Report: OpenAI holding back GPT-4 image features on fears of privacy issues
Report: OpenAI holding back GPT-4 image features on fears of privacy issues
GPT-4's image capabilities can recognize certain individuals, according to NYT.
The real danger of AI is not from it gaining sentience and murdering us, it's that it will give fascists the ability to concentrate power and control into an incomprehensible and far reaching web of censorship, profiling and surveillance from which humanity will not be equipped to escape or even really understand well enough to combat. AI will end democracy forever and people won't even recognize that it's happened
censorship, profiling, and surveillance
Also inventing evidence to scuttle anyone dangerous in the court of public opinion.
A union leader getting you down? Here's a "leaked tape" of him sexually harassing an "anonymous whistleblower."
Corporate democrat getting primaried by a socialist teacher? Sure would be a shame if it was "discovered" that he posts in a small dark web cp forum that has "posts going back years" and he "accidentally outed himself."
Once generative ML gets good enough that it's indistinguishable from legitimate evidence, it'll be possible for anyone with that kind of power to construct whatever reality is convenient. We were always at war with Eastasia.