Teaching Claude to Be Lazy

I’ve been watching AI development for a long time. I found LessWrong around 2012-2013, and managed to get myself worked up about the oncoming singularity. I managed to chill out about it, but interest and excitement for AI remained. The initial Deep Dream image generation, Alpha Go, etc, were all so exciting. And then GPT-2 came out. Over the last five years, people have been making wild claims about the utility of present AI. Not “the AI that you’ll have soon,” but the current generation st...