Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

How Claude marks AI-generated content

support.claude.com

Post reply on HN
Print view View original

81–90 of 446 posts

Re: How Claude marks AI-generated content

#81

Post by lhd1 »

simonw wrote: post #21

An interesting factor of this is competition. If Claude was the only model family they could ship a change like this and users who want to cheat (or don't like watermarks for other reasons) would just have to put up with it. In a world with many different competing models, the risk of losing customers to other providers over this is much more real. Maybe they've looked at the numbers and the portion of people who cle…

Scott Aaronson spoke about this in a colloquium where he said that this was mooted at OpenAI before the decision was made by Altman to not implement it for the reasons you describe.

https://youtu.be/9udWn1Hlj_s?si=VWOiK5-y4zcyDoHI

Re: How Claude marks AI-generated content

#84

Post by matheusmoreira »

dalemhurley wrote: post #25

People with dyslexia and dystrophia, commonly use LLMs to proofread content. Even Anthropic admits this is a limitation.

People with executive dysfunction too. LLMs bring execution costs down to near zero and are therefore assistive technology.

Re: How Claude marks AI-generated content

#85

Post by stranded22 »

dalemhurley wrote: post #25

People with dyslexia and dystrophia, commonly use LLMs to proofread content. Even Anthropic admits this is a limitation.

Yes, I’m audhd and dyslexic.

I am cancelling my Claude max 5x subscription and moving to ChatGPT pro. I have difficulty enough trying to ensure my meaning comes through correctly, along with everything else; to now have to look out for/analyse watermarks too?

I feel shamed enough by society, thanks Anthropic.

Re: How Claude marks AI-generated content

#86

Post by miohtama »

simonw wrote: post #15

> When a supported Claude model generates text, it weaves an imperceptible watermark directly into the text itself. You won’t see it, and it doesn’t change the meaning, quality, or readability of Claude’s response. I'd like to know a lot more about how that works. A lot of my interactions with Claude return pretty precise text. If I ask it to edit a project and refactor a specific function in several places I know ex…

Maybe there is a reason why Opus 5 produces such word salad conversations

Re: How Claude marks AI-generated content

#87

Post by stranded22 »

case540 wrote: post #7

I don’t like the idea of hacking a response to contain a watermark. I also don’t like the idea of false positives detections coming directly from Anthropic. If people read more AI generated content, people will probably start writing more in that style

The amount of times ‘delve’ appeared in general conversation in the last couple of years shows the influence LLMs have on society.

Re: How Claude marks AI-generated content

#88

Post by sixtyj »

simonw wrote: post #15

> When a supported Claude model generates text, it weaves an imperceptible watermark directly into the text itself. You won’t see it, and it doesn’t change the meaning, quality, or readability of Claude’s response. I'd like to know a lot more about how that works. A lot of my interactions with Claude return pretty precise text. If I ask it to edit a project and refactor a specific function in several places I know ex…

It was quick :) … https://claudewatermarkremover.app/

Re: How Claude marks AI-generated content

#89

Post by nprateem »

simonw wrote: post #15

> When a supported Claude model generates text, it weaves an imperceptible watermark directly into the text itself. You won’t see it, and it doesn’t change the meaning, quality, or readability of Claude’s response. I'd like to know a lot more about how that works. A lot of my interactions with Claude return pretty precise text. If I ask it to edit a project and refactor a specific function in several places I know ex…

Load-bearing==claude

Re: How Claude marks AI-generated content

#90

Post by dd8601fn »

baq wrote: post #29
simonw wrote: post #15

> When a supported Claude model generates text, it weaves an imperceptible watermark directly into the text itself. You won’t see it, and it doesn’t change the meaning, quality, or readability of Claude’s response. I'd like to know a lot more about how that works. A lot of my interactions with Claude return pretty precise text. If I ask it to edit a project and refactor a specific function in several places I know ex…

> I'd like to know a lot more about how that works. Count load-bearing words using two different algorithms in a belt-and-braces fashion

That’s the real shape of the problem.
Post reply on HN