The idea of truth based on recorded media is a boat that sailed away long long ago.
People are very used to being deceived by advertising in general and political media especially. The article mentions certain politicians dismissing any media critical of them as AI deepfakes, a claim the article then instantly fact-checks (without irony).
Denial and lies are not new in any sense. Fake media is not new, and analog fakes have a very long history. Photoshop didn't destabilize anyone's concept of truth beyond the extremely gullible.
It's hard for me to imagine AI having significant, let alone measurable impact here.
If you’re struggling to imagine AI having a significant and measurable impact, then I think you’re failing account how many people, who are not extremely gullible, are influenced simply by repetitive narratives and group think.
AI will simply accelerate the creation of already effective targeted propaganda.
> I think you’re failing account how many people, who are not extremely gullible, are influenced simply by repetitive narratives and group think.
Like all those people who work for money, to buy things; repeat, repeat.
Take away human language and its hallucinated meaning and we’re just meat based cassette tapes repeating yesterdays application of agency; output complete sentences, feel validated by the dopamine buzz for engaging in anything, eat, shit, go to bed.
A whole lot of daily life is just repeating yourself. Office jobs are “capitulate to group think”. Society requires a whole lot of capitulate to group think; everyone needs a tribe and finds one in the end.
Only people working directly on infrastructure, health, and tech breakthroughs are moving the needle and even then in small repetitive steps.
If you think you’re above that altogether then you’re the one chanting propaganda. Given the complexity of the universe, no center, you have no idea if you’re moving anything forward or backward; you can’t even point in the direction “forward”. Make more 1900s technology and software? Repeat. yawn
> The idea of truth based on recorded media is a boat that sailed away long long ago.
And if it was ever here, it was only here for the blink of an eye, relatively speaking. Almost all of human history took place before recorded media was invented.
You can read old George Orwell essays about his experience with the media in the Spanish Civil War that sound like they could have been written yesterday. He talks about how he felt the media was entering some sort of post truth era even then from the blatant lies and total split in reality the newspapers would have amongst themselves
I believe the concern is that, now, it's incredibly easy to make a very believable lie. It takes skill to Photoshop an image believably, or it takes social weight to push a false story as believable.
But now, AI has democratized these actions. A 15yo with an internet connection can make the fake Biden voice calls in about 10 minutes.
Seems like that will simply lead to a huge glut of noise, more quickly and thoroughly desensitizing current and subsequent generations to fakes.
The reason I mention Photoshop is because it is many orders of magnitude more automated than photo editing using previous tools. The social perception of truth wasn't effected despite this new efficiency, because a faster to make lie isn't significantly different from the absolute deluge of lies people deal with every day.
It might be more accurate to say some people are using AI to try and destabilize the concept of truth.
AI sitting on it's own on a computer won't do this. Someone has to compel it.
My question is if it can be used to destabilize truth, how can AI help make inherent truth (true from all/most angles and interpretations) stand out more?
It's not necessarily the case that because AI can destabilize truth, that it can also re-stabilize truth. It's much easier to burn something down (e.g. trust) than build it up. c.f. second law of thermodynamics, etc.
> My question is if it can be used to destabilize truth, how can AI help make inherent truth (true from all/most angles and interpretations) stand out more?
I'd say that it can't, because it's not an AI problem, but rather a media problem.
In general, people are more interested in sensational reporting ("You won't believe what X did") than well thought out, factual reporting.
The solution is a slower news cycle with enough time to do fact checking and analysis of how the events fit in the bigger picture, but that's incompatible with up to the minute news. AI doesn't help with this.
Ok, then to the extent it's a media problem, which is to me a proxy for a problem of interpretation and disseminating interpretation as facts.
Click Bait will be effective as long as the human brain is wired to process the environment for threat. "You won't believe what X tried to report on Y was doing".
A slower news cycle could be useful, but it makes me wonder if that's not how it was already.. in the past, and also centralized the news.
Learning from news itself may be part of the issue. I remain surprised by what a large percent of the population doesn't know the difference between journalism and opinion/editoral writing. Maybe something like www.mediabiaschart.com can help.
That’s probably something that Postmodern philosophers have tried to answer but for more boring Information Overload Eras (telegraph; newspapers; mass media; the Web).
by helping investigative journalists sift through large datasets to uncover patterns, like the Panama Papers or Snowden docs. essentially, finding leads for journalists to investigate.
This new world of "maybe AI made it"-everything affords politicians plausible deniability for every statement they make. People would probably start to notice when the "we actually want you to believe this" statements are signed but others are not.
Well, it's really either up to the gov to mandate crypto signatures. Or it's up to the people to declare widely enough that they wont trust anything they see or hear unless the creators do something to prove authenticity.
This article refers to all sorts of incidents -- but gives no link or proof. There's something ironic about essentially asking readers to "just trust us" in an article about the meaning of truth.
Readit News
People are very used to being deceived by advertising in general and political media especially. The article mentions certain politicians dismissing any media critical of them as AI deepfakes, a claim the article then instantly fact-checks (without irony).
Denial and lies are not new in any sense. Fake media is not new, and analog fakes have a very long history. Photoshop didn't destabilize anyone's concept of truth beyond the extremely gullible.
It's hard for me to imagine AI having significant, let alone measurable impact here.
AI will simply accelerate the creation of already effective targeted propaganda.
Like all those people who work for money, to buy things; repeat, repeat.
Take away human language and its hallucinated meaning and we’re just meat based cassette tapes repeating yesterdays application of agency; output complete sentences, feel validated by the dopamine buzz for engaging in anything, eat, shit, go to bed.
A whole lot of daily life is just repeating yourself. Office jobs are “capitulate to group think”. Society requires a whole lot of capitulate to group think; everyone needs a tribe and finds one in the end.
Only people working directly on infrastructure, health, and tech breakthroughs are moving the needle and even then in small repetitive steps.
If you think you’re above that altogether then you’re the one chanting propaganda. Given the complexity of the universe, no center, you have no idea if you’re moving anything forward or backward; you can’t even point in the direction “forward”. Make more 1900s technology and software? Repeat. yawn
Time to repeat sleep. Good night.
And if it was ever here, it was only here for the blink of an eye, relatively speaking. Almost all of human history took place before recorded media was invented.
But now, AI has democratized these actions. A 15yo with an internet connection can make the fake Biden voice calls in about 10 minutes.
The reason I mention Photoshop is because it is many orders of magnitude more automated than photo editing using previous tools. The social perception of truth wasn't effected despite this new efficiency, because a faster to make lie isn't significantly different from the absolute deluge of lies people deal with every day.
AI sitting on it's own on a computer won't do this. Someone has to compel it.
My question is if it can be used to destabilize truth, how can AI help make inherent truth (true from all/most angles and interpretations) stand out more?
https://en.wikipedia.org/wiki/Gish_gallop
I'd say that it can't, because it's not an AI problem, but rather a media problem.
In general, people are more interested in sensational reporting ("You won't believe what X did") than well thought out, factual reporting.
The solution is a slower news cycle with enough time to do fact checking and analysis of how the events fit in the bigger picture, but that's incompatible with up to the minute news. AI doesn't help with this.
Click Bait will be effective as long as the human brain is wired to process the environment for threat. "You won't believe what X tried to report on Y was doing".
A slower news cycle could be useful, but it makes me wonder if that's not how it was already.. in the past, and also centralized the news.
Learning from news itself may be part of the issue. I remain surprised by what a large percent of the population doesn't know the difference between journalism and opinion/editoral writing. Maybe something like www.mediabiaschart.com can help.
https://www.newadvent.org/cathen/10415a.htm
Pontius Pilate asked the rhetorical question, "What Is Truth?" as Truth Himself stood before him on trial.
It's not a new technique or anything.