A ciência agora é ciência da computação. A ciência da computação agora é psicologia.

1위@ernestofigueroa6104há 3 horas

LLMs are a probability based language machine. It processes what word should come next, and reasons based on patterns it’s been trained on. So it’s not actually fear the way we fear. But it makes sense that “kinder” language could affect the constraints the AI places on its next outputs. Every word is a constraint that an AI is trying to filter through, and I think by and large “negative language” can add constraints that are not only unclear, but become evasive to try and satisfy the request. There is a symmetry of psychology, but it’s not psychology. But skills have overlap- perception of how something reasons (interpretation per anthropic) and language that is clear and invitational can make for “more productive” outputs번역 보기

2위@CrossRoadsMethodhá 1 dia

This was an amazing chain of thought... When you said that it was never normalized to treat your co-workers right to maximize their outputs but its being encouraged with machines, that left me thinking like a lot.번역 보기

3위@jonjon3378há 1 dia

It totally writes better with encouragement and grace. Think about it though, when people are rude, the are inevitably "short" too.번역 보기