Alan: Good friend Frances (my traveling companion for over fifty years) has only one surviving sibling, her sister Mary, a dedicated Facebook user.
To render the following account in meaningful context, I must point out that Frances and Mary both supported Hillary Clinton at the beginning of the 2016 presidential campaign.
Then a flabbergasting thing happened.
There came a day when Mary, evidently disturbed, contacted Frances to say she had just seen a video of Hillary berating a staff member in such an insulting, degrading way that she would never be able to vote for such a person.
My regular readers know that I am a "data hound," that I have the "nose" of adept researcher and am unusually skilled at finding online information.
However... I was unable to find anything bearing remote resemblance to the video Mary swears she saw.
At intervals over the course of a year, I did my level best to find the video that "made Mary want to puke" and could find no trace of it, not a hint of the video that was directly responsible for Hillary losing Mary's vote.
These events -- Mary's testimony concerning the "abusive video" and her subsequent renunciation of Hillary -- took place long before I knew anything about "Deep Fake Videos."
And although the rules of evidence prevent the conclusion that Putin-Russia were responsible for this deep fake dirty trick, we do know that 1.) Putin wanted Trump to win and 2.) Putin's hackers used Facebook extensively to sow dissention and doubt by saturating cyberspace with cleverly-produced falsehood and inflammatory manipulation of popular sentiment.
While I acknowledge that there is no actionable proof that Mary was bamboozled by Russian operatives, "all the pieces fit" - and they fit with a remarkable degree of particularity.
Alan: There is something new under the sun.
Photographic evidence can be assembled pixel by pixel.
Photographic evidence can be assembled pixel by pixel.
And with enough computer memory and sufficient technical sophistication, it is possible to generate a completely persuasive, "foolproof" photograph (or video) of Ronald Reagan "rolling in the hay" with Mother Theresa.
Furthermore, because there is inherent "noise" in the routine compression of digital imagery, it is impossible, even for experts, to distinguish between genuine digital images and faked digital images, particularly faked video.
In effect, the inherent "noise" hides all evidence that "fake videos" have been faked.
The process of creating the "fake" pictographic evidence -- and ensuring that "viewers" are completely "faked out" by the "fake" -- is illustrated in the instructional video embedded in the following webpage.
It is well known that people routinely "believe what they want to believe."
Just reflect -- however briefly -- on the fact that one third of all Americans are as certain that Donald Trump is a Godsend (literally a Godsend) as another third is certain that Trump is this generation's P.T. Barnum, in cynical search of the sucker born every minute.
Notably, the very nature of digital photography and digital videography makes all photographic and videographic "evidence" of such uncertain origin that everyone can justifiably believe their expert-of-choice.
From now on, any debate concerning the authenticity of photograhic and videographic evidence is literally "six of one, half dozen of another."
Forevermore, our determination of "what's true" will depend on the intellectually rigorous use of language, epistemologically sound research and textual documentation.
Forevermore, our determination of "what's true" will NOT on pictographic images which can already be indetectably falsified.
Now gnosh on this.
The incremental improvement of "Deep Fake Video" techniques will only get "better" over time.
Geometrically better.
Unlike photography and videography, face-to-face verbal-and-written determination of truth will never be perfect.
Ironically, in the future our "epistemological salvation" depends on such evident imperfection.
"The perfect is enemy of the good."
Fortunately, just as we can determine the "cloud" in which 98% of an atom's electrons orbit (with no way to determine the location of the other 2%), so too does the "cloud of truth" (like Yahweh's indeterminate shekhinah) lend itself to remarkably precise, albeit imperfect, definition. https://en.wikipedia.org/wiki/Shekhinah
Furthermore, because there is inherent "noise" in the routine compression of digital imagery, it is impossible, even for experts, to distinguish between genuine digital images and faked digital images, particularly faked video.
In effect, the inherent "noise" hides all evidence that "fake videos" have been faked.
The process of creating the "fake" pictographic evidence -- and ensuring that "viewers" are completely "faked out" by the "fake" -- is illustrated in the instructional video embedded in the following webpage.
"Deep Fake" Videos Created By Artificial Intelligence Just Got Even More Terrifying
It is well known that people routinely "believe what they want to believe."
Just reflect -- however briefly -- on the fact that one third of all Americans are as certain that Donald Trump is a Godsend (literally a Godsend) as another third is certain that Trump is this generation's P.T. Barnum, in cynical search of the sucker born every minute.
Notably, the very nature of digital photography and digital videography makes all photographic and videographic "evidence" of such uncertain origin that everyone can justifiably believe their expert-of-choice.
From now on, any debate concerning the authenticity of photograhic and videographic evidence is literally "six of one, half dozen of another."
Forevermore, our determination of "what's true" will depend on the intellectually rigorous use of language, epistemologically sound research and textual documentation.
Forevermore, our determination of "what's true" will NOT on pictographic images which can already be indetectably falsified.
Now gnosh on this.
The incremental improvement of "Deep Fake Video" techniques will only get "better" over time.
Geometrically better.
The Bible's Two Versions Of The Decalogue Are Different. The 1st Commandment Is All About The Confusion Of Crafted Images With Reality
http://paxonbothhouses.blogspot.com/2015/09/the-bibles-two-versions-of-ten.html
"Idolatry: Why The First Commandment Forbids It"
Unlike photography and videography, face-to-face verbal-and-written determination of truth will never be perfect.
Ironically, in the future our "epistemological salvation" depends on such evident imperfection.
"The perfect is enemy of the good."
Fortunately, just as we can determine the "cloud" in which 98% of an atom's electrons orbit (with no way to determine the location of the other 2%), so too does the "cloud of truth" (like Yahweh's indeterminate shekhinah) lend itself to remarkably precise, albeit imperfect, definition. https://en.wikipedia.org/wiki/Shekhinah
"Tracking Down Fake Videos"
NPR
With increasing frequency, photographs and videos are subject to total fabrication.
Furthermore, it's becoming ever harder to detect their falsification.
Indeed, it has become so much harder that even professional technicians will dispute the veracity/falsity of well crafted "images."
Here Come the Fake Videos, Too - The New York Times
Lawmakers Want US Intelligence Assessment On Fake Videos
Voice Of America
Here is what the bible has to say about deep fake video.
In Both Versions, The 1st Commandment Is All About The Confusion Of Crafted Images With Reality
http://paxonbothhouses.blogspot.com/2015/09/the-bibles-two-versions-of-ten.html
"Idolatry: Why The First Commandment Forbids It"
No comments:
Post a Comment