There’s an old charge that Europeans often level at Americans. Sure, the right is spiteful, mean, and nasty. But it ofte

Image generation algorithms leapt forward in capability in 2014, with the creation of generative adversarial networks, or GANs. The idea, which A.I. researcher Ian Goodfellow originally thought up during an argument at a bar, was to pit algorithms against each other to generate the best outcome. To generate an image, you would have a “generator” and a “discriminator.” The generator would make images, and the discriminator would try to determine if it was real or fake, based on real images it had been trained on. Only the most realistic images would be accepted by the discriminator, ensuring that the final result was the cream of the A.I.-generated crop.

There are now multiple open-source, freely available methods for creating synthetic faces built on GAN architecture. And as cloud services like Amazon’s AWS and Google Cloud have become easier to access, so has the ability to train these algorithms. The most well-known in the A.I. research world is StyleGAN, made by Nvidia. It was released in December 2018, and while able to produce incredibly high-quality images of fake faces, the images also contained strange blobs and digital artifacts. Less than a year later, the Nvidia team released StyleGAN2, which fixed the algorithm’s architecture to prevent those blobs and artifacts from forming, as well as improving the fidelity of the images.

These technologies are still in their infancy, and they’ll continue to become more accurate and convincing. Some of the applications of these technologies are genuinely entertaining — check out the Avengers singing the “Sweet Child O’ Mine” scene from Step Brothers — but ultimately, these algorithms are also now much easier for anyone to use for malicious ends. And without any recourse, deepfake’s harms might outweigh their slight entertainment value.

These technologies are still in their infancy, and they’ll continue to become more accurate and convincing. Some of the applications of these technologies are genuinely entertaining — check out the Avengers singing the “Sweet Child O’ Mine” scene from Step Brothers — but ultimately, these algorithms are also now much easier for anyone to use for malicious ends. And without any recourse, deepfake’s harms might outweigh their slight entertainment value.

These algorithms are able to be adapted to different domains. By training the algorithms on pornographic images rather than just faces, the system was able to adapt to generating something it might not have been ever intended for.

But this technology didn’t show up overnight. There’s a clear path from some of the earliest modern image-generating algorithms to this phenomenon of A.I.-generated porn. Here’s what it looks like.

St. Osyth is a cursèd little village in the county of Essex, 83 miles east of London, at the edge of the North Sea. The town’s 4,600 souls live in medieval cottages arranged around a 12th Century monastery, and in cheap mobile homes that the British call caravans. St. Osyth’s exact origins remain a mystery, and over the centuries its townsfolk have survived floods, invasions, and monsters both imagined and real. In fact, every page of its wretched history is soused in the supernatural.

In the more peaceful era of 2005, 30-year-old sales executive Vanessa Mitchell arrived in St. Osyth to view a house for sale. Number 14 Colchester Road is a former medieval jail known as “the Cage.” It famously housed the 13 St. Osyth witches, including Ursula Kemp, who was hanged for murdering her neighbors with black magic. “I remembered seeing the plaque outside,” Mitchell told me on a Skype call. “I don’t remember being scared of it.” She had grown up in the village and used to walk past the mustard-colored building on her way to school. “My earliest memories are just being fascinated by the house. I remember being drawn to it,” she said.

In 2005, Vanessa Mitchell moved into her dream home, a former medieval jail where England’s witches waited to hang and burn. When paranormal phenomena forced her to flee, she became convinced it was possessed by evil spirits. This is her true story.

Goodfellow’s initial research on GANs performed well on industry benchmarks, but many of the images he created still looked like hellish blobs that only represented ideas in abstract and inhuman ways. By 2016, other researchers had started experimenting with the technique and found ways to make lifelike images, albeit at small resolutions. One of the standout papers of the time showed how researchers could generate realistic images of bedrooms, as well as rudimentary attempts at generating faces. This research again showed that GANs were able to adapt based on the kind of data they were trained on. The idea worked as well for faces as it did for bedrooms, meaning the networks were actually able to identify patterns in a variety of different types of images.

The village is named after the granddaughter of England’s last Pagan king. According to legend, Osyth was beheaded by Danish Vikings but managed to walk to the nunnery carrying her severed skull in her hands. In 1171, the village was burned down by a fire-breathing dragon. Then came the witches. During a Satanic panic in 1582, 13 local women stood trial for witchcraft and two swung from the gallows.

Earlier this week Vice reported the latest example of one of these harms: Coders were using images of sexual abuse to train algorithms to make porn. The article details how nonconsensual images were compiled by an anonymous PhD student into a dataset and combined with off-the-shelf algorithms to generate custom videos.

http://live07.colomboserboli.com/swg/videos-Piestanske-Cajky-Poprad-v-en-gb-1qvb-1.php

http://go.acaps.cat/kgr/v-ideos-piestanske-cajky-v-poprad-v-sk-sk-1jlb-7.php

http://streaming7.actiup.com/nez/Video-piestanske-cajky-v-poprad-v-sk-sk-1qmg-12.php

https://assifonte.org/media/mln/videos-Mobis-Phoebus-Dongbu-Promy-v-en-gb-wgw30122020-.php

http://news24.gruposio.es/ktn/videos-Ansan-Rushcash-Seoul-Woori-Card-Hansae-v-en-gb-tku-.php

http://live-stream.munich.es/exd/v-ideos-BCM-U-Pitesti-Phoenix-Galati-v-en-gb-wzl-.php

https://assifonte.org/media/mln/videos-Mobis-Phoebus-Dongbu-Promy-v-en-gb-car30122020-.php

http://streaming7.actiup.com/nez/videos-Asseco-Resovia-Rzeszow-MKS-Bedzin-v-en-gb-1kww30122020-.php

http://live-stream.munich.es/exd/video-BCM-U-Pitesti-Phoenix-Galati-v-en-gb-sto30122020-.php

http://go.acaps.cat/kgr/Video-Asseco-Resovia-Rzeszow-MKS-Bedzin-v-en-gb-1ndj-.php

http://main.dentisalut.com/hxd/Video-ceara-v-internacional-v-pt-br-1ozw2-4.php

http://live07.colomboserboli.com/swg/Video-kometa-brno-v-aukro-berani-zlin-v-cs-cs-1lbr-21.php

http://live-stream.munich.es/exd/v-ideos-BCM-U-Pitesti-Phoenix-Galati-v-en-gb-zwb-.php

http://go.acaps.cat/kgr/Video-Asseco-Resovia-Rzeszow-MKS-Bedzin-v-en-gb-1bqd-10.php

http://main.dentisalut.com/hxd/videos-ceara-v-internacional-v-pt-br-1nbw2-16.php

http://main.dentisalut.com/hxd/Video-BK-Decin-BK-Opava-v-en-gb-lbe-.php

https://assifonte.org/media/hvc/Video-lillehammer-v-narvik-v-no-no-1xbe-26.php

http://live07.colomboserboli.com/tie/v-ideos-orebro-HK-Malmo-Redhawks-v-en-gb-1ysv-.php

http://news24.gruposio.es/ydd/Video-Lillehammer-IK-Narvik-Arctic-Eagles-v-en-gb-1olc-.php

http://live-stream.munich.es/exd/videos-Gruner-Allianseidrettslag-Frisk-Asker-v-en-gb-1fod30122020-.php

that can help your character overcome their flaws or fears? Put them in. Let your character tell you what they don’t want to happen and then make it happen. Show their inner turmoil to finally gather the strength within themselves to overcome the challenge.

GANS have also been ported to specifically make deepfakes, through open-source projects like DeepFaceLab and Wav2Lip. The ease of using these services can’t be overstated: The Wav2Lip project’s website shows how a single line of code can be used to automatically make the subject of a video lip-sync to any audio file.

The creator of the A.I.-generated porn, who posted it on platforms like PornHub and OnlyFans, told Vice that he used StyleGAN2, an open-source algorithm built by Nvidia. If you’ve seen highly realistic fake faces online, like ThisPersonDoesNotExist.com, they’ve likely been generated by StyleGAN2.