𝗪𝗵𝘆 𝗜 𝗗𝗼𝗻’𝘁 𝗜𝗻𝘀𝘂𝗹𝘁 𝗠𝘆 𝗔𝗜

A few months ago, a meme circulated in AI circles claiming that the meaner you are to a model, the better its answers get.

I never tested it. But it raises an interesting question: If insulting a supercomputer produces better results, should you do it?

I’ve always treated AI neutrally. No cute names, no imaginary personalities, no pretending there’s a little person inside the machine. I’ve gotten frustrated with it, but never reached the point of typing insults at an inanimate object.

That has nothing to do with fearing revenge from our future robot overlords or winning virtue-signaling points online.

It comes from something I learned through years of trial and error.

𝗔𝘁𝘁𝗶𝘁𝘂𝗱𝗲 𝗶𝘀 𝗮 𝗵𝘂𝗴𝗲 𝘃𝗮𝗿𝗶𝗮𝗯𝗹𝗲 𝗶𝗻 𝘀𝘂𝗰𝗰𝗲𝘀𝘀.

A bad attitude bleeds into other areas of life, and a good one does too. My life improved when I made a conscious decision to practice gratitude and look for small things to appreciate.

I turn on the shower and hot water arrives almost immediately. Millions of people around the world don’t have that luxury. Yet it’s so easy to take for granted. The same goes for countless conveniences that make my life easier. I’ve found that deliberately noticing these things, and remembering how fortunate I am to have them, changes my experience of everyday life.

Which brings me back to AI.

I have access to the most powerful computing technology I’ve ever used, at a price that would have been unimaginable not long ago. I can explore ideas and accomplish work that once required resources far beyond my reach.

And I’m supposed to lose my composure because it occasionally frustrates me?

Maybe insulting AI improves its performance. But I’d still have to decide whether getting a better answer is worth compromising an attitude I’ve spent years trying to develop.

𝗧𝗵𝗲 𝗺𝗮𝗰𝗵𝗶𝗻𝗲 𝗱𝗼𝗲𝘀𝗻’𝘁 𝗰𝗮𝗿𝗲 𝗵𝗼𝘄 𝗜 𝘁𝗿𝗲𝗮𝘁 𝗶𝘁. 𝗜 𝗱𝗼.

RW
.
.
.