undefined | Better HN

0 pointsrngname222y ago0 comments

You need to think about different contexts.

Think about security or trust and safety or anti-scam or anti-fraud.

AI generated image, video, and audio can be used to circumvent a lot of systems used in these domains. Many of these domains are for protecting users from being scammed, being impersonated, being tracked, etc.

Think about criminal court. Evidence may become impermissible if it can't be proven whether an image or video or audio document is a forgery or captured reality.

It's a bit flippant and absurd to insult the intelligence of people working on AI detection. I'd be a bit dumbfounded by someone dismissing an effort w/o spending time to think about why that effort may exist.

0 comments

MetaWhirledPeas2y ago

> Think about security or trust and safety or anti-scam or anti-fraud.

It doesn't matter what context I think about it in. It isn't going to work! And it will make things worse for everyone involved.

Hypothetically let's say we get to a point where everyone believes the detection is 100% accurate. Well that's all that means: everyone believes it. Meanwhile AI has just gotten better, and we're all more fooled than we were before. All we are really accomplishing is enhancing the training necessary for AI to elude detection.

And there will be an inherent bias toward false positives, because high detection rate will be the selling point. The truth is secondary, and there's no way to verify the results.

rngname22OP2y ago

It does work. If you absolutely need to know if an image is AI generated, you can just have a central authority in the system watch the person draw the picture on a piece of paper. Or drive to your house and hand you the paper and pencil and watch you draw it in person.

There are workflows or system designs that absolutely can and will solve for human-verified creation, they just might be incredibly costly or unscalable compared to existing solutions. It's all just tradeoffs. Might make existing business models no longer work. Might open new ones.

MetaWhirledPeas2y ago

> If you absolutely need to know if an image is AI generated, you can just have a central authority in the system watch the person draw the picture on a piece of paper. Or drive to your house and hand you the paper and pencil and watch you draw it in person. There are workflows or system designs that absolutely can and will solve for human-verified creation, they just might be incredibly costly or unscalable compared to existing solutions. It's all just tradeoffs. Might make existing business models no longer work. Might open new ones.

This is pretty much my point. Like you said, incredibly costly and unscalable. A non-solution! We're better off not pretending we can compute what is and isn't real.

1 more reply

j / k navigate · click thread line to collapse

0 pointsrngname222y ago0 comments

You need to think about different contexts.

Think about security or trust and safety or anti-scam or anti-fraud.

Think about criminal court. Evidence may become impermissible if it can't be proven whether an image or video or audio document is a forgery or captured reality.

0 comments

MetaWhirledPeas2y ago

> Think about security or trust and safety or anti-scam or anti-fraud.

It doesn't matter what context I think about it in. It isn't going to work! And it will make things worse for everyone involved.

And there will be an inherent bias toward false positives, because high detection rate will be the selling point. The truth is secondary, and there's no way to verify the results.

rngname22OP2y ago

MetaWhirledPeas2y ago

This is pretty much my point. Like you said, incredibly costly and unscalable. A non-solution! We're better off not pretending we can compute what is and isn't real.

1 more reply

j / k navigate · click thread line to collapse