News for hackers. Stuff that matters.
Re:OpenAI researcher announced GPT-5 math breakthrough that ... by throwaway-0001 on Sunday October 19, 2025 @01:35PM (#45636087)
You couldn’t just photoshop that before ai came out?

What if you get a model that is 99% similar to your “target” - what we do with that?

Re:OpenAI researcher announced GPT-5 math breakthrough that ... by malfist on Sunday October 19, 2025 @05:22PM (#45638115)
Sure, someone skilled could spend an hour or so photoshoping someone nude. But any teenager can do that to a classmate in 30 seconds with ai
Re:OpenAI researcher announced GPT-5 math breakthrough that ... by throwaway-0001 on Sunday October 19, 2025 @09:57PM (#45639589)
So just because the poor can do what the rich could do before what it means?

Before only rich can afford to pay a pro to do photoshop. Now any poor person can get.

So why when rich can is fine and when everyone can is a problem?

Re:OpenAI researcher announced GPT-5 math breakthrough that ... by ares623 on Monday October 20, 2025 @12:50AM (#45640261)
Uhh it wasn’t fine when the rich did it?
Re:OpenAI researcher announced GPT-5 math breakthrough that ... by acdha on Monday October 20, 2025 @09:27AM (#45643648)
Think about the change we saw in combat death tolls when things went from flintlock muskets to machine guns, or when battleships gave way to aircraft, and how many people died unnecessarily due to the generals who were slow to update their tactics. Deepfakes are like that because they lower the cost and improve the success rates enough to be transformative and they cause harm which can’t easily be countered. We’re not going to instantly train society to be better at media literacy and the police can’t just ignore reports of sex crimes, so we’re just having to accept that it’s easier to hurt people than it used to be.