ncase

More AI comix, and project update

Added 2023-04-01 00:47:34 +0000 UTC

(⏱ reading time: 2 minutes)

Every. Week.

Every week that passes, my mega-explainer on AI Safety/Alignment gets more and more obsolete.

It's so weird to see how fast the Overton Window's shifted recently, with the NYT and The Atlantic and Fox and big petitions with serious-professionals all calling for AI Capabilities research to slow down or even stop, at least until AI Safety can catch up.

So: in order for my AI explainer to not be obsolete before it's done — and to keep me motivated, and to not keep y'all waiting longer — I will publish my AI explainer in three parts! (Each part ~30 minutes reading each.) Part I will be out probably late-April, next month. Probably. Ugh, you know how bad I am at self-prediction.

Apologies this is taking so long! (In March I moved places, and fell uselessly sick with brain fog — still not Covid, somehow.)

In the meantime, have a sneak peek at 3 AI Safety comics: (please don't share for now, thanks! it'll be out soon enough anyway)

I realize I probably could've finished this explainer by now if I co-wrote with GPT & co-drew with Stable Diffusion. But I shan't. I shall stand before the steamroller of progress yelling "STOP" 'til I get squooshed.

STOOOOOPPPPPPPPP,
~ Nicky Case

P.S: Quick signal boost — Slime Mold Time Mold's running a sequel citizen-science study, the Half-Potato Diet! They're still taking signups, if you're interested. Here's my data from when I tried half-tato myself. Bon a-potato-tite! 🥔🥔🥔🥔🥔