Browse Comments

Generative AI replacing creatives is backwards progress. It eliminates an entire field of people and gives far too few benefits for them back. The only people who would benefit from ai replacing writers and artists are the upper management.

Victims and Perpetrators.... So a bunch of hormy teen guys decided to do deepfakes of the token "hot" girls of their class?

That's what they said about Czechoslovakia back then. Which lead to the Nazis getting hold of the most advanced arms manufacturing industry in Europe at the time, enabling them to scale up their war machine in the first place. This time everything is documented and there won't be any "I didn't know" or "I wasn't part of this".

Of course a model is going to produce results consistent with its training data. It's hard for a human reviewer to purposely ignore the name, but trivial to make an automated tool to do so if objectivity is a goal. This seems like a win for including AI in a screening role. There's no reason to include the name if you don't want it to influence the score.

People talking to a word generation algorithm like they're afraid of it is the most beta thing I've ever heard.

This completely misunderstands the nature of fascism. Go look at what you just wrote and really think about it: which is the more important part? The part where you had an AI read 6 years of media posts...or the part where you can, without consequences worth considering, kill or otherwise inflict harm against 100 people? This is why I don't take data privacy people seriously, because they're no…

So as far as I know, chinese cars are mainly sold on the chinese market, so to their own populace, and in SEA in general. Europe and the US see only few of those (afaik at least). While this still makes them the biggest car manufacturer worldwide, how would not embracing neoliberalism have changed that in the slightest? I also kinda miss the point where China became world's leading robotic and sp…

Humans are more susceptible and AI can be fixed more easily. AI isn't a joke... And in 10 years, give or take, you will definitely not think its a joke when you have to compete with it for a paycheck.

That kind of energy would be better used on learning a new skill. This field is going to be automated away before long anyway.

innovation & interest in AI/ML has always roughly followed a f(x) = x sin(x) for decades now

\> They used the entire internet and pop culture as training data. What else barring extreme architectural advancements is supposed to sustain exponential growth? Supposedly you'd use AI's to generate a seemingly unlimited amount of (good quality) synthetic data and also agentic workflows to automate the entire training process, as well as agents conducting AI research to improve itself. Hypo…

Exception isn’t the norm. Even without chatGPT, Adam would have found ways to do it. My heart goes out to the family but restrictions have never worked out well for anyone. People will be even more incentivized to find loopholes. Iirc, I read that GPT did try to steer Adam from his agenda multiple times but maybe he asked the questions in between prompts so maybe the model didn’t follow earlier…

That's just patently false, AI is much better at understanding than it was 12 months ago.

It hurts the commercialisation of art, and makes it more difficult for people looking for a career working in producing commercial art. Which is sad for them, but the genie is out of the bottle. Getting to see these soulless corporations eat each other over AI to try and protect copyright, their greatest tool for enforcing their monopoly on the creative space, has been a joyful experience.

What happens if, let's say some nefarious actor hacks a big player in the game > said actor uses the system to hack the other 5/6 large companies LLMs > converts every LLM into one central database > has so much power and control over all the systems that are being leveraged > creates an impenetrable force of a system > that system is able to detect any attempts to repair the damag…

­G­o­­d, y­o­u pe­opl­e... *You ar­e cu­rr­e­ntly using A­­I.* R­e­­dd­­it uses A­­I to p­e­rso­nalize your feed and for mod­era­tion. That ­go­ogle sea­rch you did today uses A­I. Your email uses AI to det­ect spa­­m and feed you ad­s. A­I is integrated into the pi­­p­eli­ne of virtually every onli­ne service. It's everyw­here because it's a us­ef­ul to­ol. J­­f­c, you l­o­­w-in­­fo­r­m­at­i…

To be fair, the "don't be evil" motto was silly to begin with. And DEI is pretty evil (it's racist and quite destructive), they should never have adopted it, and they're probably paying a hefty price for that misstep. I think some of the biggest problems with Google have to do with the way it's run and staffed (which may in large part be due to the DEI problem). From what I understand, it at som…

“The DEFIANCE Act would impact individuals, like those Grok users creating deepfaked nonconsensual intimate imagery.” What about the platform providers? They are the ones enabling the content and with the $$ for lawsuit payouts. The Take It Down Act, made it a federal crime to post nonconsensual sexually explicit deepfakes and was signed into law last May. yet that did not stop the content from…

File a Better Business Bureau complaint. Those are automatically published. Outline everything. Take that leverage away from them. Feels an awful lot like extortion. Good luck!!

And us and AI. If it quacks like a duck maybe it is time to care about its feelings it is emulating so well.