Pepe Escobar: How political analysis became a target of A.I. fakes
Welcome to A.I. turning the net into an infernal machine bent on erasing meaning, culture and History – and sowing deep intellectual confusion. Exactly like Techno-Feudalism wants it.
A.I. is fast expanding as a plague all along the internet spectrum. That’s quite predictable, considering the Big Tech model for A.I. is techno-feudalism, relying on profit and mind/social control, and not on sharing/expanding knowledge and creating better conditions for a well-informed citizenry.
A.I. in many aspects is the antithesis of civitas. Prior to the A.I. boom, several layers of the internet had already been distorted into a series of minefields across a large-than-life sewer. A.I. – as controlled by Big Tech – in many aspects had already revealed itself as a fraud. Now it’s a weapon.
There are several channels on YouTube manipulated by A.I., stealing the image and voice by some of us, independent political analysts. A not-extensive list includes as targets John Mearsheimer, Larry Johnson, Richard Wolff, Glenn Diesen, Yanis Varoufakis, economist Paulo Nogueira Batista and myself.
It’s not an accident that all of us are independent geopolitical and geoeconomic analysts, mostly know each other personally, and are guests in roughly the same podcasts.
In my own case, there are channels in English, Portuguese and even Spanish: I rarely do podcasts in Spanish, so even the voice is fake. In English, usually the voice is approximately cloned. In Portuguese it comes with an accent I don’t have. In several cases, audience numbers are huge. Essentially, these come from bots.
In all cases, as far as we, the targets are concerned, all these channels are fake. I repeat: all these channels are fake. They may at least in some cases be set up by “fans” – certainly with an eye for profit via monetization.
Or the whole scam may be part of something way more sinister: a strategy bent of loss of credibility. As in an operation by the usual suspects to sow confusion amongst the – large – audience of several independent thinkers.
It’s not an accident that quite a few viewers are already deeply puzzled. Cue to the most common question: “Is this really you, or A.I.?” Many apparently have denounced these fake channels, but YouTube, so far, has done absolutely nothing about them. The algos keep suggesting these channels to large audiences.
The only realistic way to fight the scam is to file a complaint with YouTube. But that, in practice, is pretty useless. YouTube management seems to be more interested in occasionally erasing “inconvenient” channels displaying critical thinking and analysis.
Cracking the code of the scam
Quantum Bird, a physics and HPC (High Performance Computing) expert, formerly with the CERN in Geneva, has cracked the code of the scam:
“The proliferation of agents of deep learning digital neural networks capable of emulating writing, voice and video of human beings was inevitable, and their impact on scientific research, production of knowledge and art in general has a negative potential that has not been yet fully analyzed.”
He adds: “While writers and academics are detailing the springing up of texts attributed to them, and replicating to a certain extent their style and opinions, the latest fad is the blooming of whole channels on YouTube, and other notorious Big Tech platforms, that offer videos of popular content producers, communicating in their native language or other languages. In several cases, the quality of this synthesized material is sufficiently high not to allow immediate identification by an average viewer. In the context of the political analysis community, the impact is obvious: historic revisionism, erosion of reputations and distortion of news and analysis.”
And here Quantum Bird lays out the tech clincher:
“The synthetization of this type of content requires the availability of abundant samples and massive computational capacity, way beyond the reach of domestic users. While the popularity of the YouTube victims guarantees the first condition, the second one suggests the activity of large-scale state or corporate actors, since advanced deep learning models must be developed and trained by processing a huge quantity, in terms of “disk space”, of audio and video. The monetization of the content does not cover the costs of this operation. Ironically, it’s the availability and the excess exposure of voice and video online that allows this type of attack.”
Here we go. Welcome to A.I. turning the net into an infernal machine bent on erasing meaning, culture and History – and sowing deep intellectual confusion. Exactly like Techno-Feudalism wants it.
Whilst I largely agree with the above, I have found the AI Grok to be a million miles away from the likes of gemini et al. Grok appears to have a different ethos entirely and just look at Elon Musk and his actions and compare them to the usual suspects,… Read more »
Hi Claran and welcome, Grok? Who uses western AI? China is where it is at because they are using this technology as it should be used. And this is much bigger and notable than things like Grok, which is pablum for the masses. We have a conversation point now in… Read more »
At this time, imho, the dark side is in the experimental stage, sampling the effects and developing the next set of large, probably larger, scale experiments for the audience to be subjected to. The servile boffins having diabolical fun is not the goal, and probably mere confusion is not the… Read more »
Yes indeed. The mind games soar, like bombs bursting in air.. G7-slaves are now convinced “America First” = they fork over EVERYTHING.
For slave-proles too, sweet total digital immersion beckons, an endless neurostimulatory loop, like messianic lab rats hittin’ the juice button
saw it in passing on some algo recommended show, and now can’t remember which, but it was talking about the dangers of AI and comparing it to how https became the safer encryption standard for the internet, but there is little to no work at all being put into AI… Read more »
It’s not AI. It is the use of AI. It is evil scum people that do this, not the AI platforms themselves. I used some AI detectors on some videos and the detectors cannot even detect them any longer. Solution for now – we go back to writing yes? We… Read more »
Also Amarynth its the total failure from the beginning of western social media to have any kind of rules and morality governing the platforms, they don’t care in fact encourage the criminals of every kind. Western tech platforms are the wild west so welcome to the unchecked depravity of Westworld,… Read more »
I am sure the platforms were created to be the next psyops weapon when tell-lie-vision has already long since lost its effectiveness. They care alright, just not the way most who get fooled again and again, usually willfully, expect.
Or have half the AI generated content gain trust and get people to let down their guard. Like humans such as perhaps even Sal Mercogliano and Bonespurs have been doing for decades.