Even if o3 was faster, cheaper, and smarter, I’d still use Deepseek r1

I've been using chatgpt since the gpt3 release, but R1's detailed thought process is a seller feature for me. Even though o1/o3 displays its thinking as well, it is way too abstracted and vague.

Sometimes I got inaccurate answers from R1 on some very narrow and specific topics, but I was able to adjust my prompt, since I saw that R1 was building from the wrong assumptions or outdated information.