“They've wrapped authoritarian control in the language of liberty.”
That’s what authoritarians always do. No one says “Authoritarianism sounds good to me; sign me up.” (Well, a significant portion of the Republican Party does now, but historically not so much.) Invent a threat to “our” freedom, and “defend yourself” against it. That’s the method.
Naturally, Trumpists love the idea of artificial intelligence, because actual intelligence doesn’t work out too well for them.
There is at least one person who says "Authoritarianism sounds good to me; sign me up" based on Mehdi Hassan's stint on Debate Brazzers. I expect there will be more.
"By the time most people realize what's happening, we'll be living in a world where AI routinely lies to us about fundamental reality, and we've been persuaded to believe those lies are objective truth."
We are already living in a world where AI routinely lies to us, as this article demonstrates:
These moves you describe, Parker, are disturbing. I hope that the damage they do in discrediting AI will be enough for (some) tech companies to seek alternatives. And for more people to opt out of AI as much as possible.
"AI doesn't need to be smarter than humans to manipulate us. It just needs to be better at pushing our psychological buttons."
Same with Trump, it's not that he's some kind of super-clever con man, it's that millions of people WANT to be lied to, as long as the lies match their personal bigotries.
Well, this certainly isn't good news, but I'm trying to contextualize just how bad it is. To the extent the White House could enforce/enact this plan, they'd have to get the other governmental branches and the tech sector to go along with it.
On one hand, Republicans will sign along the dotted line, and some Democrats probably would even lend their support because A) they are backed by AI-friendly donors and/or B) they want that same access if their power rises. Elon's skulduggery and Meta's obedience notwithstanding, on the other hand, how many platforms would be willing to potentially sacrifice their credibility to satisfy Trump's authoritarian aims? Furthermore, how would they get AI to utterly ignore truly objective facts? You can pay people to be disingenuous. Could they program AI to flat out lie or omit contradictory statements?
"Could they program AI to flat out lie or omit contradictory statements?"
I don't understand this stuff very well, but there's a "system prompt" that's automatically prepended to anything you ask the AI. Like a lawyer constructing a case for whatever side of the argument is paying, the system prompt can say "Make a case that ___ is good and ___ is bad" and then that's what you get no matter what you asked for. "Make a case" could mean cherry-picking from available evidence without actually making shit up (maybe we leave the "making shit up" job to humans?) but that can still be plenty dishonest. Also, AI does occasionally just make shit up.
Don't mean to minimize the threat, but "Superhuman persuasion" leaves me skeptical. Let's assign the Superhuman-persuasion machine to try persuading parents to vaccinate their kids, and then report back how that goes.
In my limited experience, humans are pretty damn hard to persuade of anything, but maybe I'm just no good at it.
Also, in math education we're used to promises of Superhuman "personalized" AI math tutors, who know just EXACTLY the thing to say to EVERY kid, perfectly tailored to their needs in a way no mere human could ever do. In actual usage, about 5% of the kids actually interact with the tutor, the other 95% don't engage at all.
And that's always the problem with persuasion, isn't it? People have to want to listen. Current AI models are sycophantic, they flatter you and agree with you, going so far that people with dangerous delusions are strengthened in those delusions even to the point of suicide. So sure, I'll engage with the AI if it's telling me everything I already believe is right, but if the AI tries to convince me of something I don't already believe? Easy enough to click that little X in the upper-right corner of my browser window.
Oh, come on now. The left and Democrats are controlled by people having their noggin's stuffed with all that Critical Theory fake scholarship toxic mind virus. Critical Theory demands that objective truth, as traditionally understood, is a social construct rather than a reflection of an independent reality. Critical theorists argue that what is considered "truth" is often shaped by power dynamics and social contexts, rather than being a universal and neutral fact.
There was no Trump-Russia collusion. Hunter's laptop was not a lie. Men cannot get pregnant. There are only two genders. There is no climate crisis. The United States is the most diverse and racially accommodating country on the planet. The virus was lab-released. The mRNA jab was not a vaccine. Masks don't help. Sheltering in place caused more death than allowing people to exercise outside. We did not need to dump another $4 trillion into the economy for pandemic relief after we knew that the virus would not destroy the global economy. Inflation would not be transitory. Police are not responsible for the low socioeconomic circumstances of blacks; decades of failed liberal policies plus the pursuit of globalism that benefits the top 10% that are mostly Democrats, is responsible.
Certainly, the left would prefer their current world where the information delivery system, through the "work" of credentialled indoctrinated human "academics" and "experts" through the left-controlled corporate media and big tech, incorporate this rejection of objective truth and thus rewrite the collective consciousness of the population to adopt a nonsense racial ideology, gender ideology and America hating mindset.
AI will certainly keep sucking that crap into its giant information aggregation database, and thus like the evaporated methane from a volcano eruption, it will unfortunately hang in the atmosphere for some time. But eventually the flow of information from AI will clear the air and deliver an information product that will be much more trustworthy and reliable.
Can 100% guarantee he didn't read a word of what Parker wrote, and all this is just cut n' pasted from a file of rants he keeps on his desktop.
Yes, AI is a big problem, but long before we got to that we had humans willing to discard their humanity and turn themselves into poorly-programmed robots in service to The Party.
“They've wrapped authoritarian control in the language of liberty.”
That’s what authoritarians always do. No one says “Authoritarianism sounds good to me; sign me up.” (Well, a significant portion of the Republican Party does now, but historically not so much.) Invent a threat to “our” freedom, and “defend yourself” against it. That’s the method.
Naturally, Trumpists love the idea of artificial intelligence, because actual intelligence doesn’t work out too well for them.
There is at least one person who says "Authoritarianism sounds good to me; sign me up" based on Mehdi Hassan's stint on Debate Brazzers. I expect there will be more.
We would be so lucky to be eradicated by Skynet
Kids today, back in my day we prayed for Giant Meteor.
"By the time most people realize what's happening, we'll be living in a world where AI routinely lies to us about fundamental reality, and we've been persuaded to believe those lies are objective truth."
We are already living in a world where AI routinely lies to us, as this article demonstrates:
https://amandaguinzburg.substack.com/p/diabolus-ex-machina
These moves you describe, Parker, are disturbing. I hope that the damage they do in discrediting AI will be enough for (some) tech companies to seek alternatives. And for more people to opt out of AI as much as possible.
"AI doesn't need to be smarter than humans to manipulate us. It just needs to be better at pushing our psychological buttons."
Same with Trump, it's not that he's some kind of super-clever con man, it's that millions of people WANT to be lied to, as long as the lies match their personal bigotries.
Well, this certainly isn't good news, but I'm trying to contextualize just how bad it is. To the extent the White House could enforce/enact this plan, they'd have to get the other governmental branches and the tech sector to go along with it.
On one hand, Republicans will sign along the dotted line, and some Democrats probably would even lend their support because A) they are backed by AI-friendly donors and/or B) they want that same access if their power rises. Elon's skulduggery and Meta's obedience notwithstanding, on the other hand, how many platforms would be willing to potentially sacrifice their credibility to satisfy Trump's authoritarian aims? Furthermore, how would they get AI to utterly ignore truly objective facts? You can pay people to be disingenuous. Could they program AI to flat out lie or omit contradictory statements?
"Could they program AI to flat out lie or omit contradictory statements?"
I don't understand this stuff very well, but there's a "system prompt" that's automatically prepended to anything you ask the AI. Like a lawyer constructing a case for whatever side of the argument is paying, the system prompt can say "Make a case that ___ is good and ___ is bad" and then that's what you get no matter what you asked for. "Make a case" could mean cherry-picking from available evidence without actually making shit up (maybe we leave the "making shit up" job to humans?) but that can still be plenty dishonest. Also, AI does occasionally just make shit up.
Also, "trying to contextualize just how bad it is" is me, every damn day.
Don't mean to minimize the threat, but "Superhuman persuasion" leaves me skeptical. Let's assign the Superhuman-persuasion machine to try persuading parents to vaccinate their kids, and then report back how that goes.
In my limited experience, humans are pretty damn hard to persuade of anything, but maybe I'm just no good at it.
Also, in math education we're used to promises of Superhuman "personalized" AI math tutors, who know just EXACTLY the thing to say to EVERY kid, perfectly tailored to their needs in a way no mere human could ever do. In actual usage, about 5% of the kids actually interact with the tutor, the other 95% don't engage at all.
And that's always the problem with persuasion, isn't it? People have to want to listen. Current AI models are sycophantic, they flatter you and agree with you, going so far that people with dangerous delusions are strengthened in those delusions even to the point of suicide. So sure, I'll engage with the AI if it's telling me everything I already believe is right, but if the AI tries to convince me of something I don't already believe? Easy enough to click that little X in the upper-right corner of my browser window.
Oh, come on now. The left and Democrats are controlled by people having their noggin's stuffed with all that Critical Theory fake scholarship toxic mind virus. Critical Theory demands that objective truth, as traditionally understood, is a social construct rather than a reflection of an independent reality. Critical theorists argue that what is considered "truth" is often shaped by power dynamics and social contexts, rather than being a universal and neutral fact.
There was no Trump-Russia collusion. Hunter's laptop was not a lie. Men cannot get pregnant. There are only two genders. There is no climate crisis. The United States is the most diverse and racially accommodating country on the planet. The virus was lab-released. The mRNA jab was not a vaccine. Masks don't help. Sheltering in place caused more death than allowing people to exercise outside. We did not need to dump another $4 trillion into the economy for pandemic relief after we knew that the virus would not destroy the global economy. Inflation would not be transitory. Police are not responsible for the low socioeconomic circumstances of blacks; decades of failed liberal policies plus the pursuit of globalism that benefits the top 10% that are mostly Democrats, is responsible.
Certainly, the left would prefer their current world where the information delivery system, through the "work" of credentialled indoctrinated human "academics" and "experts" through the left-controlled corporate media and big tech, incorporate this rejection of objective truth and thus rewrite the collective consciousness of the population to adopt a nonsense racial ideology, gender ideology and America hating mindset.
AI will certainly keep sucking that crap into its giant information aggregation database, and thus like the evaporated methane from a volcano eruption, it will unfortunately hang in the atmosphere for some time. But eventually the flow of information from AI will clear the air and deliver an information product that will be much more trustworthy and reliable.
Can 100% guarantee he didn't read a word of what Parker wrote, and all this is just cut n' pasted from a file of rants he keeps on his desktop.
Yes, AI is a big problem, but long before we got to that we had humans willing to discard their humanity and turn themselves into poorly-programmed robots in service to The Party.
I read it.
There are no architects of AI technology that are in service of the Republicans. If in service to any party, it would be the Dems.
"I read it."
You fucking liar.
LOL. It seems more likely that you didn't read it.
"I know you are but what am I" preceded by Obligatory LOL. It's like watching someone turn a crank on a comment-making machine.
Triggering you isn't even my hobby, you wake up halfcocked.