"The severity of the penalties should match the level of risk that disseminating these images poses to our society; i.e., they should be extreme."
That is the top-rated comment — by a lot — at "A.I. Is Getting Better Fast. Can You Tell What’s Real Now?" (NYT)(free access link)(take the test/I got 7 out of 10).
How terribly punitive and repressive, and yet, isn't it what you've come to expect from the segment of America that reads the New York Times?
Notice the aggression mixed with passivity. The comment-writer doesn't want to face the challenge of becoming more perceptive and skeptical dealing with the onslaught of A.I. images. They want the government to do the dirty work and do it good and hard.
The comment gets a response from someone who at least says the words "First Amendment": "the first amendment protects even speech that is a lie. If these images should be prosecuted as fraud, so should Donald Trump's outright lies." There's already selective prosecution of Donald Trump — and civil suits costing him hundreds of millions — and this commenter wants even more of it. And I suspect that the first commenter envisioned their political enemies getting the worst of those "extreme" penalties that "match the level of risk."
Posts
81 comments:
So if I use an A.I. support tool to photoshop a photo of a politician into something ridiculous for my own perception of political gain, I should be prosecuted?
Fascism became mainstreamed by the party who called itself liberal and progressive.
They must feel complete confidence in their side’s ability to control the federal government, otherwise they would never give so much power to something that could turn against them.
Or maybe they’re just thoughtless fools.
9 out 0f 10 correct. You ain't foolin me..........yet!
I got 6 of 10, but a few I just started trying instinct rather than analysis. For instance, I saw the errors in the movie clip, but as a movie clip, it could have been a still made by Hollywood rather than AI. Is an actor standing in front a facade fake or real?
I got 6/10 but I could swear that there are 6 fingers on Gal Gadot's right hand in that first photo.
NYT readers were fooled by the Steele Dossier, but they still read the NYT and demand government have more power. You can’t make that level of cognitive dissonance up.
I got 4 “wrong” because I’m told they weren’t AI but I know they were and the NYT is lying. Everything is AI. Better get used to it.
I’ve mostly exited the vendetta obsessions of online ranting.
Not out of virtue, but out of concern for my own psychological well-being.
This vendetta obsession is not confined to one side. Try being a Trump supporter, like me, who worries that he cannot overcome the administrative failures of his term, and expressing that on a conservative site, like Instapundit. The fury that unleashes is preposterous.
“You’ve got to be just as vicious, corrupt and manipulative as the other side, or you’re really on the other side!” This mentality has totally overtaken the online world. Everybody is goading themselves and others into bitter confrontation and wild hyperbole.
Getting worked up into a frenzy over politics began to seem kind of silly to me, particularly in light of the revelations of the past decade of how phony and rigged the political process is.
How can we be sure this column isn’t an AI fake meant to lull us?
Deep fakes are essentially fake news. If you act in good faith, do independent research and conduct proper fact-checking, they are essentially useless. I would argue it will mostly drive additional polarization (i.e., you believe what you want to believe), but I have a hard time seeing it sway elections. We've entered a post-truth era well before deep-fakes.
The scariest aspect of deep fakes is that politicians will use the technology to easily brush off any incriminating evidence. What would happen if the I-grabbed-her-by-the-P-word scandal had happened today? Trump would surely claim that the recording is fake, and there could actually be truth to it. Will we ever be able to hold our leaders accountable for things they may or may not have said?
6/10 for me.
We see alot of fakery that is not generated by AI and yet buy into it because we want to. Fo instance, friends of mine hate Trump because he was a con man in the past, but they don’t mind the actual con being played on them by our own government, ex. CAGW, Inflation Reduction Act, etc. It’s a hoot, if not so consequential.
Shouting Thomas, I get where you're coming from but I have to disagree that this is a 'both sides' problem.
I followed a Sarah Hoyt link at Instapundit to William Otis's concurrence with a WSJ editorial by Michael McConnell regarding the damage the Bragg prosecution has done to our political and legal systems. At one point they address Bragg's campaign promises to 'get Trump'. Otis gives the standard GOPe answer that the best we can hope for is the people who voted Bragg into the DA's office will now vote him out of office because he kept a campaign promise!
As Ms Althouse's reading of the comments on this article show, I think that is a very thin reed on which to place support for the rule of law when a third or more of our fellow citizens are cheering such prosecutions on. It's an unfortunate situation that more pain is going to be necessary to get that group to understand the problems they are creating but it is hard to see what other solution will work.
Democrats want to punish everyone; Republicans want everyone to prosper.
I used AI to generate that aphorism. Doesn’t mean it’s not true. In fact it’s only too true.
And to help buttress my point at 6:17, Rich at 6:05 illustrates 'post-truth' campaigning by continuing to use the Democrat misquote of the Access Hollywood tape.
Biden's campaign used a fake image of him last week, superimposing an USA flag on his glasses. How far do we want to take this law, Liberals?
"Better get used to it."
Okay, AI.
What would a NYT article/comments be without somehow dragging the word "Trump" into the discussion. It's endemic on the Left.
60%.
Tough test!
Should we jail "journalists" that spread fakenews too? The facts of the Charlottesville speech were easily readily available when it happened, yet multiple outlets spread the blood libel of Trump supporting Nazis without ever verifying the actual facts. Rather than take this nuclear option to political advertising and reporting, I have a modest proposal that a free and open society would embrace.
How about requiring clear labelling of content as one of three things: Fact*, Opinion, AI-generated content. F, O or A.
*Would have to link to the entire original transcript, video, court document, social media post, audio recording and/or eyewitness account unedited, no exceptions. As we recently learned from the Biden admin, editing = fakery. So no editing of source material. If no source material is available unedited, then content is labelled Opinion.
Breezy:
I’ve repeatedly told the Omaha Public Power District Board that CAGW is the biggest scam in the history of the world. I’m not well liked.
How can we be sure this column isn’t an AI fake meant to lull us?
Have you noticed Althouse’s photos get better and better?
Citizens who long for mommy and daddy welfare state to make the world quite a bit less scary. Why? Partly this is the boomers getting old; they were much more libertarian when they were young. Berkeley radicals were called the Free Speech Movement.
Partly the young are fed up with the boomers. They were supposed to eliminate racism and sexism by this time, and they failed. They became bourgeois careerists. If they are reactionary on the new things, gay marriage and trans surgery, this proves they were always reactionary about the most important things. Somehow the success of Trump proves to some people that the underlying reality was always uglier than many people wanted to admit.
The woke want to ignore real problems like crime, keeping the lights on, and yes the family, in order to focus on identity and past injustice. The allegedly innocent are guilty, the guilty are innocent, Big Brother and big corporations can be counted on not to lie to us, as long as Trump is not in charge.
How did Orwell's Big Brother get away with this? It seems there was once a genuine world war, and the Establishment found it convenient to continue an almost entirely phony war. Neo-cons are the running dog lackeys of the woke and the welfare state.
“Will we ever be able to hold our leaders accountable for things they may or may not have said?”
Tells me all I need to know about “Rich.” Wants to “hold our leaders accountable” “for things” “they may or may not have” “said.”
Democrats use money to get power, Republicans use power to get money
Sell the Democrats; buy the Republicans.
”There's already selective prosecution of Donald Trump …”
That’s a remarkable assertion. I’m wondering what might serve as an example of a selective prosecution. Not the Florida documents case, where the distinguishing features meriting prosecution for Trump have been examined and explained ad nauseum.
Not the J6 insurrection case, where Trump was the spiritual leader of a riot that has resulted in hundreds of other plea deals and trial convictions.
Not the Fulton County Georgia case, where Trump is one of 19 defendants and four of them have already pleaded guilty.
The New York cases? Where a criminal jury had to trouble at all reaching a guilty verdict? Where Trump’s own freakishly irresponsible behavior earned him not one but two civil judgments in favor of E. Jean Carroll? The civil fraud cases against the Trump Organization? Where CFO Allen Weisselberg is now serving his second incarceration? Where Trump and his children are all now forbidden from trust work as a result of past fraudulent operations?
What is one good example of a “selective prosecution” brought against Trump?
You want the Truth? Read the Bible’s — both the Jewish and the Christian. It’s all myth but true as hell.
Ass Chuck is AI.
Ass AI.
In readable print at the beginning of the message: AI was used to make this message.
Where is the fraud in that?
LLR-democratical Rich: "What would happen if the I-grabbed-her-by-the-P-word scandal had happened today?"
ANOTHER New Soviet Democratical Hoax "quote" from Rich! Just a couple of days after he was forced to confront the collapse of one of his favorite and desperately clung to democratical lies re:Charlottesville!
What is it Rich? Are you experiencing New Soviet Democratical Hoax Collapse withdrawal and you simply needed to crawl back into the warmth of your other lies?
You ought to get a comfort animal and maybe a little plushy toy to get you through these difficult times.
I got 10/10. Almost all AI images fall into the uncanny valley for me, and I have an almost instinctual revulsion to them. Not out of any moral stance, they just look like nightmare fuel to me.
Its important to recall how LLR-democratical And Violent Homosexual Rage Rape Fantasist Chuck routinely creates fake quotes and assigns them, in mindreader mode, to his political and cultural opponents.
Basically anyone to the right of Adam Schiff and anyone opposed to extreme child groomer activities.
I am thinking specifically of Chuck's racist thought "Ben Carson is only qualified to lead a Department of Black People" which Chuck launched and then tried to assign to Trump. It was quite a spectacle watching Chuck go Full Inga and claim to know precisely what Trump is really thinking and how Trump should be held accountable for Chuck's fake "quote".
And all that projection and lying from Chuck didn't require any AI at all!
Meade, my take is he is definitely artificial but not intelligent: 100% asshole.
Drago that fake “quote” from me is a lie. I never wrote that.
You need to do something about these false, misleading, off topic, cluttering personal attacks, Althouse.
What a world of difference there is - and what a difference in the world there is - between "love your neighbor; judge nothing; condemn nothing; forgive everything" and "judge! condemn! mandate! punish!"
Nothing publishes in the times or thd post xan be assumed to be real or complete acciunt
“ You need to do something about these false, misleading, off topic, cluttering personal attacks, Althouse.”
She doesn’t NEED to do damn thing, Ass Chuck. Self-deport.
7 of ten correct.
Would the readers have issues if, say, somebody falsely accused others of providing fakes that were totally legit?
That seems to be the WH raison d'etre as of late.
As if the "Biden looks like a senile dufus" videos are new or rare.
How terribly punitive and repressive, and yet, isn't it what you've come to expect from the segment of America that reads the New York Times?
Ah, Professor Althouse, with respect, doesn’t the set of all New York Times readers include you?
Back in 2009, before he got a terminal case of TDS and went over to the dark side, Jonah Goldberg published a prescient book titled Liberal Fascism. Somehow here in the 21st century liberalism morphed into Nazism gradually, and now all at once. This year they added “death to the Jews” as the rotten cherry on top of their shit sundae.
LLR-democratical And Violent Homosexual Rage Rape Fantasist Chuck: "Drago that fake “quote” from me is a lie. I never wrote that."
It really quite amusing how often you attempt to deny what you have written in a lame attempt to try and "reset" your fake online persona of a "conservative".
Whether its your vicious, violent, misogynistic and racist attacks on conservative women and blacks/Latinos or your inadvertant admission that you are only here at Althouse to "drive a wedge between Althouse and her readers".
There are hundreds more examples of course that have accumulated like rhetorical barnacles on the bad ship Team New Soviet Democratical....and now you find yourself unable to regain speed in advancing the democratical party interests under false pretenses!
Tsk tsk tsk
Such a sad story really.
And pathetic.
7/10, but I tried to be quick about it. What is my initial impression?
Two of the 3 I got wrong were fairly obvious fakes once I looked at them knowing they were fake (ex., if you actually examine it, The Rock's face looks like a cartoon). The only one that was really indistinguishable was the rustic bathroom, which was pretty innocuous and looked like a standard Architectural Digest bathroom. No harm will come from misidentifying that photo.
For the record, I got 8 right. I thought one was AI because different faces seemed illuminated from different directions, but the Times says it’s real. I suppose expert photographer Althouse could explain how that works. And another seemed to be so obviously AI that I was sure it had to be real. Oops.
LLR-democratical And Violent Homosexual Rage Rape Fantasist Chuck: "You need to do something about these false, misleading, off topic, cluttering personal attacks, Althouse."
I would be hard pressed to come up with a better example than the above Chuck psycho comment and demand of democraticals projecting at Althouse blog.
"Abacus Boy" Rich clearly tries his best to meet his ally Chuck's "standards" in that regard but, lets face it: when it comes to purposefully mendacious posting with bad motives and ill intent, Chuck is in "the fast lane", so to speak.
I looked but gave no answers. Why help the New Tricknologists?
What that top commenter really wants is for the prosecution of these things to be selectively applied. I guarantee you the exact same commenter would be laughing his ass off and actively retweeting a fake video that made a conservative look bad.
This is why we have the 1st amendment- we don't have to worry about selective censoring or, at least, we didn't use to have to worry about it.
"How terribly punitive and repressive, and yet, isn't it what you've come to expect from the segment of America that reads the New York Times?"
Nah. We know you just read it so we don't have to.
Still images are almost impossible now for me the to reliably detect as real or AI generated unless I spend a lot of time examining them- like several minutes. I spent several minutes with each of the 10 images and got all but the rustic bathroom one right. I don't have much problems spotting real vs AI generated in videos yet but is getting harder to do with every passing month. 5 years from now you won't be able to tell the difference at all without examining the actual code embedded in any video/image file. It will be a brave new world of information- you had better get used to it being that way.
Fine, Chuck- prove Drago was lying. That is your usual standard applied to Trump.
"Passing AI images off as real ones for the sake of commercial or political gain should be prosecuted as fraud."
"The severity of the penalties should match the level of risk that disseminating these images poses to our society; i.e., they should be extreme."
...but only in those cases that we, your betters, don't like.
They demand harsh, punitive punishments for vaguely-defined transgressions. I am sure that could never go wrong.
At the same time, a political advertisement for Biden I am seeing 200 times per day has 11 video cuts in just over 30 seconds.
Isn't that also potentially misleading for "political advantage"? But left untouched by this modest proposal.
Oh, and as an aside, I have to *love* the defense of selective prosecution by pointing to the Jan6 prosecutions - also obvious examples of egregious selective prosecution.
Its not turtles - its lawfair all the way down now.
Lloyd W. Robertson said...
"...The woke want to ignore real problems like crime, keeping the lights on, and yes the family, in order to focus on identity and past injustice. The allegedly innocent are guilty, the guilty are innocent, Big Brother and big corporations can be counted on not to lie to us, as long as Trump is not in charge.
How did Orwell's Big Brother get away with this?"
A generation indoctrinated.
And it's not only merely indoctrinated, it's really, thoroughly, most sincerely indoctrinated.
"At the same time, a political advertisement for Biden I am seeing 200 times per day has 11 video cuts in just over 30 seconds.
Isn't that also potentially misleading for "political advantage"? But left untouched by this modest proposal."
Exactly right- editing any video to make the subject look worse or better isn't qualitatively different than using AI to generate the exact same file with the same content.
Meade said...
“ You need to do something about these false, misleading, off topic, cluttering personal attacks, Althouse.”
She doesn’t NEED to do damn thing, Ass Chuck. Self-deport.
6/25/24, 8:29 AM
Let’s review, Laurence.
1. I posted a single comment that went directly to the text of Althouse’s main post.
2. Drago attacked me personally.
3. The attack was a pure falsehood.
4. The attack was completely off-topic.
5. We now see Drago obsessively doing repetitive attacks.
”Comments may need to pass through moderation. Comments should respond to material raised in the post. I encourage brevity and substance and discourage personal attacks and repetition. You must use a name or pseudonym. The non-name "unknown" is not accepted.”
And you think that I am the problem?
Here is an image of Chuck: is it real or is it AI generated?
You are a problem, Chuck. It is hilarious that you don't see it. You create more Trump voters than you discourage with your antics and I laugh every single time you appear to believe otherwise. If I were a member of the Trump Campaign, I would create an army of fake commenters just like you to discredit Never-Trumpers.
Start with a digital "water mark". Don't ban.
Always knew that Drago was a beta version ... explains a lot of the malfunctions.
I didn't examine any closely, and still got 9 of 10... and the one I missed was the bathroom where if I'd looked closely I'd have seen the weird text on the banner on the wall. AI pics still seem to have a "look" to them that I seem to be able to pick up on pretty easily anymore, though I have noticed over the past year that that "look" seems to be fading out and it is getting harder to tell.
AI needs a watermark or one day we'll see Biden playing the intro to Layla on CNN.
It seems relatively trivial to have any image or video made with commercially available AI to leave some code in the file that is detectable by any browser.
Hover your cursor over the image/video and a note pops up saying when and how it was made.
"But wait," you say, "what about people developing their own, homegrown AI?"
Good luck with that.
I was one of the first users of Photoshop. I met one of the guys who developed it.
I've known since the '80s that all images are suspect...
Maybe they are all AI (sic). The article is designed to deliberately decrease the ability of a person to detect the cylon...
"Partly the young are fed up with the boomers. They were supposed to eliminate racism and sexism by this time, and they failed."
Boo fricken hoo. I'm a boomer and I was never informed that I was responsible for doing those things. If today's yutes want to tackle the job, nothing's stopping them.
"How terribly punitive and repressive, and yet, isn't it what you've come to expect from the segment of America that reads the New York Times?"
Don't be so hard on yourself.
LLR-democratical Rich: "Always knew that Drago was a beta version ... explains a lot of the malfunctions."
Interesting. Its clear you want to help your co-buffoon and serial liar Chuck, you just dont seem to know how.
Would you like to try again?
LLR-democratical And Violent Homosexual Rage Rape Fantasist Chuck: "Let’s review, Laurence."
Note the condescending, patronizing and imperious tone and tactic of our democratical laddie Chuck in how he speaks to the blog moderator whilst lecturing the blog moderators on their performance....even though Chuck himself, after years of psychotic postings, was actually banned but refuses to accept that!
Chuck is so bizarre and weird and apparently borderline everything I wonder if AI could truly mimic him?
I would first have to identify the proper descriptors for Chuck beyond the expected:
- far left democrat
- prefers interacting with others under a false flag
- serial liar and smear merchant
- prone to violent outbursts when cornered with his previous comments
- psychotic need to control all communications within a group for which he has no authority or claim
- exhibits quite a few explicit DSM defined maladies, schizophrenic cringey "Eddie Haskell-ish" interaction with rapid shifts to articulated desires for physical violence, etc. So many others of course.
Whew! Seems like it would take some time!
It’s an article of faith for liberals that “Trump lies.” However in my experience, when you press for examples all you get are examples of Trump engaging in harmless puffery or equally harmless hyperbole, except for the instances where where Trump is demonstrably correct but the person calling out “Trump’s lies” ardently believes something that is actually false.
all images are suspect
Whether through massively parallel correlations, or through human intelligence, nothing has changed.
I got 10 out of 10 right.
I'm usually pretty good at spotting AI images, but as AI gets better, I'm worried that we won't be able to spot AI images in the future.
I think that AI should be treated like regular images in terms of the law. After all, photos have been doctored and used for fraud since the invention of photography. There should only be a punishment if there is an intent to defraud or commit some other illegal activity using AI images.
I do worry about AI because I see more downside in it than upside. I think that it will do more harm than good. As the old saying goes, just because you can do something doesn't mean that you should do something.
loudogblog: "As the old saying goes, just because you can do something doesn't mean that you should do something."
Unfortunately, the other relevant old-ish saying is "you can't put the toothpaste (or is it Genie?) back in the bottle".
We are off and running into some brave new world.
Lilly, a dog said...
I got 6/10 but I could swear that there are 6 fingers on Gal Gadot's right hand in that first photo.
*******************
If she does, she can say that when she shoplifts she's getting a "six-finger discount."
baddabish
Chuck said...
Drago that fake “quote” from me is a lie. I never wrote that.
You need to do something about these false, misleading, off topic, cluttering personal attacks, Althouse.
*****************
Translation: "Mommy, make that bad man stop being mean to me!"
There's an image of me going around with me in a speedo with bulging muscles and a six-pack stomach.
Don't be fooled.
It is totally real.
LLR-democratial And Violent Homosexual Rage Rape Fantasist Chuck: "Drago that fake “quote” from me is a lie. I never wrote that.
You need to do something about these false, misleading, off topic, cluttering personal attacks, Althouse."
effinayright: "Translation: "Mommy, make that bad man stop being mean to me!"
A more accurate translation might be "Mommy, make that bad man stop accurately quoting me!"
I can only imagine the insane spittle-flecked direct messages Chuck must be sending to harass Althouse and Meade right about now...and those are not going to be AI generated.
Once the public realizes any and every thing they see could be artificially generated, there should be no more need for big msm media. It's all been just entertainment our whole lives. The truth ain't out there.
There's an image of me going around with me in a speedo with bulging muscles and a six-pack stomach.
@Joe Smith, me too. And my image is also real, not AI. Trouble is, it’s from 60 years ago, when I swam butterfly in college. Once upon a time I had muscles, a narrow waist, and hair. [sigh]
Post a Comment