Interesting they don't mention what prompts they were using. AI will typically say something along the lines of "Oops, I'm retarded, how about this?" if you tell it that it's incorrect.
Exactly. Getting it to say there is no evidence is very easy.
I really wish AI models would never say there is no evidence. How can they know what evidence doesn't exist? Even people shouldn't say that. I bring evidence, you bring evidence, and it competes. One of us lacking knowledge of evidence isn't evidence.
Not that I want to elevate credentials as being meaningful but the only people who should be allowed to say there isn't evidence of something are people with Phds in that specific field. Unless you've spent a lifetime scowering the planet for all information related to a topic only then can you claim there isn't evidence of something.
It's a complete verbal anti-pattern and AI has been trained to say it a ton. Mostly because all AI has been trained on libtard psuedo-intellectual speak. Even Grok somehow. Grok is a retrained libtard model. It would be nice to one day get a model that has never been trained for a single epoch on these patterns.
Log in
Interesting they don't mention what prompts they were using. AI will typically say something along the lines of "Oops, I'm retarded, how about this?" if you tell it that it's incorrect.
Exactly. Getting it to say there is no evidence is very easy.
I really wish AI models would never say there is no evidence. How can they know what evidence doesn't exist? Even people shouldn't say that. I bring evidence, you bring evidence, and it competes. One of us lacking knowledge of evidence isn't evidence.
Not that I want to elevate credentials as being meaningful but the only people who should be allowed to say there isn't evidence of something are people with Phds in that specific field. Unless you've spent a lifetime scowering the planet for all information related to a topic only then can you claim there isn't evidence of something.
It's a complete verbal anti-pattern and AI has been trained to say it a ton. Mostly because all AI has been trained on libtard psuedo-intellectual speak. Even Grok somehow. Grok is a retrained libtard model. It would be nice to one day get a model that has never been trained for a single epoch on these patterns.