Ask Lemmy
A Fediverse community for open-ended, thought provoking questions
Rules: (interactive)
1) Be nice and; have fun
Doxxing, trolling, sealioning, racism, and toxicity are not welcomed in AskLemmy. Remember what your mother said: if you can't say something nice, don't say anything at all. In addition, the site-wide Lemmy.world terms of service also apply here. Please familiarize yourself with them
2) All posts must end with a '?'
This is sort of like Jeopardy. Please phrase all post titles in the form of a proper question ending with ?
3) No spam
Please do not flood the community with nonsense. Actual suspected spammers will be banned on site. No astroturfing.
4) NSFW is okay, within reason
Just remember to tag posts with either a content warning or a [NSFW] tag. Overtly sexual posts are not allowed, please direct them to either !asklemmyafterdark@lemmy.world or !asklemmynsfw@lemmynsfw.com.
NSFW comments should be restricted to posts tagged [NSFW].
5) This is not a support community.
It is not a place for 'how do I?', type questions.
If you have any questions regarding the site itself or would like to report a community, please direct them to Lemmy.world Support or email info@lemmy.world. For other questions check our partnered communities list, or use the search function.
6) No US Politics.
Please don't post about current US Politics. If you need to do this, try !politicaldiscussion@lemmy.world or !askusa@discuss.online
Reminder: The terms of service apply here too.
Partnered Communities:
Logo design credit goes to: tubbadu
view the rest of the comments
No it can't do that. It's an LLM, it can only generate the next word in a sequence.
Also this doesn't solve OPs problem at all. If it's in the top 10 results on a major search engine then anyone can find it in minimal time.
Fucking AI bros being like I remade looking at 10 google links but this time it burns down a forest and tells me what a genius I am for asking.
As much as I understand your hate for LLMs, this is wrong.
Explain to me which forest burns down when I run an AI on my local computer that uses the same power (or less) as running a video game?
AI / LLMs aren't evil or unethical or immoral - commercializing them into enormous behemoths that eat resources 24x7 is.
Even local models are trained on stolen art and content. That's the immoral part.
There's many models that use open source training sets and weights. You can choose them.
No one seems to get this part.
AI models aren't trained on anything "stolen". When you steal something, the original owner doesn't have it anymore. That's not being pedantic, it's the truth.
Also, if you actually understand how AI training works, you wouldn't even use this sort of analogy in the first place. It's so wrong it's like describing a Flintstones car and saying that's how automobiles work.
Let's say you wrote a book and I used it as part of my AI model (LLM) training set. As my code processes your novel, token-by-token (not word-by-word!), it'll increase or decrease a floating point value by something like 0.001. That's it. That's all that's happening.
To a layman, that makes no sense whatever but it's the truth. How can a huge list of floating point values be used to generate semi-intelligent text? That's the actually really fucking complicated part.
Before you can even use a model you need to tokenize the prompt and then perform an inference step which then gets processed a zillion ways before that .safetensors file (which is the AI model) gets used at all.
When an AI model is outputting text, it's using a random number generator in conjunction with a word prediction algorithm that's based on the floating point values inside the model. It doesn't even "copy" anything. It's literally built upon the back of an RNG!
If an LLM successfully copies something via it's model that is just random chance. The more copies of something that went into its training, the higher the chance of it happening (and that's considered a bug, not a feature).
There's also a problem that can occur on the opposite end: When a single set of tokens gets associated with just one tiny bit of the training set. That's how you can get it to output the same thing relatively consistently when given the same prompt (associated with that set of tokens). This is also considered a bug and AI researchers are always trying to find ways to prevent this sort of thing from happening.
Your knowledge is out of date, friend. These days you can configure an LLM to run tools like
curl,nmap,ping, or even write then execute shell scripts and Python (though, in a sandbox for security).Some tools that help you manage the models are preconfigured to make it easy for them to search the web on your behalf. I wouldn't be surprised if there's a whole ecosystem of AI tools just for searching the web that will emerge soon.
What Mozilla is implementing in Firefox will likely start with cloud-based services but eventually it'll just be using local models, running on your PC. Then all those specialized AI search tools will become less popular as Firefox's built-in features end up being "good enough".
I understand how agents work. I'm just saying it cannot "verify relevancy". That's a qualitative assessment that an LLM is incapable of doing. The scripts and regex that form the backbone of the "agent" can absolutely download a webpage and add the text contents to the context. But after that it's just random bullshit.
It does way more than that.
I have it write scripts in 30 seconds that would take me 2 days to write and verify.
I can quickly parse through what it writes (takes me about a minute) to verify it hasn't done anything wonky, then test it in a VM I use for testing my own scripts.
It does this because the question I ask is very clear and explicit: exact script language/version, exact input, exact output, how the script should flow, what commenting should look like. It takes me about 1 minute to write a good question like this.
I've setup Projects in it with specific rules so I don't have to state those rules every time - I have one for each scripting language. I've saved that definition so when I update the rules I have a local definition for reuse or share with my peers.
For informational searches I have it provide source links automatically - I have a lot of general knowledge so whenever it produces something that doesn't sit right I will steelman the information. It's surprising what it can come up with this way.
My friends say "you like to argue with it" - well sometimes arguing is necessary.
Haha, downvoters. What a joke.