Discover and read the best of Twitter Threads about #Deepfakes

Most recents (24)

Hallo ich bin Anna Prianka Schmidt @anna_prianka. Diese Woche übernehme ich den Account des SPPs und freue mich darauf, einen Einblick in meine Arbeit zu geben! 🔎 #takeover #scicomm #dasdigitalebild /1
Ich arbeite mit @evisualyn in dem Projekt “Hinter dem digitalen Bild. Fotografien auf Community-Plattformen und auf Twitter als Repositorien für maschinelles Lernen und journalistische Publikationen” /2
bit.ly/3qFXi6a
Was ich dort mache? Tatsächlich bin ich viel hier unterwegs, denn ich suche nach Konversationen zwischen Fotograf*innen (amateur oder profi) und Agenturen, die diese Bilder nutzen wollen /3 #photography #photographers #datacollection #search
Read 6 tweets
Fastest EU-membership negotiations ever, even faster than @Eurosurveillanc's rapid response.

Congrats @vonderleyen, @ProfKlausSchwab, @georgesoros.
Ah, btw. Uncle @ProfKlausSchwab, why did the @KremlinRussia_E hide this link?

It's just 4 months ago. Memory-holed?

Asking for my mommy.

web.archive.org/web/2021122421… Image
Ah, btw. @ProfKlausSchwab, why did you hide Putin's @wef-profile?

Why would you do that? Klaus?

I had to extra-retrieve it from Archive.is. That's not user-friendly.

Deleted:
weforum.org/people/vladimi…
Archive:
archive.fo/s7s9L ImageImage
Read 74 tweets
Have you been seeing artwork like this on your timeline and wondered how it was created?

Let's learn about one of the most popular algorithms for AI-generated art! ⬇ ⬇ ⬇
The technique used is known as "VQGAN+CLIP" and is actually a combination of two deep learning models/algorithms both released earlier this year 2/11
VQGAN - Vector-Quantized Generative Adversarial Network

VQGAN is a type of GAN, which is a class of *generative* neural networks that have been used for #deepfakes and other AI-generated art techniques

You pass a vector/code and VQGAN generates an image. 3/11
Read 12 tweets
😱Video of New Zealand PM 'smoking crack'😱

#Real or #fake? Let’s find out. 👇👇👇 1/...

#OSINT 🔎
#Verification 📹
#SpeurJeMee? 🧐 #HowToOSINT Image
On 1 Oct. someone posts a video on #Twitter.

We see two videos superimposed. According to the caption , we would see the New Zealand Prime Minister Jacinda Ardern twice.

At the top she appears to be talking to #BillGates and at the bottom she appears to be using #crack. 🔞2/. Image
We also find the same video in various other posts on #Twitter, #YouTube, #Facebook and on the alternative video site #BitChute.

In total, these videos were viewed more than 31,000 times. Let's verify! 3/...
Read 34 tweets
Today the European Commission officially presented its proposal for harmonised rules on artificial intelligence that has the ambition to create a framework for #artificialintelligence that respects European values. Is it fit for purpose? At first sight: No. [Thread] 1/23
First of all, very few applications are prohibited. Among them, the ban of remote biometric identification is in. In theory. In practice the exceptions granted to law enforcement agencies are too broad to call this provision a ban. #reclaimyourface 2/23
There is no ban for gender identification and recognition of sexual orientation either. Subject to a ban are subliminal techniques beyond a person’s consciousness, but only if they cause physical or psychological harm to that person. 3/23
Read 23 tweets
Daily Bookmarks to GAVNet 03/09/2021 greeneracresvaluenetwork.wordpress.com/2021/03/09/dai…
Scientists find evidence of mathematical structures in classic books | James Joyce | The Guardian

theguardian.com/books/2016/jan…

#fractal #joyce #evidence #books #structures
Read 10 tweets
Threats to share intimate images can have life-threatening consequences and should be criminalised:
Baroness Nicky Morgan: Change revenge porn laws so we 'stop failing women and girls' telegraph.co.uk/news/2021/01/0…
My research with colleagues has also found that almost 1 in 5 people (19%) across the UK, New Zealand and Australia have experienced threats to share intimate images #LifeThreatening
routledge.com/Image-based-se…
Reminder that this is not a new issue and there have been campaigns for years to get a change in the law. Here is just one example of where it could have made a difference:
telegraph.co.uk/news/2019/03/1…
Read 4 tweets
#DeepFake Tutorial.

In this thread I'm going to post some tips & tricks to identify Deep Fakes using these examples I found online.

📊Vote in this other thread first:
There are many ways to detect deep fakes. Here are three of them:
1) Impossible static poses.
2) Impossible movements.
3) Technology artefacts.

The first category seems to be easier to detect, but the second is more reliable. The third may go away soon!
VIDEO #1

This one seems easy enough...

TIP: Use the speed controls in your favorite video player to slow it down while it's playing.
Read 19 tweets
Daily Bookmarks to GAVNet 11/23/2020 greeneracresvaluenetwork.wordpress.com/2020/11/23/dai…
Balancing Epistemic Humility and Prior Knowledge - Insight

zeynep.substack.com/p/balancing-ep…

#humility #balancing #knowledge #coronavirus #prior
Maybe There is No Technological Slowdown - New Things Under the Sun

mattsclancy.substack.com/p/maybe-there-…

#growth #technology #decline
Read 8 tweets
1) I think this is #DictatorDan's actual testimony. Still weird as, but. He looks skinnier, paler, way younger. And his hands look large in comparison to his body. I suspect tech heads could explain weirdness as result of Skype or similar software with low resolution.
2) But it's not remotely batshit to wonder what is going on here. Look at insane stuff going in #Victoria lately. To say it seems surreal (zif dystopian satire) is yuge understatement. Also, many #MSM glitches involving pollies esp those linked to Cabal, Dems (as Dan is).
3) Check out this "glitch in the matrix" featuring #HillaryClinton.
Read 42 tweets
[🚨New Book🚨] Today is the BIG DAY! With the help of so many amazing people, I'm SO excited to launch my first book, THE HYPE MACHINE, today!

sinanaral.io/books

For those of you who like to sample before you buy, here is a sneak peak at the book in 12 acts 1/
[Chapter 1] opens with the annexation of Crimea as a leading example of the geopolitical impact of the New Social Age. It provides new evidence on the role of social media in the Crimean annexation - the first forceful redrawing of European borders since WWII. 2/
[Chapter 2] traces the Rise of Fake News all the way to the End of Reality, covering its impact on elections & democracy, business & markets and our public health like #Covid_19. It covers the science of fake news and the implications of the rise cheap, ubiquitous #deepfakes 3/
Read 14 tweets
STARTING NOW (11 AM Pacific) Deepfakes and the U.S. Election hosted by the University of Washington’s Center for an Informed Public in collaboration with Microsoft’s Defending Democracy Program.

aka.ms/deepfakeforum
This live event is open to the public. Image
Learn more about @UW @UW_iSchool Center for an Informed Public
cip.uw.edu
History of video deep fakes begins with porn (as so much has done) - about 2.5 years old.
@NinaDSchick
We have an implicit trust of video - as a society we have been "extremely slow" to respond to this threat.
Read 18 tweets
LUMS 1, Digital Deepfakes 0
SSE Researchers Part of the Winning Team of the @Facebook Research Award 2020

Have you ever used filters on your @Snapchat? Of course, you have! We love filters that can make us look older, younger, or just outright funny.
1.10
However, a sinister extension to this technology, appropriately named ‘deepfake’, can recreate anyone’s voice and video with unprecedented accuracy and can be used to project fake identities of influential people to create anarchy, distrust and political polarisation.
2.10
Bad, bad stuff!

But worry not – Team LUMS to the rescue!
3.
#LUMS #LifeAtSBASSE @lifeatlums #deepfake @Facebook
Read 10 tweets
RECOMENDACIÓN: Se los explico al detalle, en su mayoría sabemos que al descargar APPs otorgamos datos, en otras ocasiones ingreso a nuestro Movil o geo localización, asunto que es delicado pero que es medible el riesgo a corto y largo plazo, Image
todos en facebook hemos entregado datos y un centenar de información a cambio de un servicio, la diferencia al Usar FaceApp no sólo es que regalas tus datos personales, sino que se te hace una toma de rostro biometrica en 180 grados
(que no es lo mismo que usar una de las likes de fotografías que haz subido a la web), y autorizas el uso de tu rostro para ejercicios de identidad digital (machine learning + I.A) que se queda grabado en el universo de los 1 y los 0,
Read 7 tweets
#Deepfakes are a fundamental threat to democracy, especially as it pertains to women. If you want to be safe from finding your face manipulated onto violent #pornography you cannot have it anywhere online. This threatens womens participation in public life
There has always been violently sexualized ridicule of women in public life, but for the longest time it was handdrawn. Now with sufficient footage you can manipulate any woman's face onto that of a woman in porn - take personal revenge on an ex or humiliate a female politician.
Feminists have warned society about the impact of porn and its ripple effects on women in and out of the industry for decades, but we weren't heard. Because it was physically primarily happening to a woman in the industry and when reenacted on one outside excuses were made.
Read 10 tweets
1. A series of tweets on my interview with @hugoreasoning about his new book which suggests that people are far less gullible than we think. Interview at washingtonpost.com/politics/2020/… and book at amzn.to/3c5NZTN (nb that Hugo is a friend, and someone who I am writing with).
2. The argument of the book is straightforward - that human beings are far less likely to believe crazy things than most commentary presumes. Hugo is a cognitive psychologist, and draws on work there and across the social sciences to make what seems to me to be a compelling case.
3. His basic claim is straightforward - that if we take cognitive psychology and evolution seriously, it would be bizarre to expect that people would be consistently gullible. We have a variety of mental tools that allow us to assess the likely validity of claims made to us.
Read 21 tweets
Google's Jigsaw unveils new tools to detect manipulated images and deepfakes. @daveyalba

As the tech to create #deepfakes becomes increasingly plug-and-play, accessible to the average citizen, it's essential for detection to keep up.
nytimes.com/2020/02/04/tec…
The core tool is a collaborative research platform, Assembler, that brings in work from academic partners to detect manipulation.

"Fact-checkers and journalists need to understand if an image is manipulated and how much trust to put on that conclusion."
jigsaw.google.com/assembler/#app…
As emerging technologies put more sophisticated tools in the hands of purveyors of disinformation, it becomes more challenging to weed out fact from fiction -- but it's something our democracy depends on.

latimes.com/politics/story…
Read 4 tweets
Oh for the love of Cthulhu, I had plans for this morning, but apparently we're doing this now.

Gather round children, for a thread on deeply irresponsible "activism".
So, these #deepfakes are made by @FutureAdvocacy, a "think tank" aiming to "promote responsible AI policy".

You can watch them in the thread from the quoted tweet above. And they are in and of themselves a danger to democracy.
Who am I to be telling you this? I'm the person who's watched a substantial chunk of the existing porn #deepfakes on the internet so you don't have to. (Paper forthcoming in Porn Studies, like next month.)
Read 18 tweets
The @StanfordHAI (Human-centered #AI) Conference on #AiEthics, Policy, & Governance is getting started, live stream below. Speaker list is fantastic. Will be tweeting.

#machinelearning #data4good
@StanfordHAI "Silicon Valley, we have a problem" opens @erikbryn

Speaks to the 'great decoupling' - where productivity and GDP expand with technology, but don't take household incomes with them.

Part of the problem is lower-paid occupations have more automate-able tasks.
@StanfordHAI @erikbryn Both @Susan_Athey and @erikbryn speaking to ML adoption in private sector and the important distinction between labor replacing vs. labor augmenting #automation.

The "Suitability for Machine Learning" or SML scores from this
@MIT_IDE is useful here: ide.mit.edu/sites/default/… #ai
Read 22 tweets
HAPPENING NOW: Day ✌🏾 of #MediaLitWk is kicking off at .@thomsonreuters with remarks from Editor-in-Chief Steve Adler and .@ciullalipkin ImageImage
“Being media literate empowers us to be an engaged participant in the media-saturated world.” #MediaLitWk
A panel of journalists from .@Reuters .@CJR and .@VICENews are dropping gems on media literacy and how it relates to the business and content of media in today’s world! #MediaLitWk ImageImage
Read 9 tweets
Here's a funny story from the making of #TheReport, told in honor of the work that @BobbyChesney and @daniellecitron have done on #DeepFakes. As we were beginning work on Volume II of The Report, our executive producer, Ian Enright of @GoatRodeoDC, brought to our attention...
A Twitter feed called @StableVoices, which is well, a #DeepFake that imitates the voice of President Trump. We began fantasizing about using for all of Trump's lines in Volume II a character we dubbed #DeepFakeTrump, which would be the lines as synthesized by @StableVoices.
So Ian got in touch and asked @StableVoices whether #DeepFakeTrump was possible. Ian sent a key line to see how well it might work. Here's what @StableVoices sent back in response:
drive.google.com/file/d/1ku0tr4…
Read 7 tweets

Related hashtags

Did Thread Reader help you today?

Support us! We are indie developers!


This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3.00/month or $30.00/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!