3 ms·
Alright, here’s a take I haven’t seen in this thread yet: how could this be used for fingerprinting, beyond an existence check for the API itself?
by Klonoar 2y ago
Alright, here’s a take I haven’t seen in this thread yet: how could this be used for fingerprinting, beyond an existence check for the API itself?
- INTPenis 2y agoThis assumes the model is different on each computer. And that made me realize that Google might start training it with your browser history. Anything is possible at this point.
- TonyTrapp 2y agoEven if the model data is the same - we have seen that fingerprinting can be applied to WebGL, so if hardware acceleration is used to run those models, it might be possible to fingerprint the hardware based on the outputs?
- poikroequ 2y agoThese models make heavy use of RNG (random number generator), so it would be difficult to fingerprint based on the output tokens. It may be possible to use specially crafted prompts that yield predictable results. Otherwise, just timing how long it takes to generate tokens locally. There's already so many ways to fingerprint users which are far more reliable though.