dack

joined 1 year ago
[–] [email protected] 62 points 10 months ago (7 children)

This is why Google has been using their browser monopoly to push their "Web Integrity API". If that gets adopted, they can fully control the client side and prevent all ad blocking.

[–] [email protected] 8 points 10 months ago

No, it's significant because attackers can pump out way more emails while also making them customized to their targets and constantly changing to help avoid detectors.

[–] [email protected] 50 points 11 months ago

They almost certainly won't. Every so often they make a big show of these raids and then quietly drop it later. Check out some of Jim Browning's videos to see how the raids work out.

[–] [email protected] 6 points 11 months ago (1 children)

Greatly increasing taxes for the super wealthy and closing tax loopholes would be a good start.

[–] [email protected] 59 points 11 months ago

Honestly, I think his communication here is fine. He's probably going to offend some people at NIST, but it seems like he's already tried the cooperative route and is now willing to burn some bridges to bring things to light.

It reads like he's playing mathematics and not politics, which is exactly what you want from a cryptography researcher.

[–] [email protected] 1 points 11 months ago

What do you define as "source" for an AI model? Training code? Training data set?

[–] [email protected] 5 points 1 year ago

That's much easier said than done. For game developers that already have games based on unity released or in development, changing to another engine is an expensive and time consuming development effort.

[–] [email protected] 5 points 1 year ago (4 children)

Who ever said signal is anonymous? Secure, private, encrypted - yes. But definitely not anonymous.

[–] [email protected] 3 points 1 year ago

Syncthing is another good cloud-free option.

[–] [email protected] 7 points 1 year ago (1 children)

At a very high level, training is something like:

  • generate some output
  • give the output a score based on how much it looks like real human text
  • adjust the parameters slightly to improve the score
  • repeat

Step #2 is also exactly what an "AI detector" does. If someone is able to write code that reliably distinguishes between AI and human text, then AI developers would plug it in to that training step in order to improve their AI.

In other words, if some theoretical machine perfectly "knows" the difference between generated and human text, then the same machine can also be used to make text that is indistinguishable from human text.

view more: next ›