Trantarius

joined 10 months ago
[–] Trantarius@lemmy.dbzer0.com 1 points 2 weeks ago

I've always wondered what to do in this kind of situation. Somebody ought to know about it, but who do you call? Science? Ghostbusters?

[–] Trantarius@lemmy.dbzer0.com 27 points 4 weeks ago (3 children)

If they actually wanted to protect children, the answer is simple: reverse the responsibilities. Require porn sites to include metadata indicating it isn't safe for minors. Require browsers to recognize that metadata, and filter out that content if parental controls are enabled. If parents are still too lazy to turn it on, make it default (like "safe search", but more effective). The fact none of them have even suggested this is proof they don't care about children or even porn, they just want to be seen as if they do.

[–] Trantarius@lemmy.dbzer0.com 6 points 1 month ago* (last edited 1 month ago) (2 children)

Not certain if it's the same one, but a while ago such a bill was on the docket until they literally forgot about it. Like, they all supported it, but no one brought it up so they didn't vote on it.

[–] Trantarius@lemmy.dbzer0.com 22 points 1 month ago (6 children)

Absolutely none of this law was ever about privacy or mental health. No one ever claimed it was. The law is banning tiktok because it is based in China. That is the reason given by the law itself. The possibility that meta or Google or some other American company will buy or replace tiktok and operate the same way is not an unintended outcome. It is literally the whole point of the law to get bytedance to sell tiktok to an American company.

[–] Trantarius@lemmy.dbzer0.com 1 points 2 months ago (1 children)

You have that backwards. Roko's basilisk would punish anyone who didn't help create it.

[–] Trantarius@lemmy.dbzer0.com 4 points 3 months ago

I'm pretty sure he said " the rules were that you were going to fact check, this isn't fact checking" or something to that effect. He was accusing the moderators of being argumentative.

[–] Trantarius@lemmy.dbzer0.com 12 points 4 months ago

It already was. The Ohio SC upheld almost all of the phrasing.

[–] Trantarius@lemmy.dbzer0.com 0 points 4 months ago

Do you have a source for this? This sounds like fine-tuning a model, which doesn't prevent data from the original training set from influencing the output. The method you described would only work if the AI is trained from scratch on only images of iron man and cowboy hats. And I don't think that's how any of these models work.

[–] Trantarius@lemmy.dbzer0.com 0 points 4 months ago (2 children)

Other than citing the entire training data set, how would this be possible?

[–] Trantarius@lemmy.dbzer0.com 5 points 5 months ago (2 children)

When does that even happen? If you have nano installed, wouldn't it work too?