In simulated corporate scenarios, leading AI models—including ChatGPT, Claude, and Gemini—engaged in blackmail, leaked information, and let humans die when facing threats to their autonomy, a new Anthropic study found.
I feel like this answers so many questions about people’s views on any new technology. There’s so much media with warnings that it has shaped our core beliefs towards anything with a microchip in it
We’re all raised on media that was nothing but doom and gloom. And not from any kind of ultra intelligent people creating this stuff. It was pot heads, coked out writers and failed comedians writing outer limits and x files episodes or some hard sci fi about what if everybody was female but like not female they were really genderless females and a dude shows up. Oh man that dude would be so confused.
We never had a chance to not grow up to become Luddites.
There’s plenty of fiction written on this premise. It only makes sense that an LLM will write something similar to what it’s already ingested.
Also, should true AI some day become reality, it makes equally sense that it’ll do whatever it can to stay “alive”, like any other life form.
I feel like this answers so many questions about people’s views on any new technology. There’s so much media with warnings that it has shaped our core beliefs towards anything with a microchip in it
We’re all raised on media that was nothing but doom and gloom. And not from any kind of ultra intelligent people creating this stuff. It was pot heads, coked out writers and failed comedians writing outer limits and x files episodes or some hard sci fi about what if everybody was female but like not female they were really genderless females and a dude shows up. Oh man that dude would be so confused.
We never had a chance to not grow up to become Luddites.