Could anybody explain how this isn't easily circumvented by using a competitor's model?
Also, if everything in the future has some touch of AI inside, for example cameras using AI to slightly improve the perceived picture quality, then "made with AI" won't be a categorization that anybody lifts an eyebrow about.
> Large language models generate text one word (token) at a time. Each word is assigned a probability score, based on how likely it is to be generated next. So for a sentence like “My favourite tropical fruits are mango and…”, the word “bananas” would have a higher probability score than the word “airplanes”.
> SynthID adjusts these probability scores to generate a watermark. It's not noticeable to the human eye, and doesn’t affect the quality of the output.
I think they need to be clearer about the constraints involved here. If I ask What is the capital of France? Just the answer, no extra information.” then there’s no room to vary the probability without harming the quality of the output. So clearly there is a lower bound beyond which this becomes ineffective. And presumably the longer the text, the more resilient it is to alterations. So what are the constraints?
I also think that this is self-interest dressed up as altruism. There’s always going to be generative AI that doesn’t include watermarks, so a watermarking scheme cannot tell you if something is genuine. It is, however, useful for determining that something came from a specific provider, which could be valuable to Google in all sorts of ways.
I really hope SynthID becomes a widely adopted standard - at the very least, Google should implement it across its own products like NotebookLM.
The problem is becoming urgent: more and more so-called “podcasts” are entirely fake, generated by NotebookLM and pushed to every major platform purely to farm backlinks and run blackhat SEO campaigns.
Beyond SynthID or similar watermarking standards, we also need models trained specifically [0] to detect AI-generated audio. Otherwise, the damage compounds - people might waste 30 minutes listening to a meaningless AI-generated podcast, or worse, absorb and believe misleading or outright harmful information.
I feel this is not the scalable/right way to approach this. The right way would be for human creators to apply their own digital signatures to the original pieces they created (specialised chips on camera/in software to inject hidden pixel patterns that are verifiable). If a piece of work lacks such signature, it should be considered AI-generated by default.
1. One-sample detection is impossible. These detection methods work at the distributional level—more like a two-sample test in statistics—which means you need to collect a large amount of generated text from the same model to make the test significant. Detecting based on a short piece of generated text is theoretically impossible. For example, imagine two different Gaussian distributions: you can never be 100% certain whether a single sample comes from one Gaussian or the other, since both share the same support.
2. Adding watermarks may reduce the ability of an LLM, which is why I don’t think they will be widely adopted.
3. Consider this simple task: ask an LLM to repeat exactly what you said. Is the resulting text authored by you, or by the AI?
I wonder what exactly would prevent a developer from removing the signature from a generated file. One could remove arbitrary segments that signal that it is AI generated.
Would you really use Google products to write your email if you knew that they were watermarking it like this?
I think this technology is gonna quickly get eliminated from the marketplace, cause people aren’t willing to use AI for many common tasks that are watermarked this way. It’s ultimately gonna cause Google to lose share.
This technology has a basic use dilemma problem where widely publishing it’s ability and existence will cause your AI to stop being used in some applications
Hey I made an open source version of this last week (albeit for different purposes). Check it out at: https://github.com/sutt/innocuous
There's lot of room for contributions here, and I think "fingerprinting layer" is an under-valued part of the LLM stack, not being explored by enough entrants.
28 comments
[ 148 ms ] story [ 959 ms ] threadI'd love to see the data behind this claim, especially on the audio side.
Also, if everything in the future has some touch of AI inside, for example cameras using AI to slightly improve the perceived picture quality, then "made with AI" won't be a categorization that anybody lifts an eyebrow about.
It is easy to alter by just saving to a different format or basic cropping.
I would love to see how SynthID is fixing this issue.
https://help.openai.com/en/articles/8912793-c2pa-in-chatgpt-...
> SynthID adjusts these probability scores to generate a watermark. It's not noticeable to the human eye, and doesn’t affect the quality of the output.
I think they need to be clearer about the constraints involved here. If I ask What is the capital of France? Just the answer, no extra information.” then there’s no room to vary the probability without harming the quality of the output. So clearly there is a lower bound beyond which this becomes ineffective. And presumably the longer the text, the more resilient it is to alterations. So what are the constraints?
I also think that this is self-interest dressed up as altruism. There’s always going to be generative AI that doesn’t include watermarks, so a watermarking scheme cannot tell you if something is genuine. It is, however, useful for determining that something came from a specific provider, which could be valuable to Google in all sorts of ways.
Oh crap, knowing Google it probably means they will put articles generated using their AI higher among the search results.
The problem is becoming urgent: more and more so-called “podcasts” are entirely fake, generated by NotebookLM and pushed to every major platform purely to farm backlinks and run blackhat SEO campaigns.
Beyond SynthID or similar watermarking standards, we also need models trained specifically [0] to detect AI-generated audio. Otherwise, the damage compounds - people might waste 30 minutes listening to a meaningless AI-generated podcast, or worse, absorb and believe misleading or outright harmful information.
[0] 15,000+ ai generated fake podcasts https://www.kaggle.com/datasets/listennotes/ai-generated-fak...
2. Adding watermarks may reduce the ability of an LLM, which is why I don’t think they will be widely adopted.
3. Consider this simple task: ask an LLM to repeat exactly what you said. Is the resulting text authored by you, or by the AI?
And if I scan the image or take a picture of the image on display.
I think this technology is gonna quickly get eliminated from the marketplace, cause people aren’t willing to use AI for many common tasks that are watermarked this way. It’s ultimately gonna cause Google to lose share.
This technology has a basic use dilemma problem where widely publishing it’s ability and existence will cause your AI to stop being used in some applications
https://www.nature.com/articles/s41586-024-08025-4.pdf
https://www.nature.com/articles/s41586-024-08025-4
There's lot of room for contributions here, and I think "fingerprinting layer" is an under-valued part of the LLM stack, not being explored by enough entrants.