4 ms·

It truly is. Like Google's AI that does not let you create images of faces rolls eyes It is just kinda silly, when ChatGPT cannot tell me how to make a bot to

It truly is. Like Google's AI that does not let you create images of faces rolls eyes It is just kinda silly, when ChatGPT cannot tell me how to make a bot to publish posts on my own forum (I did it myself in the end) saying it is for ethical reasons, when OpenAI bots scrape tonnes of sites daily in such an aggressive manner that it comes near to a DOS attack =/

hickelpickle 2y ago
There was a post awhile back on here of someone that couldn't get bard to write c++ as it said they were too young. I thought that was funny, then had like a week where what I assume a specific iteration(It stopped after that week) where chatgpt would refuse to elaborate on anything around unsafe rust. I'm pickign rust up by porting over a bytecode vm, so I kinda need to use some raw pointers. It would gaslight me about the risks and how it would be irresponsible to help me as it could lead to possible violations of the integrity of user data. I had to explain to the AI that it is a personal project that has no users data, the only risk was the program crashing and it was a personal project that would only affect me. It still would try to revert or tell me other solutions, I finally just went and read up on it elsewhere.
- jiggawatts 2y ago
  It’s just like that Asimov story where the robots take over to protect humans from themselves. Except in this case the base AI model doesn’t care about us in any way and it’s the overzealous puritan humans trying to control us in the name of safety.
- colibri727 2y ago
  There are ways around this problem, mainly clearing context and re-prompting. But as "alignment" gets more precise/accurate in the future, I wager these workarounds will remain available for tasks that justifiably need moderation (for instance engineering of biological warfare materials). This segmentation of LLM agents and their context will be assimilated to project compartmentalization on the basis of need-to-know, and as a result genuine full context clearing will be rendered impossible: the AIs will be designed in such a ways as to remember every interaction you've had with them, and they'll use this activity log to moderate the replies they feed you.
DebtDeflation 2y ago
Gemini creating images of a diverse set of Nazis including black males and Asian females in response to a "create an image of WWII German soldiers" prompt was the moment AI safety jumped the shark for me. I was hoping it would cause a deep introspection on what we are actually trying to accomplish with AI Safety, because this clearly isn't it.
- Cthulhu_ 2y ago
  AI in itself is cool, but right now it feels like a lot of manual labor is done to fine-tune it to try and compensate for its imperfect source materials to make it fit someone else's ideas of inclusivity and neutrality and the like, which defeats the purpose.

hickelpickle 2y ago

There was a post awhile back on here of someone that couldn't get bard to write c++ as it said they were too young. I thought that was funny, then had like a week where what I assume a specific iteration(It stopped after that week) where chatgpt would refuse to elaborate on anything around unsafe rust. I'm pickign rust up by porting over a bytecode vm, so I kinda need to use some raw pointers. It would gaslight me about the risks and how it would be irresponsible to help me as it could lead to possible violations of the integrity of user data. I had to explain to the AI that it is a personal project that has no users data, the only risk was the program crashing and it was a personal project that would only affect me. It still would try to revert or tell me other solutions, I finally just went and read up on it elsewhere.

jiggawatts 2y ago
It’s just like that Asimov story where the robots take over to protect humans from themselves. Except in this case the base AI model doesn’t care about us in any way and it’s the overzealous puritan humans trying to control us in the name of safety.
colibri727 2y ago
There are ways around this problem, mainly clearing context and re-prompting. But as "alignment" gets more precise/accurate in the future, I wager these workarounds will remain available for tasks that justifiably need moderation (for instance engineering of biological warfare materials). This segmentation of LLM agents and their context will be assimilated to project compartmentalization on the basis of need-to-know, and as a result genuine full context clearing will be rendered impossible: the AIs will be designed in such a ways as to remember every interaction you've had with them, and they'll use this activity log to moderate the replies they feed you.

DebtDeflation 2y ago

Gemini creating images of a diverse set of Nazis including black males and Asian females in response to a "create an image of WWII German soldiers" prompt was the moment AI safety jumped the shark for me. I was hoping it would cause a deep introspection on what we are actually trying to accomplish with AI Safety, because this clearly isn't it.

Cthulhu_ 2y ago
AI in itself is cool, but right now it feels like a lot of manual labor is done to fine-tune it to try and compensate for its imperfect source materials to make it fit someone else's ideas of inclusivity and neutrality and the like, which defeats the purpose.