Perhaps a writing prompt.
From CNN Technology:
Platforms like Facebook, Twitter and YouTube have been moving recently to deal with the threats posed by misinformation and meddling that they didn’t see coming, but now they face an emerging form of disinformation they know is on the horizon: Deepfakes — doctored videos that will eventually fool even the sharpest eyes. As the technology to create deepfakes advances, experts say, it won’t be long before it’s used to foment discord or even affect an election.
“The opportunity for malicious liars is going to grow by leaps and bounds,” said Bobby Chesney, professor and associate dean of the University of Texas School of Law who has been closely researching deepfakes.
. . . .
[A]sked by CNNMoney just what they’re doing to prepare for this looming problem, none of the major social media platforms would discuss it in detail. The companies wouldn’t name the researchers they’re working with, how much money they’ll pour into detection, or even say how many people they’ve assigned to figure it out.
None of them offered much more than vague explanations along the lines of Facebook’s promise to “protect the community from real-world harm.”
. . . .
The word “deepfakes” refers to using deep learning, a type of machine learning, to add anyone’s face and voice to video. It has been mostly found on the internet’s dark corners, where some people have used it to insert ex-girlfriends and celebrities into pornography. But BuzzFeed provided a glimpse of a possible future in April when it created a video that supposedly showed Obama mocking Trump, but in reality, Obama’s face was superimposed onto footage of Hollywood filmmaker Jordan Peele using deepfake technology.
Deepfakes could pose a greater threat than the fake news and Photoshopped memes that littered the 2016 presidential election because they can be hard to spot and because people are — for now — inclined to believe that video is real. But it’s not just about individual videos that will spread misinformation: it’s also the possibility that videos like these will convince people that they simply can’t trust anything they read, hear or see unless it supports the opinions they already hold.
Experts say fake videos that will be all but impossible to identify as such are as little as 12 months away.
. . . .
Companies like Facebook and Twitter often argue that they are platforms, not publishers, and note that Section 230 of the 1996 Communications Decency Act absolves them of responsibility for content posted by users.
. . . .
Satya Venneti of Carnegie Mellon University has seen some success identifying fakes by analyzing the pulses of people in deepfake demonstration videos. People typically exhibit similar blood flow in their forehead, cheeks, and neck. But she found “widely varying heart rate signals” in spoofed videos, something that happened when a video was layered with images.
. . . .
In some cases, she saw heart rates of 57 to 60 beats per minute in the cheeks and 108 in the forehead. “We don’t expect to see such wide differences,” she said.
Siwei Lyu, director of Computer Vision and Machine Learning Lab at University at Albany SUNY, outlined another trick in a paper he co-wrote in June: Look for regular blinks. “If a video is 30 seconds and you never see the person blink, that is suspicious,” he told CNNMoney.
After his paper was released, Lyu said deepfake developers used his research to successfully improve their own work to get around his detection system. Now, his team is exploring other ways of identifying fakes, but he declined to elaborate because he doesn’t want to give away anything that might help people create more convincing fakes. “We’re on the front lines,” he said, adding that Google has expressed interest in collaborating with him.
Link to the rest at CNN Technology
So that photo of Mars orbiting the Earth like a second moon is fake? Dang it all.
For sci-fi world-building purposes I’ve been wondering how it would be if one of the super earths Kepler keeps finding had a moon the size of Mars — and was in the Goldilocks zone. Those pictures swapping Mars for the moon seem designed to taunt me 😉
Would make an interesting ‘world’. Would have to be a lot further out than our moon or the ‘tides’ would tears the planet up (near constant earthquakes!) and if that far out would it have as big a disk as the moon as seen from the earth?
(never mind that large a gravity well so close to Earth and us trying to keep a station in orbit … 😉 )
Why stop at just one habitable moon? Go for two or five!
Any system with more than one habitable planet would be a prime target for colonization…
…and spectacularly bloody interplanetary wars a thousand years later. 😀
Jerry Pournelle’s CoDominium made that mistake. The system was featured in THE MOTE IN GODS EYE.
I’ve read “Mote,” and I’ve been meaning to look into the rest of the CoDominium series. When I checked a few years ago some of the books weren’t available, so I was hoping that when Pournelle died it would light a fire under Baen or someone to put them all out in e-book form. I would have cheerfully bought a tribute bundle at the very least. You’ve reminded me to check again.
They do have a CoDominium bundle. I got it a couple years back:
https://www.baen.com/codominium-future-history-bundle.html
Individual titles, too. But the bundle is a really good deal.
King David’s Spaceship is particularly good.
Btw, MOTE has two sequels: THE GRIPING HAND by NIVEN & POURNELLE and OUTIES by Jennifer R. Pournelle.
The latter was Indie pubbed because Jerry thought the tradpub deals they offered his daughter were egregiously bad.
Ah! Thanks for that! Investigating now…
Heck, make your own! 😉
https://www.daz3d.com/solar-system-super-bundle
“… a fiery, glowing sun and all the remaining planets, from Mercury to Neptune and everything in between – including all the moons found orbiting each planet …”
And you can even land your very own space shuttle on Mars!
https://www.daz3d.com/space-shuttle
Sometimes, visiting the Passive Voice is really bad for one’s budget discipline 🙂
Depends on how badly you want the ‘shiny’, they (and others) have free stuff and sales (and you can even ‘create’ your own toys – though it can take a lot of time/practice. (Yes, I hunt and more often than not ‘buy’ … 😉 ))
“…anyone with a brain will not automatically accept anything that supports their ideology.”
Hypothetically, if we were to restrict voting rights to only those people ‘high minded’ enough to notice their own biases and question themselves as much as they do others, we’d have a smaller electorate than ancient Greece.
Few humans will willingly forsake their ideological comfort blanket. In the future people will either choose to believe what they are told or shown, or they will ignore all newsfeeds regardless of source and cut themselves off (I have already met such people). In either case, how does anyone know what they’re voting for? Elections will become (more of a) farce and the ruling elites will rule an ignorant and disenfranchised population… again, because that’s how it’s been throughout most of history. Democracy as we know it is basically doomed.
That is how populist systems operate: an oligarchy of insiders manipulating the masses and buying votes via “government services”. Once the oligarchy institutionalizes their nominal ideology the elections can be as open as anybody might want to no significant effect. Argentina is a perfect example: nominal leadership changes but the core policies don’t. It might as well be a single party state.
Three easy steps:
1- Party loyalty, which is a standard feature of parliamentary systems and personality cults
2- Delegitimizing dissent to minimize discussion of issues
3- Jungle primaries or equivalent gaming of the rules, to squeeze moderates and the opposition out of the electoral process
In the 70’s the UK and Sweden both came close to falling into that trap but managed to step back from the abyss. Now it’s the US’s turn to choose.
It’s not as if the warning signs aren’t all around us.
In SF there is a detailed example of a “shining democracy” that went all the way down that path, even further than Argentina (whose Falklands campaign followed the same logic of the Legislaturists of Haven in David Weber’s Honorverse novels.) Weber is a historian and it doesn’t take much analysis to see where he gets the sociopolitical setttings for his galactic societies.
Fun reads, btw.
I’ll clue in CNN, and the rest of the MSM – they don’t have the trust now, so they aren’t going to lose all that much with this technology.
They’ve been editing things for years to match their desired narratives (Zimmerman and Pulse Nightclub tapes, photos of the Selma anniversary march, etc.). They’ve also been touting actual forgeries as the real thing when it fits the narrative (Bush National Guard memo). They may not create them themselves, but anyone with a brain will not automatically accept anything that supports their ideology. That train left the station a long time ago.
Actually, CBS did it to me ages ago.
They edited my answer to combine parts of two separate sentences and now I can’t show my face in San Diego. 😉
We already have them – go watch a movie.
What we are getting now is to where they can be made quickly enough to be used to fool the public in real time – not that the public seems to need much fooling these days.
And it’s not just for the pros/governments!
https://www.daz3d.com/facegen-artist-pro
“Import your own photo or multiple for even more accuracy to FaceGen. Then watch as FaceGen generates a shape and texture that you can easily apply right onto a Genesis, Genesis 2, Genesis 3, or Genesis 8 figure.
FaceGen even analyzes your pictures to create an overall skin texture that can be applied to your entire figure to avoid transitions.
If you want to make your figure look like your pictures, that’s all you need.
FaceGen lets you take things even further. You can adjust skin tone and feature shape gradually as you dial up and down various ethnicity adjustments to your photo’s face. Adjust its level of masculinity or femininity. Adjust age, color, shape, and even tween between multiple photos to combine more than one person’s structure and skin tones.
The ultimate in control, FaceGen even lets you modify the face with a pinch and pull style experience so you can mold and fine-tune features to meet your needs.”
So, got a couple hi-res shots of someone you’d like to add to your next book cover? 😉
The character builders for many video games are maybe one generation away from true photorealism.
And the new generation of consumer video cards announced this week look to have the horsepower to get there. The software may or not be here but if not its not far off.
https://finance.yahoo.com/news/amd-introduces-radeon-pro-wx-130000786.html
By now most everybody has heard of Photoshopping and movie CGI and knows not to trust photos or movies. Time to extend that skepticism to anything not personally witnessed.
Ah, and the stage magician will teach them that even what you see can’t be believed (never mind the fact that eye witnesses get more wrong than right far too often.)
Nice cards, too bad I don’t have enough ‘mad money’ to go pick one up, but even older/lower end cards (I have a pair of GTX1050ti in one desktop) change change hours to picture to minutes.