When your only tools are a reciprocating saw and a cello, all your problems start looking like digital electronics.

These faces are breaking my brain. I never thought I'd have use for the phrase, "your axes of reality are not orthogonal".

Who needs friends when you can make them with Jupyter and StyleGAN? youtu.be/4spmAO1JIfc

Listening to Superstition by Stevie Wonder.

"When you believe in things that you don't understand, then you suffer. Superstition ain't the way."

Words to live by, for any scientist.

Mandatory Windows 10 update wakes my laptop (and me) from sleep, fans blazing. I log in and try to ssh out via linux services (as usual). Then this happens. Even after reboot.

Congratulations, Windows team. You broke the only thing I use it for (aside from games).

Emitter base prototypes, left to right: fdm print, resin print, slip cast porcelain (hand finished), milled wood, milled ceramic. Let's hope this one works. Thanks, @sirus for the mill work!

Turns out Macor does machine like a dream. Left prototype is wood, right is ceramic.

(though i bet we could get conservatives to support rescinding DST en masse if we convinced them that yoko ono invented it)

for those wondering how to establish trade relations with crows:

- feed crows food A regularly
- when crows decide to bring you gifts, feed them food B
- when crows bring you different categories of gifts, feed them a food-per-category, ex: food C for jewelry, food D for paper money, etc
- crows will recognize these exchange patterns and opt to bring you things in order to acquire desired treats

good luck! 🐦

Also, for the record: The average cat is a grey tabby. The average car is a grey BMW. The average bedroom is brown and has two table lamps.

You can also ask for the average of the dataset. Here are the average Flickr face and average celebrity. The "gal" on the left shows up in many StyleGAN articles. Not too surprising to see brown hair and a brown background. Dark skin must be underrepresented in the training data.

StyleGAN's "truncation trick" pushes weights towards the average, resulting in better quality images. You can of course push weights away from the average too. I find this works best with the celebrity faces dataset.

StyleGAN photobomb faces nightmare fuel Show more

Sent the last toot to gpt-2. The reply:

But the more complex things are, the more we find ourselves in a very different world. We need to be like, "Fuck you! You're the reason I made this. I'm the reason you're here."

I know this is all bullshit, but I know I can help you.

Show more

Where the 1337 meet to federate. Home of the most interesting local feed on the Fediverse!