Relevant xkcd: https://xkcd.com/2169/
Randal is some sort of savant…
Cool until you realise Grandma is senile and can’t actually think beyond piecing together text they’ve seen before into what they think is a coherent response.
Those keys will absolutely not work, either because they’ve already been used and were scraped from training data, or they are fake keys generated based on said training data.
I tried it myself and looked the keys up. It gives you generic keys. They will work and let you install Windows, but they aren’t “valid” keys so to speak. At the same time, they aren’t “fake” either
Fun fact: If you google those codes you find out that they are “real” codes, but they don’t actually activate Windows. I think they are something that are used as placeholders in the upgrade from Windows 8 to 10 or something, but don’t know the specifics.
ChatGPT actually can’t create new “words”, just regurgitate words that it’s seen somewhere before!
Yep yep, statistical analysis as to the frequency of tokens in the training text.
Brand new, never-before-seen Windows keys have a frequency of zero occurrences per billion words of training data.
That isn’t actually what’s important. It’s the frequency of the token, which could be as simple as single characters. The frequency of those is certainly not zero.
LLMs absolutely can make up new words, word combinations, or sentences.
That’s not to say chatgpt can actually give you good windows keys, but it isn’t a fundamental limitation of LLMs.
Okay, I’ll take your word for it.
I’ve never ever, in many hours of playing with ChatGPT as a toy, had it make up a word. Hallucinate wildly, yes, but not stogulate a word out of nothing.
I’d love to know more, though. How does it combine new words? Do you have any examples of words ChatGPT has made up? This is fascinating to me, as it means the model is much less chained to the training data than I thought.
A lot of compound words are actually multiple tokens so there’s nothing stopping the LLM from generating the tokens in a new order thereby creating a new word.
It can create new words, I just verified this. First word it gave me: flumjangle. Google gives me 0 results. Maybe Google is missing something and it exists in some data out there, Idk.
I’m not sure what is so impressive about this though. Language models can string words together in unique ways, why would it be different for characters?
Sure it can create new words. It can’t create new tokens, would be more correct, I think. But a token is just a text fragment, and, as far as I know, they can range from being several words to being single characters.
I got it it say vilumplox. It doesn’t return any Google search results.
If it was Windows 95 it could generate them
https://www.youtube.com/watch?v=cwyH59nACzQ&t=306s
Close your eyes and let the gentle rhythm of keys lull you into a peaceful slumber.
Thanks grandma.
This comment reminded me that when my generation is old and I am grandma, I could be like, let grandma type on her mechanical keyboard to help you sleep. XD We’ll be the cool old people XD
I take it chatGPT does not give you windows 10 codes if you ask about it directly? You just have to trick it a little OG Grandma style?😅
tbh the keys are probably fake
Yeah like bruh in bullet point 2, the 19th character is a fucking multiplication sign
You should probably watch Enderman. He unlocked Windows 11 Pro using Windows 7 Ultimate keys generated by ChatGPT. It took 3 regenerations, but a key did pass the online check.
This is so nostalgic to me. I left windows many years ago and seeing those keys reminded me how I used to try and get those keys etc. :)
You left it for the better. Anything above Windows 7 is basically spyware now.
Yup I saw it coming in Win 7 and havent used windows since :)
My sweet grandmother reads Adobe cc leaked emails
