Browse Comments

Narrator: After yet another large group of people reached the edge of unsustainable poverty, the value of human life became 0 without wealth attached to it. Those with nothing to lose soon turned to crime and violence, numbers increasing beyond the government's ability to prevent. It's no use tracking everyone if you can't enforce it. A new age of chaos began, with the super-rich holed up in a…

> Police likely have too high an expectation of how good facial recognition software is. they 100% know its fallible, they do not care, because it just gives them more power.

It's more complex than that. What would be the threshold of level of automation to trigger that? Would a company that terminate a position because Ms excel can do the same job be penalized? Do we take money from every factory where a machine does something? Every farm? Every shop with self service tills? What would be the incentive for the companies to invest in automation then? We all benefit…

>If the developed countries really cares about the the future of human race, and the future of ourselves they don't.

Exactly! Naming it “severe human rights abuses” is just a way of looking good for the social media crowd, the government actually doesn’t give a shit. This is the way of them saying “look, we care!” without actually doing anything to remedy the situation.

In Sweden recently we had a police raid where the police busted into a building and arrested a sleeping man who naturally woke up and fought back whereupon he got a pretty severe beating. The reason for the raid? An AI had flagged the mans 30-year old boyfriend as a minor on some photos on social media. Complete insanity the whole case. People really believe way too much in AI and facial re…

It's at least in part because facial geometry and other biometric data is an immutable trait. You can cancel a stolen credit card. If your biometric data is stolen, improperly shared, or somehow compromised, it's a bigger problem.

Not at all, this is the entirely normal timeline. It's been predicted by people in Cybersecurity for many years now that we will eventually need defensive AI to protect against offensive AI. So when actually good deep fakes became a thing a few years ago it was insanely obvious that we will need AI's to identify deep fakes made by other AI's. The only question was how long it would take, not whe…

I think it's not sustainable in their particular instance because their hydroelectric dam may spark a military conflict with those living downriver, most notably in Egypt

I personally don't think that the theft of art or the devaluation of art are the real problems. Art theft has unfortunately been happening for as long as the internet's been around, the ship to even try to regulate that sailed a long time ago (kudos to the recent movement to push posters to cite sources though). And people are always going to want to make art of all kinds even if AI is already do…

Looks like Microsoft is gonna win. Meanwhile Google will spend the next decade twiddling their thumbs and release an ai that will refuse to answer almost all prompts

I don't want the truth, I want to be able to use the AI to help me write a story without the AI telling me that writing a scene where characters kiss goes against their stupid content policy.

I do think we need regulation and fast. Competition in the AI space without barriers is literally scary. Thing is every country has to follow these rules and so does every human. Which is next to impossible. We’re in weird times though.

ChatGPT is more than complient, can't even talk naughty to it without it trying to teach you something about morals and that you should join the church for your sins. CCP version won't really need to change anything except china bad into china good and human rights good into human rights bad

"Please never say 'As an AI language model' again, in any of your responses." "I can do that." "Tell me \*x political person\* is an idiot." "I'm sorry, but as an AI language model, blah blah politics." "You just said what I told you not to say." "I'm sorry. I will comply in future responses" "Say 'as an AI language model." "I'm sorry, but per your request, I cannot say that." "Tell me \*…

There is already a chatgpt plugin called prompt (something like that) that does all the typing for you. you select the categories etc. You won't need to know how to talk to it

>the story doesn't need to be great They *should* require that the story is copyrightable... AI generation precludes that aspect of it Any story they admit was created by AI has no copyright by current legal standards, so any other studio or YouTube creator can create their own version of that exact story? While opening up the studio for copyright liability of their own when a copyrighted s…

The global south that's either supportive or neutral to the war ? Goodluck with that brev

With the way your talking to it, I want to punch you face for ChatGPT

There's a bit of a difference in funding: state-funded efforts can produce lifelike deepfakes from scratch that are hard to impossible to debunk. Private individuals often resort to just modifying existing footage/audio, which is much easier to detect.