Well, of course. The stigma surrounding AI use will only get worse with stuff like this. I'm not interested in having people single me out for using AI. So glad I switched away from Anthropic.
Claude users are mad that Anthropic's new watermarks will catch them using it
21–30 of 99 posts
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#22Then a whole industry will rise at tracking and removing those watermarks.
Already happening. Just have a quick search for llm watermark remover and discovered multiple repos already available.
It’s very easy, just change each third word into something different than what you primarily wrote.
It’s very simple, just change every third word for something different to what you originally wrote.
It’s very straightforward, just change each and every third word for something different instead of what you actually wrote.
And so on ad infinitum.
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#23Well, of course. The stigma surrounding AI use will only get worse with stuff like this. I'm not interested in having people single me out for using AI. So glad I switched away from Anthropic.
Did you write this with AI? jk
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#24Earlier quoted context omitted.
https://en.wikipedia.org/wiki/Silly_season
The actual original dunk from that page is great enough to be reproduced here again verbatim: > during the months of autumn [...] Work is left to feebler hands. ... In those months the great oracle becomes —what at other times it is not—simply silly. In spring and early summer, the Times is often violent, unfair, fallacious, inconsistent, intentionally unmeaning, even positively blundering, but it is very seldom mere…
[dead]
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#25Well, of course. The stigma surrounding AI use will only get worse with stuff like this. I'm not interested in having people single me out for using AI. So glad I switched away from Anthropic.
The stigma surrounding AI use will get worse if... People know AI is being used? Sorry how does maintaining heightened paranoia by omitting clear signals help exactly?
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#26So, it seems the watermark is less what it sounds like (a stamp) and more an "imperceptible statistical pattern woven into the choice of words and sentence structures." Does that mean AI responses will sound even more "AI?" Like, will it become even easier to detect on a read-through because of the word choices and patterning? I see the word "imperceptible" there, but what does this mean in this context? My non-tech…
> Like, will it become even easier to detect on a read-through because of the word choices and patterning?
It's turned on right now. Can you tell a difference? I can't.
How many ways could I write this paragraph and still convey the same idea? Way more than we're aware of. Hundreds? Thousands? Maybe a lot more? The number of semantically similar variants increases exponentially with each word.
I suspect anthropic could turn their fingerprinting up or down if they want. If it were turned way up, claude would use weird phrasing but it would take very little text to tell if something were AI generated. If they turned it down, it would seem imperceptible to humans, but you would need a large sample to determine (with high accuracy) that a passage was AI generated. There's probably a very large middle ground where humans can't tell, and where it doesn't take a large text sample to know (with high probability) that some text was AI generated.
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#27This is aimed at academia with all the cheating kids and youngsters. I really dont have a big problem with it but I doubt it will be successful.
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#28Well, of course. The stigma surrounding AI use will only get worse with stuff like this. I'm not interested in having people single me out for using AI. So glad I switched away from Anthropic.
Hiding your use of AI will only make things worse when people find out. Look at how badly the CEO of Saber Interactive is crashing out after they lied about firing a writer to replace them with AI.
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#29So we finally tackle the issue of ai-text pollution and probably found a way to clean up the internet (from now on), yet people start complaining that "their" output is marked as spam.
Well, the solution is quite easy: start thinking on your own again and write the lines yourself.
I welcome this watermarking. Finally it's an easy detectable signal that someone just generated some request/answer to waste my time by forcing me thinking for the other person too. Now that time is over.
(It never was hard to spot this type of texts but now there is evidence)
Re: Claude users are mad that Anthropic's new watermarks will catch them using it
#30Earlier quoted context omitted.
Is it possible for them to sound even more AI?
Don’t forget the survivor bias is at play, you don’t see the ones that are good at passing for human written
Sure, but if my experience (I know, I know) is anything to go by, it's increasingly difficult to get Claude to write in anything other than it's own house style.