Interesting Links for 01-12-2022

[andrewducker]

1. 1 in 4 US hiring managers say they are less likely to move forward with Jewish applicants

(tags:usa Jews bigotry jobs OhForFucksSake )

2. A list of practical feminist behaviours for men

(tags:men feminism advice )

3. Copyright, cost, and ageing code: why some games disappear from Steam

(tags:games history copyright )

4. Ofgem reveals plan for cheaper, cleaner electricity supplies

(tags:electricity renewables UK thefuture )

5. Twitter Thrills Far-Right Trolls by Silencing Left-Wing Voices

(tags:Twitter censorship ElonMusk )

6. Disney's new neural network can change an actor's age with ease

(tags:age cgi ai Disney )

Comments

[dewline]

1. And still the cruelty + stupidity...

2. Thanks. Been doing the work imperfectly, I suspect. The reminders of what's good practice do matter.

3. Some tragedy, some cruelty, some...not sure what to make of it.

5. Yep. Been seeing the consequences of some of that toxic behaviour. ("Shitty" doesn't go far enough in naming the evils going on here.)

[bens_dad]

6. Faces re-aged well, but they appear not to have attempted to re-age the hair. I noticed one of the alternative ageing methods *did* age the hair and that seemed much more convincing to me.

[ My PhD was in this area; interesting to see what is being done now. ]

[andrewducker]

Hair definitely makes a big difference. I can't help but feel that the face shapes haven't changed as much as I'd expect, but it may well be that I'm expecting a more exaggerated effect than actually happens in real life.

Your PhD sounds fascinating!

[bens_dad]

I got the impression from some of the videos that the ageing stopped between the eyebrows and the forehead, as if the face was cut out of the head, re-aged and put back.

What is odd, is that the training data was generated with StyleGAN2 but the hair morphs much better in the StyleGAN2 video. Otherwise I might have said that although the AI probably doesn't explicitly cut the face out, perhaps the training set did.

> Your PhD sounds fascinating!
Text at https://www.aitchison.me.uk/thesis/thesis.pdf
Code for appendix C in the same directory.

[calimac]

2) This is a good list, though I am a bit uncomfortable with #35, "Self-identify as a feminist." I don't think it's up to men to decide whether they qualify as feminists, owing in part to too many self-identified feminist men who turned out to be nothing of the kind (cf Joss Whedon). They've kind of spoiled the self-puffery inherent in such self-identification. The most I say is that I do my best to support feminist causes, and I don't think it's an evasive dismissal of the word I actually used to put it that way.

[andrewducker]

I'm always torn on that one. Some people really object to it, some people think it's great, and I'm not sure there's a right answer for everyone!