PinKushinSince this came up, I can explain how I interact with all LLMs, because I don't trust them. I run them on the scientific method as laid out by Bacon and refined by Popper. In this specific project, Valve is the authority, not me and not the AI, and my own guesses are hypotheses like anyone's. Every claim gets tested, more than once, and every ported rule cites the SDK or a binary address. A flaky test is a defect, not noise: code is deterministic, so a test that isn't deterministic means something underneath is wrong, and we chase it until it isn't. I have the AI attack my conclusions as hard as I attack its, and hooks enforce all of it so we can't forget or quietly weaken the rules.
You understand how posting obviously LLM-generated text in this thread like this is disrespectful and doesn't help your cause, right? I've never signed up for ChatGPT or Claude in my life, yet I've seen so much nauseating LLM output over the past 4 years simply by virtue of wanting to spend time online (my fault, I guess). If you can't be bothered to flesh out your prompts into coherent paragraphs using your own human mind, how can you expect others to spend their own time reading anything you post? When I see LLM-generated text, my default expectation is that either a) this is a bot reply/some sort of scam, b) this is a non-native speaker naïvely trying to translate their text into English without realizing that LLMs will often distort the original tone of what they were trying to express, or c) the person hasn't really thought deeply about the implications of what they're submitting, as otherwise they would have defended the idea with their own words. Only in the case of b) do I really have any sympathy. Even then, if we're talking about something "serious" like someone submitting a pull request to a project, I'd prefer it if the person posted their original text in their native language (maybe as a collapsed spoiler) alongside whatever their LLM gave them so that we have additional assurance that the submitter isn't yet another person wasting the time of a maintainer.
None of this is a comment on how useful your AI-assisted project may or may not be, how capable frontier models are at this point in time, or whether it's justifiable for people to use AI for writing code if the resulting code yields a product that can benefit others and likely would not otherwise exist. This is primarily lamenting the fact that throughout history, humans have continually come up with new and innovative ways to be rude, annoying, and generally antisocial to each other. We now live in an era where people have to reflexively distrust any text they read, any image they view, or any song they listen to until they've observed enough signals that indicate that what they're looking at is authentic and worth paying attention to. (Maybe this is less of a problem for people who were never paying that much attention in the first place to the things they consume.) Photoshop released over 30 years ago, but generative AI has massively increased the ease with which people can put slop in front of everyone's eyes (e.g., the awful restaurant menus that Gaz and springrolls mentioned) to the point where I feel like the level of trust erosion isn't really in the same realm. You can clearly see why many people view generative AI with justified apprehension and hostility, and it's worth reflecting on how you can better interact with others if you want them to hear you out. Step #1 in that process: don't post AI-generated text without a disclaimer.
GazAs for the vibecoding projects ban, there should be clear rules to it. I'd disagree to ban every project with AI code outright. I believe projects that serve a clear, useful purpose should stay, as they might've not existed in the first place without AI.
I agree with you that clear enforceable rules are important. In a perfect world, we'd have some sort of tagging system where people would tag their projects with an "ai-assisted" tag if they used AI, and those of us who would rather not see those projects could filter them out. Since that probably isn't in the cards, then at minimum, people posting projects should be upfront about the degree to which they used AI so that others can quickly decide if they want to keep reading or spend their time looking at something else. Maybe some of these posters will feel that this is an unfair demand if, in their view, AI is "inevitable" and already everywhere ("why do I need to mention my AI usage if everyone I know is using it or will be using it one day?"), but I really believe that it's a considerate thing to do in a community space and people should want to do so even without a hard rule in place.
wickedplayer494I was already giving some thought to how timing will work going forward, given that I have been eyeing the possibility of some greener pastures in Genshin and/or HSR[...]
As someone who used to play HSR and quit before Castorice but still looks at leaks on occasion, all I can say is godspeed.