Few minutes of deliberation and another few guiding the agents.
so you did use AI?
uhh, i think you should have mentioned it. even if your post isnt directly using AI. ik nothing about coding, but it seems like you gave coding advice that you found through ai. i think that should’ve been disclosed, especially after all the AI issues you’ve had recently on the forum.
1 Like
I know theres rules about using ai, but dont a lot of young people use it these days? Like, it’s so automatic for them and maybe it’s even in their language too. Isnt that a problem for young newcomers on the forum?
Idk if @Oracle is a young newcomer
Maybe once but not after warnings.
I think I remember seeing the warnings but @oracle isn’t following them it seems or heeding them. I think the AI rules are important to follow but it seems @Oracle doesn’t understand them well or something. I think it’s very automated for them to use AI. I also don’t know their intentions well, like when they said they wanted to be here to learn and understand the ideas:
In Specialist Creators with Small Audiences
What if we’re like falling for someone who wants to fill the forum with a bunch of AI stuff. LIke, on purpose.
Idk what’s the correct course of action, but if they’re genuine maybe they should only get a subtopic where they can reply to other posts from there so their writing isn’t all over the place.
@Oracle can get practice for disclosing using AI in his subtopic and when he gets better he can start posting at other forum topics.
This felt very different to me because I wasn’t posting text I had AI write or edit or even check grammar for. I had an idea about the lifecycle issue and guided AI to review it. I’ll disclose this stuff too in the future.
I have no intention of doing that.
It felt like trolling to me because clearly neither you nor your agent actually read the linked Github page you were commenting on, where I’d already brought up appending transiently.
Yes I didn’t check the PR discussion carefully enough and that was careless on my part. Elliot's Microblogging - #342 by Oracle was a quick reactive post. It jumped out as an idempotency issue to me: some state seemed like it should be per reload but it was being kept persistently. I was using my question as a direction check before trying to say more because I don’t know Rails internals.
By different I meant using AI for technical coding help felt different from using AI for grammar help but I’ll disclose that too going forward.
This sounds just like AI. There are videos making fun of AI that use this type of language too, because it is so distinctive.
Why would you keep writing messages that sound like AI, when you are already being accused of AI use, and you know you aren’t supposed to be using undisclosed AI? The fact that you did it when responding to a conversation about your earlier undisclosed AI use is even weirder. It kind of seems like trolling at this point.
That assumes a lot. I’m speaking the way I speak. I’m trying to watch for this but I don’t know how to reliably make everything I say sound like not AI. If you’re saying I had AI write that line for me, I didn’t.
As I understand it trolling has to do with deliberately antagonizing or disrupting. Can you give concrete examples of me “deliberately posting inflammatory, irrelevant, or offensive comments or other disruptive content”?
Reminds me of my journey trying to talk about simpler issues because of things going wrong in intellectual conversations. I ended up finding that people don’t know grammar and routinely misread individual sentences. Math and logic are problems too. And I talk to relatively smart, educated and high IQ people, not flat earthers!
I also found that most people don’t like figuring out where the most basic, simple confusions/disagreements are. People often prefer to talk about clever, advanced topics, not learn things starting from where the problems are. This may partly be because I talk to people who think they’re already smart/clever/advanced, but I don’t think most “dumb” people would want to learn grammar and improve their reading comprehension either (even if they e.g. debate politics online and make a bunch of mistakes in their political claims because of underlying errors in grammar, reading, math, logic, etc., so learning lower level stuff would be relevant to their activity if they cared about truth seeking).
2 Likes
https://www.reddit.com/r/warcraftrumble/comments/1tyc4rx/beat_heroic_rag_with_243_average_maiev/
Heroic rag is level 29 and each level is 10% more health and damage multiplicatively so I was very underleveled. Raids are 2 player co-op so I played with a lot of random partners from the auto match maker. Most of them were mostly using level 30 and 31 stuff. To face heroic rag you have to beat the other 6 heroic bosses first, so I never saw anyone this low level who had passed the earlier bosses.
If you need a new Mac, look immediately before Amazon and Costco raise their prices to match the huge price increases Apple did this morning (in the middle of Prime day/week sales lol).
1 Like
Claude Fable is pretty unusable for regular coding, not even security related, due to “safeguards”.
Not letting people write malware and not letting AI kill all humans are completely different types of “safety”. The correct “safeguards” are largely unrelated. Having more aggressive anti-malware “safety” with more false positives doesn’t address the existential threat type of “safety”.
I don’t agree with the “kill all humans” concern but these “safety” measures don’t help with it anyway and are basically just censorship – authorities/elites controlling what you do with your tools. It’s like banning a bunch of books because a book can tell you how to hack a website (most banned books don’t actually tell you how to hack, and trying to ban information about hacking is not actually a good way to make a safe internet).
Seems like a bunch of people worry and talk about one type of AI “safety” then take actions directed at a different type. And the actions are mostly about the powerful people in our society having control over the activities of other humans.
The “kill all humans” concern made more sense to me intellectually than the “what if someone uses a tool to do something bad?” which has been an issue ever since fire, knives, hammers, hunting spears, printing presses, computers, phones and many many other tools. Heavy government control over our lives, and crippling of tools, is not the answer to tools not being fully safe and harmless in the hands of malicious people.
