ADDED: It's funny to see "sewing discord" for "sowing discord"! If you were sewing discord, you'd be mending it, not scattering it about.
October 13, 2023
"Rapid advances in artificial intelligence have made it easy to generate believable audio, allowing anyone from foreign actors to music fans to copy somebody’s voice..."
"... leading to a flood of faked content on the web, sewing [sic] discord, confusion and anger.
Last week, the actor Tom Hanks warned his social media followers that bad actors used his voice to falsely imitate him hawking dental plans. Over the summer, TikTok accounts used AI narrators to display fake news reports that erroneously linked former president Barack Obama to the death of his personal chef.
On Thursday, a bipartisan group of senators announced a draft bill, called the No Fakes Act, that would penalize people for producing or distributing an AI-generated replica of someone in an audiovisual or voice recording without their consent...."
Subscribe to:
Post Comments (Atom)
Posts
43 comments:
If only Disney had used the technology to de-age Harrison Ford's voice in the latest Indy movie. They de-aged his face, sort of.
Call me skeptical that it is believable.
And this is one reason why many online accounts suddenly shifted to mandatory two-factor authentication in the last year. The only way they can avoid fraud (voices, photos) is to link to an established phone or email account.
All sorts of search engine results are fraudulent today, with pop-up websites stealing source material in hopes of capturing credit card numbers and personal information.
Welcome to the early Blade Runner phase of house-of-lies society. I await an Amish-like or Luddite-like backlash -- in times like these the old ways may be the safe ways.
Seems like we've been getting these warnings for almost a decade now. I've always suspected they're being issued because something is about to drop that someone wants to pre-emptively discredit.
George Orwell would have a field day with all this.
Over the summer, TikTok accounts used AI narrators to display fake news reports that erroneously linked former president Barack Obama to the death of his personal chef.
I never paid attention to the chef death at the time but now that WaPo saw fit to stuff two truckloads of duping language into one brief sentence in a story about AI I’m convinced Obama’s handlers had his personal chef killed because the chef knew too much…
I’m sewin’ dis cord!
🎶if I had a sowing machine,
I’d sow in the morning
I’d sow in the evening
All over this town
former president Barack Obama to the death of his personal chef
Barack Obama IS linked to the death of his chef. Because it happened at Barack Obama's house.
That's the link.
If somebody mysteriously dies at your house and there's no investigation, you are now linked to that death. We'll be watching to see how many OTHER mysterious deaths you become linked to.
The US news media spent 4 years reporting on Russia Collusion. The New York Times even got a Pulitzer Prize for their breathless coverage. And despite the fact that it was all faked and completely not true, they refuse to return the Pulitzer.
Literally NOTHING you read in our news media can be believed. And that's exactly the way our government wants it.
There's so many lies and liars now, I really wonder if this is necessary.
Blogger R C Belaire said...
"George Orwell would have a field day with all this."
Indeed, but, somehow, they're still calling it "progress"
I always pause before typing "sowing discord" or "reap what you sow". That "ow" stops me, and I have to convince myself that the spelling is proper. Even right now it looks weird.
They're sewers, all right.
Ann said…
“It's funny to see ‘sewing discord’ for ‘sowing discord’!”
They should of been more careful with their words. It shouldn’t phase me, but it does.
Too bad nobody has thought to apply A.I. to grammar and spelling checkers.
Watch The Capture, a BBC thriller streaming free on Peacock, for more on the subject.
Series 1 was great. Series 2 even better. But watch Series 1 first.
You’ll thank me later.
This is just another point of the case that things that are the real deal should gain dramatically in desirability. Classics? Too well-known to be faked, The Iliad and the Odyssey seem likely to regain popularity in coming generations with their classical insights into humans' minds and souls. Hard-earned and real, not shallow and ersatz like the drivel from the AI mumblecore Althouse occasionally posts on this site. People will crave the real. Their authority may be forgotten or ignored for periods, but it is impossible to deny.
That consent issue? It will be included in the Microsoft and Google and Apple Terms and Facebook Terms of Service, right under the bit about all content generated belonging to the corporation and usable for any purpose, which is already included.
Can comedians still do impersonations? Politicians (and their 'lawyers') are notoriously bad at writing laws.
This is also a leftist/totalitarian desire, of course, because it affects them.
Fix the damned laws so that criminals who steal peoples identity, wrecking their lives, are actually caught and prosecuted. Until then, fuck off.
I'm laughing at the sew/sow error. I wonder why the Post doesn't employ copy editors. They're sorely needed.
At least we know, a human being wrote the article since they made that egregious mistake
Not really. But it does make me feel better and almost nostalgic about stupid mistakes. In my experience AI does make a lot of mistakes.
If you were sewing discord, you'd be mending it, not scattering it about.
Damned Autocorrect.
"At least we know, a human being wrote the article since they made that egregious mistake"
"ChatGTP, write an article in about deep fakes in the style of a typical WaPo journalist."
Eric Swalwell on the AOC fart video.
Scroll up to see the original AOC fart video @0:38.
Whatsoever a man ripth, that shall he also sew.
In my experience AI does make a lot of mistakes.
Mine too, but there is a bit of irony with the statement. The mistakes AI makes is due to flaws in its programming made by humans.
Too bad nobody has thought to apply A.I. to grammar and spelling checkers.
I take this as snark. I would prefer AI got removed from automatically trying to correct my grammar and spelling when I attempt to post comments with a mobile device. And going back and trying to edit out the AI mistakes real time causes more gibberish from the AI.
There are some hysterical ones of a cartoon Tucker Carlson commenting on current events in the Roman Empire.
If you were sewing discord, you'd be mending it, not scattering it about.
Damned Autocorrect.
That’s right! No one should be allowed to generate a fake voice except the government! And they will only do it for the good of us all.
Maybe AI generated voices should be limited to comedy.
Sewing low sweet chariot.
Part of the problem is the sci-fi name that has been applied, both by the vendors and the media. None of the large basket of tech that now calls itself AI is actually Intelligent: it's mostly statistical pattern matching, no logic involved.
Another problem is the size of that basket. Things that were previously called Machine Learning (a very powerful technique in itself) have now decided that they are actually AI, because it raises valuations. This is forgiveable as they're at least as powerful as generative "AI" but we also get processes that are almost certainly simple rules based heuristics being branded as AI (mortgage applications etc).
I'm quite sure, for example, that a LLM AI is not tuning the performance of jet engines, because that wouldn't be possible. It is a category error, like asking your dog to write a biography of Alexander the Great.
The Obama chef story died pretty quickly (no pun intended)...
At least we know, a human being wrote the article since they made that egregious mistake?
If a human wrote the article, it means his mistakes are his bread and butter, and as we all know, when one makes Campbell alphabet soup mistakes, it can be said one is made to eat one's words, which eventually wind-up in... the sewer.
Kate there’s a mnemonic for that. Sow rhymes with show but sew rhymes with snow.
I've already heard an ad on YouTube that perfectly mimics Jordan Peterson's distinctive voice.
The ad doesn't claim his endorsement, but that's exactly what they are implying.
Celebs often sue for "unauthorized appropriation of image" when their pusses appear in ads.
Peterson might offer that as an analogy for unauthroized use of his voice.
"I would prefer AI got removed from automatically trying to correct my grammar and spelling when I attempt to post comments with a mobile device."
You can turn autocorrect off. I did, because it was making me look stupid.
I grew up watching and hearing Rich Little impersonate everyone from Howard Cosell to Richard Milhouse Nixon, making their voices say things just close enough to reality to make us all laugh. There is nothing new under the sun.
"ChatGTP, write an article in about deep fakes in the style of a typical WaPo journalist."
Yada, yada, yada. Deep fakes may be a problem. Yada, yada, yada. Thump is a Nazi. Yada, yada, yada.
mikee said...
I grew up watching and hearing Rich Little impersonate everyone from Howard Cosell to Richard Milhouse Nixon, making their voices say things just close enough to reality to make us all laugh. There is nothing new under the sun.
*************
Humans impersonating people for parody or satire is exactly the same as machines impersonating people for commercial gain..
Nothing to see here---snort.
What is the relationship between algorithm-driven information dissemination and the operation of First Amendment-based free speech? Does algorithm-driven information have free speech rights?
What is the role of foreign-controlled algorithms on domestic US information dissemination? Do American-owned algorithms have some intrinsic superiority or privilege over foreign-owned algorithms?
Why Fox and not Beijing? How do you tell the difference?
Post a Comment