#quedadequatro

[perpetuallyfive]

quedadequatro replied to your post: Someone needs to stop the straights.

Not to "well, actually" this, but it's worth looking at their actual findings https://docs.google.com/document/d/11oGZ1Ke3wK9E3BtOFfGfUQuuaSMR8AO2WfWH3aVke6U/edit#

They didn't create an AI, they studied existing technologies and are concerned about what they're capable of and how they might be used to violate privacy.

There’s promising follow-up with people really examining their methods and interpretation, too callingbullshit.org/cas…

Okay, this is going to be long because, for the sake of humor, I was being much more flip and reductive than my actual position, which is that the only thing this study of pre-existing technology proves is 1.) computers are capable of picking up on nuances to a greater degree than the human eye and 2.) a particular small subset of a minority group belonging to a particular racial profile, self-identifying, and self-selecting by engaging with the same dating site in the hopes of seeking out people attracted to the same sex -- and would be inclined to highlight features they considered desirable through framing, lighting, etc -- are likely to share certain features.

This doesn’t suggest anything about the community as a whole, as people who do not align with a particular beauty standard do not always engage with mainstream dating websites largely targeted at a heterosexual demographic in the first place, and it discounts the spectrum of gender and sexuality entirety while also only using a white sample group since facial recognition technology has so far largely been designed around the use of white faces. In terms of a representative sample of the “gay” community, it’s abysmal. Their reading of sexual identity as a binary option is laughable and shows a lack of interest or understanding of the community they are attempting to analyze. Being overly reductive is only more scientific if you assume that there are important conclusions to be drawn between studying the fraction of the population that would qualify as a Kinsey 0 or 6.

The conclusions that they do reach about visual differences in photographs taken by people seeking out sexual partners suggesting specific hormonal realities feels speculative at best. If you were to look at a sample of photos from myspace in the 90s, you might be under the impression that all women at the time had very large breasts. 

Their suggestion that their work is in some way beneficial to the gay community at large because it points out dangers inherent to facial recognition software is outright comical as it suggests that somehow gay people didn’t already know that being outed is dangerous and we needed a group of scientists to tell us this shit. It feels like a ridiculous attempt to justify the intentions as inherently noble. I understand the argument that what they’re doing is intended in some ways to identify the dangers to the public at large as though that might somehow curtail those who would use this technology for ill, but I just don’t believe it could have any realistically beneficial impact.

All it seems to do is prove computers have better eyesight than people (something we already knew) while at the same time validating incorrect assumptions about the overall makeup of the LGBTQ community, reaffirm biases in the minds of bigots, provide an outline for governments of countries that are eager to eradicate their queer populations, and I guess maybe give Google a moment of pause before they absolutely continue to do exactly as they intend. This Genie is long since out of the bottle and I don’t think this analysis of a subset of a fraction of a minority does anything to actually help anyone except the scientists who wanted a lot of people to read their research.

Which all boiled down to and largely remains: someone needs to stop the straights. (Addendum: it’s way too late.)

[perpetuallyfive]

quedadequatro replied to your post: Vampire bc you reblog really gay shit (and im here...

I will defer to my dad, Louis Virtel on this one “Vampires went to, like, the Berklee School of Music. They’re like gay, or very straight. Fast, but they’re slow. I’m very scared but also, "God you’re so thin!”“

This captures so much truth and feels so succinct. 

[perpetuallyfive]

quedadequatro replied to your post “Tag yourself I’m Heather O’Reilly looking like she’s about to fucking...”

Her serving the high ponytail Ariana realness is giving me life.

FUCKING YES.

[phaaelbarba]

Diary entry 26/11/2015

Toshio, Queimado, eu, Sumu

2 new moves today.

1️⃣Negativa Frontão: same as regular negativa, but moving down and back from ginga to face front.

As you sink down, front foot pulls back a little to close the negativa. Push forward again on rising.

2️⃣Martelo from Queda-de-4: To get decent power and accuracy there are several points to consider..

1. The queda-de-4 base must be quite “folded” in order to be able to stretch out with body and leg together. If hands and feet start far apart, the body must also be open (less protected) and therefore has less ability to extend. Think of an accordion that is already fully stretched, it cannot make any more sound until it is compressed.

2. In order to attack the side of partner’s head/body (which is directly in front of you), your hips must be rolled/rocked to the kicking side. This also helps you to kick with the top of your foot.

3. Extension is not just of the leg, but the whole body (as in so many moves). As 1. says, body starts folded and compressed, then lift bum/lower back up, stretching whole body to greatly increase reach and power.

👣 An effective ganxo (?) can also be made using these techniques