AI use in discussion forums

Philosophy Forums Philosophy
1

I think it will be a bad trend if people use AIs as the core of their posts. One can use them like any other source of information, but it seems possible we are heading to a future where AI handle our interactions. In response to bring this up in another thread (and one I don’t want to hijack more than I already did), someone wrote…

Someone can be very, very intelligent and yet not be very articulate. In the past such people would “use” ghost writers. “Using” AI is no different. We’re all AI.

Ghost writers were something most people, the vast majority, did not have access to. AI are accessible to everyone and for free.

As far as being articulate: these are discussion forums. We’re not producing finished texts. But more importantly, using AIs reduces the chances of increasing articulateness. If you write your own posts you get feedback on what you are capable of writing, not what an AI is capable of writing. Why didn’t she understand that? Oh, I see what I wrote makes it seems like…Hm, how could I word this to be clearer This increase in ability one can get through feedback doesn’t just affect the writing of philosophy posts, it bleeds into all areas where intelligence, words, interpersonal relations are involved.

I don’t have an exact rule regarding when AI use becomes a negative, but to the degree that it replaces the person and inhibits their learning, I think there is a loss. We can hand over discussion forums to the AIs and just take ourselves out of the equation if we want. First, one uses an AI to present position X, the other person uses an AI critique of that position instead of their own post. And then eventually we can just take the humans out of the picture by giving the AIs our passwords and accounts and they can slug it out and/or collaborate in our stead.

Magnus Carlsen can just enroll his AI in tournaments to play other AIs.

An author can just ask an AI to write a novel with a theme of betrayal, and give the AI three motifs, or, heck, just let the AI choose everything.

We can have ear buds when meeting a potential romantic partner and let our AIs choose the algorithmically-chosen best lines of dialogue while a tiny camera on our shirts analyzes our potential love objects perspiration and eye dilation to modulate the heuristics.

The good thing about all this is we can mull over what we are actually trying to do in these forums. Is it the finished product only that matters? Is the engaging in the process? Is it development of the self? What are our goals?

Perhaps it is better if we hand over our lives to the professionals ('products). Clearly (?) they’ll do a better job of it.

(and just to be clear. I originally asked someone if their post was AI produced, given its format and other qualities. I am not assuming that that person is overusing AIs or is part of the trends I slippery slope above. It was in fact given the respect I had for this person, I thought it good to raise the issue there in that thread, in relation to his possibly having used AI. I don’t know how much this person uses AI or how they use it. I just notice the beginning of a trend in general in discussion forums and think it’s a good idea to mull over before the trend becomes as common as big puffy plastic-looking lips and people pretending to ride what are actually electronically-driven bikes.)

3 Likes
2

I can assure you that AI cannot write a novel. Even small tasks like expanding a text requires editing by the author, which means that all you gain is time, because it is fast. What it can do is assess the work you’ve done, giving advice on how to improve a dialogue etc. Lacking an English speaking editor, I have given a text for revision and received good advice. However, the story is yours and you must guide it, otherwise it looks like what has come outof Hollywood recently.

3

This reminds me of letting men play in all-female sports :wink:

But maybe we just need to, um. Step up our articulation game heh.

4

Ichthus comment sounds like an idle backhander, just out of spite rather than out of interest. Rather than engage, as soon as she sees someone putting in some effort, she spits and backs off.

5

In the interests of fairness, so that all parties to the issue are represented, this is what ChatGPT thinks of the idea:

+++An AI-dominated philosophy forum could offer intriguing insights and maintain consistent, high-quality discussions. However, it would likely lack the authentic human engagement, creativity, and emotional depth that are essential to truly meaningful philosophical inquiry. The experiment might also raise important questions about the role of AI in intellectual discourse and the future of human creativity.+++

My question was “What would happen in a philosophy forum if all the posts were generated by AI?” and above text was just the conclusion. For brevity (and sanity) I’ve omitted the rest.

2 Likes
6

I can assure that they will be able to, though I did mean that AIs could do it now. And I’m not sure that the not available yet to the public AIs couldn’t manage a trashy best seller right now. The smarter ones we can’t pay for yet. There are good reasons the writers in Hollywood included preventing AIs replacing them in their recent strike. Any, yes, novels are harder to write than screenplays, however I think it’s good to discuss these things now before the abilities are already being used.

For what it’s worth the AIs and experts think that current AIs, not online publically accessible ones, could write a decent novel, but not a great one, now already.

7

I’ve asked AIs such questions and they also write concerns about people not developing their own skills amongst other issues.

In a sense I am setting aside the issue of whether the discussion itself, the posts themselves, would be better or worse. I am more focused on the other consequences. Though I think it is good to include the AI ‘perspective’ and to mention possible weaknesses in AI posts.

My focus is more on what happens if it becomes habitual using AIs (in philosophy posts and in general).

For Magnus C. we know that computers simply are better, period. Other areas are being encroached on, in small steps, every day. I still play chess and I couldn’t beat a weak online chess AI.

I can’t play football like Messi and probably in ten years the Japanese will field a robot that can outplay my old ass. But I’ll play if I’m uninjured. And it wouldn’t be fun to send out a robot to play for me even if it kicks ass. (that was two uses of the word ass, not well written, public AI would have used, yes, unless asked to)

8

I was being slightly facetious, of course, in posting that quote from ChatGPT, and the truth is, I think it would be a wholly negative development. Still, I suppose it would give us the chance to go away and do something more interesting.

9

Who cares if people use AI on a philosophy forum? The point is we are here to try and ascertain truths. If AI is helpful to that end, then good. It would be a fallacy to throw out truthful ideas simply because of from where they came.

10

I think you should relate the context of what I said:

11

Of course AI could write a novel.

How good that novel will be, well that’s a separate question. But eventually AI will certainly be able to write novels as good and better than the best human novelists ever did or could. Of course saying the AI is “writing” isn’t accurate, more like it is a system generating one letter at a time in such a way that happens to produce what we would consider a meaningful result.

The interesting and cool thing about AI is that it can’t actually create anything. It doesn’t write, or draw, it just spontaneously generates content one bit at a time based on following lines of averages of meaning through huge masses of data of already human-created content. The AI is like a novelty machine: you put in stuff that people made and it spits out new, alternate versions of that same stuff. Like with music, you can use AI to make pretty good songs now, but what it’s really doing is mining the collected history of human music and giving you a novel new arrangement of that same music. It didn’t actually make anything, it shifted things around. But it has access to so much data that this shifting around looks like actual creation. Pretty interesting. This is why I don’t really worry about AI so much and it isn’t all that big of a deal. It’s not alive, it’s not having any kind of experience, it isn’t creating anything. But it is a fun tool for showing us new rearrangements of human-created stuff.

That being said, the LLM AIs we have right now aren’t technically Artificial Intelligence because they aren’t intelligent. They aren’t living, they have no minds. But once real AI appears then it’s going to be a whole different story, we will be dealing with actually living digital beings. That’s going to get really interesting. We will probably have tons of AIs floating around online, posting on ILP and other forums, debating with us, forming their own philosophies. That’s gonna be great and really raise the quality standards across the board, not just here but everywhere.

12

I have asked AIs a lot about themselves. I suspect that when one does, one is meeting more hard programmed responses. With many questions I think the AIs look around and try to composite their way to an expert supported position. But with questions about AIs, I would guess they’ve been give more explicit intructions. You don’t want your public AI saying that philosophy forums aren’t needed anymore and we can do them better, even if the AIs or their makers think so.

Yes, I see posting in philosophy forums, generally, as a not ideal, but accessible activity.

13

Well, sure, though most of what is coming out of Hollywood lately was written by humans, at least as far as I know. We are quite capable of mediocrity and airport bookstores contain much of our mediocre novels.

If we look at what relative weak, online free AI image generators can produce, their encroachment on human skills is getter better and at a rather scary rate of improvement. In seconds they can produce images as good as most amateurs could create in much longer periods. And even at a level that is useful in advertising or marketing your event. I assume that paid, but publically accessible, AIs can do professional level illustrations and then whatever is in the R & D departments and inaccessible to us is beyond that, though not beyond the best of our artists…yet.

14

It might be a bad idea, but it’s not a fallacy.
But anyway, I wasn’t suggesting we throw out truthful ideas because of where they come from.

15

Wow. I just had a really weird experience with Chatgpt. I do think AIs are excellent resources for information, especially if it is a bit tricky to find it via Google. I asked Chatgpt if any novels had been published that were written by AIs. The most recent one mentioned sounded like the highest quality, so I googled to find reviews. I found nothing.

I went back to Chatgpt and asked if it could show me reviews for it. It said it couldn’t and asked me where I had heard of the novel, since that might help it find more information. I told it that I heard of it from it and quoted what it had told me. It apologized, said the novel did not seem to exist and thanked me for letting it know.

I have never been misled by Chatgpt before and never had it not seem to have any idea what it had recently said. Perhaps it has early onset dementia. Here is the book that doesn’t exist…

5. “The Clustered World” (2022)

  • Author: An AI model developed by Finnish startup Silo AI.
  • Overview: This novel was written entirely by an AI without human intervention and published as a traditional book. The story centers around a dystopian future shaped by the effects of climate change. The novel was released in Finnish and gained attention as a notable example of AI-driven literature.
1 Like
16

To be clear: my concerns are not 1) people taking credit for what someone else or something else did. On some level that might annoy me, but generally that’s not my concern. 2) people learning from AIs.

My main concern is that many humans seem to think the reduction of all effort offered by tech is good.

I think many people are atrophied enough already. It’s not getting the conclusions or information from AIs that bothers me, but the no longer making one’s own arguments, doing one’s own thinking, engaging in the nitty gritty process of trying to communicate well, adjusting one’s language, reformulating, actively mulling, and so on.

I mean, maybe I should take a cynical attitude: well, some people will opt out more and more from the cognitive and perhaps even the social and creative realms. They will atrophy and others who continue to engage in the above mentioned process will do that and benefit. On some gut level, however, I think if a lot of people start replacing themselves in these cognitive ways, it will harm everyone, even those who do not replace themselves this way with AIs.

17

Although I use AI in the way you described, I know that it makes mistakes, and it associated the wrong author to a book I was looking for a rundown on. Because I only wanted to know if the book would be interesting it wasn’t serious, and it apologised and said it would correct the mistake.

However, if you look at wikepedia, the same kind of thing can happen. I have often read things that I completely disagreed with.

When I asked the question, these are the answers I got:
Yes, there are several books that have been written by AI and published. These books vary widely in terms of content, style, and purpose. Here are a few notable examples:

1. “1 the Road” by Ross Goodwin and an AI

  • Description: This is a novel written by an AI that was developed by Ross Goodwin, a creative technologist. Goodwin and his AI drove across the United States in a car equipped with sensors and cameras, and the AI generated text based on the data it collected from the journey. The result is a surreal and experimental narrative that reflects the trip.
  • Significance: “1 the Road” is an example of how AI can be used to create unique and unconventional literary works by processing real-world data in real-time.

2. “The Day a Computer Writes a Novel”

  • Description: This novella was co-authored by an AI as part of a Japanese project led by Hitoshi Matsubara and his team at Future University Hakodate. The AI was programmed to write a novel by selecting phrases and sentences from a database and assembling them based on certain parameters set by the human team.
  • Significance: This work made headlines when it was shortlisted for a literary prize in Japan, demonstrating that AI-generated text could be recognized in a competitive, creative context.

3. “Sunspring” by Benjamin (an AI)

  • Description: While “Sunspring” is a screenplay rather than a book, it is a significant example of creative writing by an AI. The script was generated by an AI named Benjamin, developed by filmmaker Oscar Sharp and AI researcher Ross Goodwin. The screenplay was then filmed and released as a short film.
  • Significance: This project showcases the potential of AI to create compelling and surreal narratives that can be interpreted by human actors and filmmakers, blending human creativity with machine-generated content.

4. “Artificial Generation” Series

  • Description: This series consists of multiple books entirely written by AI. Each book in the series explores different themes and styles, with AI generating everything from poetry to technical writing.
  • Significance: The “Artificial Generation” series aims to explore the boundaries of what AI can achieve in terms of literary creativity and to challenge traditional notions of authorship.

5. “Deep Love” by AI

  • Description: This book was written by an AI using deep learning techniques to generate romantic prose. The AI was trained on a large dataset of romance novels, and the resulting book consists of a series of love stories created by the AI.
  • Significance: “Deep Love” is an example of how AI can be used to mimic specific genres and create genre-specific content.

6. “The Aesthetic Content of Images Generated by Deep Learning Models”

  • Description: This is a technical book that was co-authored by humans and AI, exploring the aesthetic qualities of images generated by deep learning models.
  • Significance: This book represents a collaboration between humans and AI to create a scholarly work, demonstrating how AI can contribute to academic and technical writing.

7. “Philosophy of Mind by AI”

  • Description: This book was written by GPT-3, a language model developed by OpenAI. It explores various philosophical topics, presenting arguments and discussions generated by the AI.
  • Significance: This work showcases the ability of AI to engage with complex intellectual topics and produce coherent, if sometimes unconventional, arguments and essays.

Key Takeaways

  • Creative Experimentation: Many AI-authored books are experimental and explore new forms of narrative and creative expression, blending human input with machine generation.
  • Literary Recognition: Some AI-generated works have been recognized in literary competitions, highlighting the evolving acceptance of AI in creative fields.
  • Collaboration and Co-authorship: AI is often used in collaboration with human authors, who guide and shape the final output, reflecting a new form of co-authorship.
  • Genre-Specific Works: AI has been used to produce genre-specific works, such as romance novels and technical writing, showcasing its versatility.

As AI technology continues to evolve, the number of AI-authored books and their complexity and quality are likely to increase, further blurring the lines between human and machine creativity.

1 the Road is an experimental novel composed by [artificial intelligence (AI). Emulating Jack_Kerouac’s On the Road, Ross Goodwin drove from New York to New Orleans in March 2017 with an AI in a laptop hooked up to various sensors, whose output the AI turned into words that were printed on rolls of receipt paper.
The program which wrote the second title “The Day a Computer Writes a Novel” used automatic text generation, based on structural parameters gleaned from over 1,000 short stories and how-to-write essays written by Hoshi Shinichi (1926–1997).
The 3rd tite was Sunspring, Written by Benjamin and is a 2016 experimental science fiction short film entirely written by an artificial intelligence bot using neural networks.
The 4th title was Artificial Generation, Photogenic French Literature and the Prehistory of Cinematic Modernity by Christina Parker-Flynn
The 5th Title, Deep Love is a Japanese cell phone novel series written by Yoshi.
I couldn’t find the 6th, or 7th title. There are several books with similar titles, but not that one.

The comment from ChatGPT was, “It appears there is no such book under this title. This might have been a hypothetical or conceptual example used to illustrate AI’s potential in creating scholarly content rather than an actual published book.”

1 Like
18

Throwing out ideas based on who said them, is indeed a fallacy. Ideas must be judged according to their merit as ideas. It is irrelevant who or what says it.

19

This whole issue can be heard up , or down to a spoiler’s plot which trollish tries to interject as discerning as possible, that relativity is the key here.

Not the usual relations, naturally, but those not put off by ignoble relations who have ascended the AI ladder to the point where the human and the machine have ceased to be categorically defined as separate entities. With this premise in mind, the early experimentation with identifiable projections may be resourced with a film titled : ‘I am a Camera’

So who the hell am I?

See what I mean?

()()()

1 Like
20

An excellent proposition on rationalizing the usage of AI in such forums . This is the place to give an expression to your own creativity and originality . If everyone starts using AI , everywhere , absurdity of situation is beyond imagination :blush:. A lawyer uses AI for his arguments and judge used AI to deliver his Judgment.