AI is currently able to do what Moderators do in the Forum when a new member completes his registration process on the site.
I don't know how much it actually would cost to train a learning machine though.
I believe that it is (much) more difficult to "recognize" a post "written" by AI if not for its inherent lack of bad faith, that instead is a typically human "feature".
But, if it is difficult to recognize a text produced by AI, assuming it really is, then it will also be difficult to ban those tests produced by AI.
Regardless of whether that hypothetical post was analyzed by AI itself or by a human operator.
However, I just read something that changed my mind about AI.
https://builtin.com/articles/claude-ai
I had a kind of sudden awareness and I suddenly convinced myself that nothing and no one will be able to hinder the diffusion of AI and that my previous statements are only the fruit of a (pathetic?) idealistic attitude of mine.
Also because I've always refused to use ChatGPT and therefore I didn't even know its user interface.
Despite not being able, like anyone else, to distinguish between what is good and what is bad in preventive terms relating to the diffusion of a new technology.
I believe the perfect synthesis was that of Osvaldo de Banfield here:
because the spread of AI is at least comparable, like any new technology, to the spread of the smartphone, that is, impossible to stop or even slow down.
Of course, you can make a personal choice to give up, but in the end who will "win" is a given forecast.
Let's take Google, everyone knows that it collects everyone's personal data and yet we all (I guess) use it simply because it works better, is more efficient and doesn't force you to search twice.
Anyway, if I as an individual intercept a moral reason (assuming that a certain one exists) or simply an "intimate" reason to choose whether or not to use a smartphone or AI, it will not change the world of others.
But, perhaps, it may (or may not) change mine, that is, "the whole world".
I believe that the only reason why the diffusion of AI has not yet exploded is in people, but not for moral reasons, but for the ignorance of companies manager in choosing the methods and means and therefore how to insert the data and which data to insert to enhance the efficiency of any activity.
Also because there is still a lack of a large number of so-called specialized technicians, but there will soon be many more.
I obviously believe that the companies that will adapt first will be those that produce products or services with the highest profitability and, of course, diagnostic medicine is at the top.
Although not exactly for philanthropic reasons.
I don't know how much it actually would cost to train a learning machine though.
I believe that it is (much) more difficult to "recognize" a post "written" by AI if not for its inherent lack of bad faith, that instead is a typically human "feature".
But, if it is difficult to recognize a text produced by AI, assuming it really is, then it will also be difficult to ban those tests produced by AI.
Regardless of whether that hypothetical post was analyzed by AI itself or by a human operator.
However, I just read something that changed my mind about AI.
https://builtin.com/articles/claude-ai
I had a kind of sudden awareness and I suddenly convinced myself that nothing and no one will be able to hinder the diffusion of AI and that my previous statements are only the fruit of a (pathetic?) idealistic attitude of mine.
Also because I've always refused to use ChatGPT and therefore I didn't even know its user interface.
Despite not being able, like anyone else, to distinguish between what is good and what is bad in preventive terms relating to the diffusion of a new technology.
I believe the perfect synthesis was that of Osvaldo de Banfield here:
because the spread of AI is at least comparable, like any new technology, to the spread of the smartphone, that is, impossible to stop or even slow down.
Of course, you can make a personal choice to give up, but in the end who will "win" is a given forecast.
Let's take Google, everyone knows that it collects everyone's personal data and yet we all (I guess) use it simply because it works better, is more efficient and doesn't force you to search twice.
Anyway, if I as an individual intercept a moral reason (assuming that a certain one exists) or simply an "intimate" reason to choose whether or not to use a smartphone or AI, it will not change the world of others.
But, perhaps, it may (or may not) change mine, that is, "the whole world".
I believe that the only reason why the diffusion of AI has not yet exploded is in people, but not for moral reasons, but for the ignorance of companies manager in choosing the methods and means and therefore how to insert the data and which data to insert to enhance the efficiency of any activity.
Also because there is still a lack of a large number of so-called specialized technicians, but there will soon be many more.
I obviously believe that the companies that will adapt first will be those that produce products or services with the highest profitability and, of course, diagnostic medicine is at the top.
Although not exactly for philanthropic reasons.
Some vision which will come true:
if its all about control it would be impossible to live without smartfone and people who do not use it will be hunted down or be caserned.
Later refugees from these modern dictatorship will be searched by satellite and drones and will be eliminated.
Everyone ignoring the predator qualities of humans is naive.
First time in history the cost of control is so cheap that it will be abundantly abused and executed to excess. And it will be impossible to escape from it.
Its all been thought through in literature and films and political philosophy.
And if someone asks me as someone who studied policy I would say in one sentence:
"The dystopic visions of the future have won"
https://en.m.wikipedia.org/wiki/Homo_homini_lupus
Sigmund Freud summed it nicely up:
"Men are .... creatures among whose instinctual endowments is to be reckoned a powerful share of aggressiveness."
if its all about control it would be impossible to live without smartfone and people who do not use it will be hunted down or be caserned.
Later refugees from these modern dictatorship will be searched by satellite and drones and will be eliminated.
Everyone ignoring the predator qualities of humans is naive.
First time in history the cost of control is so cheap that it will be abundantly abused and executed to excess. And it will be impossible to escape from it.
Its all been thought through in literature and films and political philosophy.
And if someone asks me as someone who studied policy I would say in one sentence:
"The dystopic visions of the future have won"
https://en.m.wikipedia.org/wiki/Homo_homini_lupus
Sigmund Freud summed it nicely up:
"Men are .... creatures among whose instinctual endowments is to be reckoned a powerful share of aggressiveness."
Like any tool / instrument, I guess it depends on how one uses it. I find Google search indispensable to provide information instantaneously, that would be difficult to otherwise obtain. I've developed my Google "fu" well enough that I can create a prompt that will yield what I need pretty effectively, most of the time.
Likewise with AI. It's a dumb as a post without a well "engineered" prompt. The skill with which an human engineers a prompt, directly determines the quality of the result. You still have to create, to get it to create better than a garbage output.
I suppose if they built an AI like a really good acoustic guitar, where what comes out depends entirely on the skill of the operator. Versus something like a Casio / Yamaha keyboard, with 999 preset "perfect" rhythms / sounds and "one key orchestra". I dont see anyone being particularly successful as a musician making use of that and I suspect AI generated music would help someone "make it big" about as much.
Unless of course that too can be made all a function of how well the initial prompt has been "engineered", or created by the human using AI music generation simply as a tool.
It is to wonder if AI could be made with "idiot-proof" at its input. That is, if it detects a zero-effort prompt ("entertain me") it just kicks you out at the outset, with a message "Put a little effort into your prompts, else I wont waste my resources on the level being presented". I got that idea from a reddit community rule, where they talked about "zero-effort posts" being removed by the mods.
Likewise with AI. It's a dumb as a post without a well "engineered" prompt. The skill with which an human engineers a prompt, directly determines the quality of the result. You still have to create, to get it to create better than a garbage output.
I suppose if they built an AI like a really good acoustic guitar, where what comes out depends entirely on the skill of the operator. Versus something like a Casio / Yamaha keyboard, with 999 preset "perfect" rhythms / sounds and "one key orchestra". I dont see anyone being particularly successful as a musician making use of that and I suspect AI generated music would help someone "make it big" about as much.
Unless of course that too can be made all a function of how well the initial prompt has been "engineered", or created by the human using AI music generation simply as a tool.
It is to wonder if AI could be made with "idiot-proof" at its input. That is, if it detects a zero-effort prompt ("entertain me") it just kicks you out at the outset, with a message "Put a little effort into your prompts, else I wont waste my resources on the level being presented". I got that idea from a reddit community rule, where they talked about "zero-effort posts" being removed by the mods.
One can always revise the previous decisions - flexibility it is called
The recent religious crusades in this thread are beyond my level of acceptance.