I think there might be value in requesting that people voluntarily disclose they are using AI and to what extent, because it makes it easier to help them if you know… it also means you can give them responses that they can explore with their LLM instead of taking more time from volunteers.
The best we can hope for is that folks feel comfortable sharing as much context about their problems as possible.
I’m with @dubois on this. I use LLMs at work, and don’t really have a choice.
For my own personal stuff I use a mix. Turning it off completely now feels like stepping out of a car onto a bicycle.
You’ve somewhat oversimplified my remark about wealth, brother; it was more of a metaphor…
But getting back to the point, you can’t stop it, no one can. Soon everything will be automated; when we live to see it, you’ll remember my words. Look at Airbus or Dreamliner pilots. Do they have any idea what processes are happening in the software when they move the control wheel? They have no clue. Similarly, I’ll give you another five, maybe ten years, and no one will explain to you how a colorful streak follows your cursor. Your approach will be like something from an open-air museum, something like I do now when I look back at VHS tapes with nostalogy. Don’t take this as an attack or anything like that; the whole point of my previous argument is to consider why people fear progress, and I’m not just talking about IT, but in general, from a broader perspective. As for your remark about the callousness of spam, do you ever have open documentation while working? I almost always do, and I know many people who do the same. Following your reasoning, this also makes our work soulless, doesn’t it? After all, it lacks the pure absoluteness of ideas; it’s tainted by cheat sheets. If, for example, Yuri Artiukh is livecoding, with one eye on the road ahead, where his whiteboard probably stands, and the other on his laptop, does that make his stream soulless spam? So what’s the difference between major support and minor support—support to support, static support and automatic support? I understand what you’re trying to say, and I’m absolutely not saying that people should be bots, mindless plants like bulrushes growing on the edge of a pond. But there are certain processes in the development of civilization that represent critical moments, like the invention of writing, the light bulb, or other such things. And that’s what this is: the automation of the human-machine relationship. I’m absolutely certain that the near future will create human hologram models. Just like in Spielberg’s brilliant work “A.I. Artificial Intelligence.” But the question is, why treat this as a bad thing? I have a feeling, decoding your words about love, that you feel a bit—I don’t know how to put it in English, but “abandoned,” though I’m not sure that’s the right word. Like something was taken from you, leaving you and your love of coding behind, while the world moves on and spits out pseudo-developers who are usurping your domain.
I think there’s a bit of a category mismatch in how we’re framing this and the conversation tends to drift away from the core of the proposal.
This isn’t about stopping AI use, or judging it, or pretending it won’t become more common. It clearly will, and it already is.
It’s closer to something like: in a hockey club, everyone is still playing the same game, but some players train with advanced simulation tools, some don’t, and some mix both. No one is saying the tools are “bad”, but if you’re trying to understand how a team performs, or match with players of similar training context, that information is useful.
That’s all the proposal is really aiming for: optional context, not restriction, definitely not stigma.
We’re not trying to solve the future of automation here, just make it easier to navigate and discuss projects in the present
I’m just wondering where the participants want to draw a line. Is this something like when photographers were against photoshop and processing in general?
I’ve been autocompleting my code for years now with copilot. Last three months I only did a few leet code problems manually, I’m reading specs and unit tests. Doubt I’d be able to post anything. Plus I picked up an old project with agents now.
And that’s what I’m talking about. Reading this, I thought to myself, “Oh my God, how small I am!” And if you wrote this, then you’re at a level I can’t reach; if this AI wrote it, then that’s a level I want to be a part of me. As for understanding the thread, I think everyone here understands it fully, even beyond. I have the impression that every time a thread like this starts, the discourse takes on a similar tone. A tone of motivations and hidden meanings, where everyone tries to slip in what they don’t want or are ashamed to say directly. People simply try to add a superstructure. We all know there’s something deeper going on. After all, whether there’s a tag or not is largely irrelevant, just like whether someone posts it or not. Notice that behind every thread like this there’s an emotional tone, not just a subtle one, but a dense one. Digressions are common, and straying from the main thread ends most discussions. People are engaging in philosophical and technological reflections on the ontology of digital existence, and all it took was the keyword AI. And that means something. It means nothing less than that in many cases, the context is personal. Perhaps we’re all ashamed of this, because we all want to be masters of our own ideas.
Could you at least take the effort to make paragraphs . I hear a lot of aggression and anger, obviously many vibe generator feel offended about this topic especially when they completely depend on it.
As red said this topic isn’t about a general ai debatte, but a solution to transparency, fairness, security and everything else that comes from this. And honestly it is really a offence if you want to be seen equally to people who pour their heart into their projects with real skills and spending years or decades on learning all of it. while not even understanding it.
It isn’t about “staying behind”, don’t worry most will get automated away anyway, there is no bright future for vibe corders either the value is in the ai not the user. Most places these days are either “human-only” or flooded with slop in the end, the tags only work to some degree but trust goes down anyway compared to very strict platforms.
I’m not sure if discourse offers adding custom fields, but having a ratio toggle to explicitly state human or ai (or via tags) would be the most fair solution.
Att his point we all use AI… is not about that, is about the idea that AI will replace a specialist in any field which I dont see happen, even now most of the code is written by agents but i always have to be in charge check things and make sure everythign is right so this is not somethign that can be replaced!
To keep us moving toward an actual solution instead of endlessly debating around the problem statement, I want to acknowledge a legitimate challenge raised by @dubois: where exactly do we draw the line? Fair question.
My recommendation would be to avoid rigid formulas entirely and keep this intentionally lightweight and based on good faith:
If a project is fundamentally your own work and understanding, even if AI was used occasionally as a tool or assistant, then no tag is really needed. If a project was generated heavily through prompting to the point where authorship and understanding become blurry, then using the tag probably makes sense. Not as punishment or stigma, just as context.
This can follow a crawl, walk, run approach. Start with an optional tag and soft guidelines, see how people actually use it in practice, then iterate from there if needed.
Thoughts? Does anyone have a more pragmatic solution?
. I hear a lot of aggression and anger, obviously many vibe generator feel offended about this topic especially when they completely depend on it.
, I will try remember 