Thanks for sharing your thoughts. I completely agree that this forum should prioritize human-to-human interaction and knowledge sharing. If someone needs AI-generated answers, they can directly use OpenAI’s or Anthropic’s websites for that purpose. This forum thrives on the expertise and collaboration of its community members, and maintaining that focus is crucial. Let’s keep this space dedicated to genuine, human-driven discussions and solutions.
1 Like
Wont this damage the information in the forum in the long run? At one point google will just spit out poorly worded questions with chatgpt answers.
Why not just like… ignore the poorly worded question? Better yet ask, or help improve the question, so that future readers can benefit?
Stack overflow is pretty amazing in this regard, it filters our the highly regarded questions. I know this forum is more flexible, but just how much flexibility is too much?
This seems like an interesting topic, so I’ll throw in my two cents. Generally speaking, the problem is not whether the answers generated by AI are good or bad, but rather whether someone uses them as their own ideas, as Pavel already mentioned. The medium through which you develop yourself is secondary. The source of knowledge is everything, regardless of its form. It seems to me that the problem lies deeper. It is a bit of a division between new and old era programmers, where the latter do not necessarily want to accept the fact that with technological progress comes efficiency and, above all, the speed of knowledge acquisition. I remember learning JS from regular paper books with screenshots showing the output, and I have to accept the fact that today’s generation is much more efficient than mine. Is it better from an epistemological point of view? This question is pointless because it touches on other sources.
I can venture to say that everyone uses it now, and admitting it may be embarrassing for some, especially for old-style notepad-coding developers. But is this a reason for self-criticism? Of course not. A source is a source, and when using it, it is important to express it honestly, at least to yourself. I followed Yuri’s stream today, and he was using Copilot as usual. Did anyone in the chat feel bad for him? Of course not. When creating his great course, could Bruno Simon use any tools to achieve his intended goal? Of course he could! And Paul and his R3F, Mugen with his extensive knowledge, the authors of this topic, and other participants of this forum? It is important not to lose sight of how relatively small the group of specialists here is and how they acquire and manifest their knowledge. After all, who cares how information flows on this funny planet?
But back to the thread, for me, it is a valuable source, at least from the point of view of working with GLSL. Many times I caught myself losing connection with what I was doing and only changing values without analysis, expecting a result that I had not even thought about before. And when I went deep enough into this rabbit hole, I simply pasted the code into the AI to explain it to me. We can also ask ourselves: how can we be sure that AI will create the right explanation of this problem? But we may as well ask whether the author of the book really knew what he was writing about, whether the authors of Three.js understood WebGL “well,” or other similar dilemmas. We never know this because it depends on a series of interpretive conditions.
The conclusion is this: use what you can, don’t take what isn’t yours, and if you do take it, check if it works.
1 Like
Your perspective provides a thoughtful and nuanced take on the debate surrounding the use of AI-generated content in programming and other technical fields. Let’s break down and respond to the key points you’ve raised:
AI and Knowledge Acquisition
You rightly point out that the crux of the issue isn’t necessarily the quality of AI-generated answers but the integrity of claiming them as one’s own. This speaks to a deeper ethical concern about intellectual honesty and the responsible use of AI tools.
Efficiency and Technological Progress
You highlight the generational divide in accepting new technologies and their impact on efficiency. The rapid pace of technological advancement has indeed transformed how knowledge is acquired and applied. Today’s developers have access to a wealth of resources and tools that enable faster learning and problem-solving. This is a natural progression and should be embraced as a means to enhance productivity and innovation.
Use of AI Tools
You mention public figures like Yuri and Bruno Simon who openly use AI tools like Copilot. This openness helps normalize the use of AI in development and removes any stigma associated with it. The key takeaway here is transparency and the acknowledgment that these tools are aids, not replacements for genuine understanding and creativity.
Epistemological Considerations
The point about the epistemological perspective—whether the source of knowledge matters if the knowledge itself is sound—is thought-provoking. AI, like any other source, requires critical evaluation. Just as one would question and verify information from a book, AI-generated content should be scrutinized and validated.
Practical Application
Your personal experience with GLSL and using AI to understand complex code highlights a practical use case where AI serves as a valuable educational tool. This reinforces the idea that AI can be a powerful supplement to learning, especially when used to bridge gaps in understanding.
Conclusion
Your conclusion encapsulates a balanced approach: use AI tools where beneficial, maintain integrity by not misrepresenting AI-generated content as entirely your own, and always validate the information. This approach ensures that AI serves as a beneficial tool without compromising ethical standards or personal growth.
Response to Your Post
Your detailed reflection on the role of AI in modern programming is both insightful and balanced. You effectively address the ethical considerations while recognizing the practical benefits of using AI tools. It’s important to acknowledge the generational shift in learning methodologies and to embrace the efficiencies that new technologies bring. By being transparent about the use of AI and rigorously validating its outputs, we can leverage these tools to enhance our capabilities without compromising on the integrity or quality of our work.
I’m pretty sure @discobot could be extended to make checks and create tags… Much like reddit’s bot
Hi! To find out what I can do, say @discobot display help.
It seems like you might be referring to Discobot, a bot used in Discourse forums. Discobot helps users learn how to use the forum software through interactive tutorials.
Here’s how you can interact with Discobot:
- Start Tutorial:
- You can start a tutorial by sending a message to Discobot.
- For the basic tutorial, send
@discobot start tutorial. - For the advanced tutorial, send
@discobot start advanced tutorial.
- Help Command:
- You can ask Discobot for help by sending
@discobot display help.
This will prompt Discobot to guide you through various commands and features available on the forum.
If you have any other questions or need further assistance, feel free to ask!
@discobot display help
I currently know how to do the following things:
@discobot start {name-of-tutorial}
Starts an interactive tutorial just for you, in a personal message.
{name-of-tutorial}can be one of:tutorial, advanced tutorial.
@discobot roll 2d6
3, 6
@discobot quote
Fear of failure is one attitude that will keep you at the same point in your life. — Byron Pulsifer
@discobot fortune
You may rely on it
My 3 cents without reading this entire thread (it was already discussed before.)
Please just don’t post GPT / AI answers. There’s no point in that, as simple as that 
There’s no shortage of experienced people here, covering plenty of timezones, willing to answer the questions relatively quickly (usually <1h) and in a proper way. By posting AI content you’re making their work harder by adding gibberish noise and hiding unanswered questions.
(If you’d like to experience how AI-generated mess looks, see Quora AI answers / Twitter Topics / Google AI Suggestions - it’s noise that rarely brings value.)
5 Likes
If I want ChatGPT opinion I’m going there and ask for it. I’m doing this already for pretty much everything and it is very helpful most of the time. But some time it is not. Then I go to specialised forums (like this one) and ask for people opinion/feedback. I expect people to answer and getting valuable information that ChatGPT could nor provide. So no, AI generated responses here would be just garbage.
7 Likes
I think sometimes you don’t know enough about a problem to even know what to ask chatGPT.
For instance here:
I could have dug through my old source to find the GPU unmasking code, but it was easier to just GPT it.
However, I doubt OP even knew that this was something that is possible to do…
So I think it’s still legit to use GPT in answering some questions.
Blindly pasting a question into chatGPT and just pasting the output isn’t super cool tho, but it’s also kinda the same as posting any usual spam. Spam posts get deleted.
I think some of the fear of people using GPT is overblown, because what is the incentive?
Spam posts are actually dangerous and malicious.
But It’s not like GPT is harvesting emails or linking to dodgy websites.
People using it to synthesize answers are mostly just trying to help I think, however misguided.
At least you went through the motion of removing the:
What you just asked is indeed an interesting question. Lets see what the answer is!
- bla bla
- lorem ipsum
- foo bar
People using it to synthesize answers are mostly just trying to help I think, however misguided.
Are they? Why is this such a controversial issue then, for example, banned on SO etc. We do in the modern world set some boundaries, like you have get training to become a doctor in order to treat people. If anyone could just do this based on a google search or a chatgpt prompt… i dont think it would be good.
Can we all agree that a thousand users all copy pasting into and from chat gpt would be disruptive? Would you visit this forum if every answer was Certainly! What you asked is indeed interesting. Let me answer it for you! ?
So we agree that its ok to have this feature. We wont wait for discours or whatever to update their software. But in order to avoid the thousands upon thousands of bots doing this, what do we do? Limit it to one person? Then we wont have a shift covered when they sleep. Have a few people cover all the time zones? I guess.
Would it then be fair to ask people to add GPT_COPY_PASTER in the nickname so people know whats it about?
I’m bothered by just a couple iof people doing this in a regular basis, i dread to think what happens when this number grows exponentially 
1 Like
Yea… to be clear only the code was source from gpt. The words were my own. And ya… I find the boilerplate answers super annoying.
Also, i forgot.
You raise some very valid points about the utility and potential pitfalls of using AI like GPT for answering technical questions. Here are some reflections and considerations on the topics you brought up:
Using GPT for Technical Questions
- Knowledge Gaps and Hidden Solutions:
- Often, people don’t know what they don’t know, making it challenging to ask the right questions. This is where GPT can be particularly helpful, as it can provide insights and solutions that might not have been initially considered.
- For instance, GPU driver issues and their quirks are a niche area that might not be immediately obvious to everyone. GPT can surface this kind of nuanced information that might otherwise be missed.
- Quality and Reliability:
- While GPT can provide useful information, it’s crucial to verify the outputs, especially for highly technical or critical applications. GPT’s responses can sometimes lack context or miss subtle but important details.
- It’s a good practice to cross-reference GPT’s suggestions with trusted sources or personal experience.
- Learning and Adaptation:
- Using GPT can be a learning tool. It can introduce new concepts and solutions that users can then explore further, deepening their understanding of a topic.
- For developers, it can be a time-saver to get a quick overview or a starting point for more in-depth research.
The Issue of Misuse and Spam
- Spam and Malicious Use:
- While GPT is not inherently malicious, the concern arises when its outputs are used irresponsibly. For example, unverified or incorrect information can propagate if users blindly trust and share GPT-generated content without proper validation.
- The intent behind using GPT is often to help, but there must be a responsibility to ensure the accuracy and appropriateness of the shared information.
- Community Guidelines and Best Practices:
- Platforms and communities can develop guidelines to encourage responsible use of AI tools. This includes verifying information, citing sources, and understanding the limits of AI-generated content.
- Encouraging users to provide context and details in their queries can lead to more accurate and helpful responses from AI systems.
Conclusion
Using AI tools like GPT can enhance productivity and provide valuable insights, especially in areas where users might lack detailed knowledge. However, it is essential to use these tools responsibly, validate their outputs, and continuously learn from the suggestions they provide. This balanced approach can lead to better outcomes and a more informed user base.
Stackoverflow got overrun by AI bots because they created an incentive for it, by gamifying SO. So peoples SO rankings and stuff actually translate to monetary value and power.
The main incentive in this discourse is to help people, and occasionally flex, so I kinda doubt 1000s of GPT people are gonna descend on us because they’re not gaining any social currency from posting (unless it actually helps someone).
I dunno tho, time will tell I suppose…
fingers crossed
You’re posting gpt results to prove a point and it’s becoming counter productive in itself.
You’re posting gpt results to prove a point and it’s becoming counter productive in itself.
But being counter productive is the point that i am trying to prove! I dont know if becoming counter productive is then actually a good thing or a bad thing
Ouch, my brain hurts.
If I think emotionally, increased misuse of AI here will reduce my willingness to participate in the forum. If I think rationally … I’d do the same.
On a side note, I’d be curious what would AI respond to the job offers here.
2 Likes
The Internet is quickly filling with AI. I’d love to keep forums like this to humans so that the signal to noise ratio here is better than endless answers that don’t have any stake in actually helping.
At the minimum, there should be a way to mark or filter AI answers. I’d say it’s better to have a “quality” filter but newbies asking questions here wouldn’t be able to tell the difference between a bogus generated answer and one made by an actual expert.