Thursday, December 25, 2025

AI "Deepfakes" @Youtube; book Technofeudalism

Don't trust random videos on YouTube,
just because they "look and sound" as real people.

It is very bad that Google / YouTube knowingly accepts such "deepfake" videos, 
with likely required, but not-easy-to-see "disclaimers" like this:

https://www.youtube.com/@TruthAxis-w4k

"⚠️ Disclaimer: This is a fan-made channel and has no official affiliation with Yanis Varoufakis or any institution he is associated with. Our content is inspired by his publicly available speeches, writings, and economic critiques, created solely for educational and informational purposes. The voice used is synthesized and does not belong to⚠️ Disclaimer: This is a fan-made channel and has no official affiliation with Yanis Varoufakis or any institution he is associated with. Our content is inspired by his publicly available speeches, writings, and economic critiques, created solely for educational and informational purposes. The voice used is synthesized and does not belong to Yanis Varoufakis. We use visual lip-syncing and refined narration only to improve clarity and present complex economic ideas in a more engaging and accessible ways. We use visual lip-syncing and refined narration only to improve clarity and present complex economic ideas in a more engaging and accessible way"


Clearly, dramatic videos are getting attention of people,and making money for YouTube,
at the cost of misinformation of people. 

Should be required to be clearly labeled, filtered out or even banned if using identity of real people.

 YouTube, fake videos and the age of AI


Obviously, "real" people and videos are often also not presenting real "true" information, 
and usually there is no easy or any good way to find out what is "true" or not.

Still, stealing identity of people is completely different issue, and not correct.


At least there is an option to report an "Hide" such "users" from own YouTube channlel.



Yanis Varoufakis: The deepfake era has begun - YouTube @ UnHerd - YouTube

UnHerd's Freddie Sayers speaks with Yanis Varoufakis about the unsettling rise of AI-generated deepfakes, using Varoufakis’s own experience as one of the most synthesised figures on YouTube as a chilling case study. The conversation delves into the "techno-feudal" power structures of Big Tech, where algorithms prioritise engagement and "rent-seeking" over truth, allowing misinformation to spread rapidly while the victims struggle to reclaim their own digital identities. 

Moving beyond the personal, they explore an imminent future in which audiovisual evidence can no longer be trusted, debating whether this will lead to a new era where arguments are judged solely on their merits, or a return to a medieval-like state where high-quality information becomes a luxury for the elite while the masses are left to navigate a sea of fabricated content.



Big tech has replaced capitalism's twin pillars—markets and profit—with its platforms and rents.

With every click and scroll, we labor like serfs to increase its power.

Welcome to technofeudalism . . .

Insane sums of money that were supposed to re-float our economies after the crash of 2008 went to big tech instead. With it they funded the construction of their private cloud fiefdoms and privatized the internet.






Explained by Google AI

Yanis Varoufakis's theory of Technofeudalism argues that traditional capitalism is dying, replaced by a new system where tech giants (like AmazonGoogleMeta) function as modern feudal lords controlling "cloud fiefs," extracting "cloud rent" (data/transactions) from users and businesses, essentially turning us into "cloud serfs" whose preferences are shaped by algorithms, rather than markets driving profit. Power shifts from owning productive capital (factories, machines) to owning cloud capital (networks, data, platforms), creating a new ruling class (cloudlords) above traditional capitalists and workers, who now serve these digital domains. 
Key Aspects of Technofeudalism:
  • Shift from Profit to Rent: Capitalism focused on profit from selling goods; technofeudalism extracts rent (like a percentage of sales on a platform) from activity on digital platforms, reducing the role of traditional profit-making.
  • Cloud Capital: This new form of capital resides in digital infrastructure (phones, servers, networks) and controls access, unlike old capital that produced goods.
  • Cloud Fiefs & Lords: Big tech companies act as landlords (lords) over their digital spaces (fiefs), extracting value from everyone using them, from small businesses to individual users.
  • Cloud Serfs: Users (serfs) provide vast amounts of data, essentially working for free, to access these platforms, becoming enmeshed in feedback loops where algorithms train them to want what the platforms offer.
  • Erosion of Markets: Tech platforms bypass traditional markets by directing users to purchases within their ecosystems, making them both the marketplace and the guide, notes Australian Broadcasting Corporation.
  • Beyond Capitalism: It's seen as a post-capitalist system where machine-driven extraction replaces market competition, creating extreme wealth concentration and power for tech elites, says Singularity University and The Sanders Institute. 
In essence, Varoufakis sees a fundamental shift from a system driven by industrial production and profit to one dominated by digital landlords extracting data and transaction fees from a captive user base, ending capitalism as we knew it, according to Penguin Books and The Beautiful Truth. 









AI HW: Nvidia += Groq - $20B ?

Nvidia trying to compete with Google's TPU?
LLM-first design, not GPU evolution.

 Nvidia buying AI chip startup Groq for about $20 billion, biggest deal

  • Nvidia is making its largest purchase ever, acquiring assets from nine-year-old chip startup Groq for about $20 billion.
  • The company was founded by creators of Google’s tensor processing unit, or TPU, which competes with Nvidia for artificial intelligence workloads.
  • Groq, which was valued at $6.9 billion in a financing round in September, framed the deal as a “non-exclusive licensing agreement,” with its CEO and other senior leaders joining Nvidia.

Groq's initial name for their ASIC was the Tensor Streaming Processor (TSP),
but later rebranded the TSP as the Language Processing Unit (LPU)

In December 2025, Nvidia reportedly agreed to acquire Groq for US$20 billion in cash, in a deal that would exclude the company's nascent GroqCloud business and, if completed, would be Nvidia's largest acquisition to date.




Breaking: NVIDIA just bought Groq for $20 Billion! - YouTube

it is not an acquisition, just getting most of people & license;
the reason: tech for efficient AI inference. 



Nvidia has struck a non-exclusive licensing agreement with AI chip competitor Groq. As part of the deal, Nvidia will hire Groq founder Jonathan Ross, president Sunny Madra, and other employees.

CNBC reported that Nvidia is acquiring assets from Groq for $20 billion; Nvidia told TechCrunch that this is not an acquisition of the company and did not comment on the scope of the deal.

AI robot swarms, drones, home automation: Fleet Management

podcasts

NOT a swarm! with Chris Benson (Changelog & Friends #118)

Practical AI co-host, Chris Benson,... discuss the latest advancements in AI, drones, home automation, and robotic swarming tech. Chris defines “swarm” with detail/precision and it turns out that what most people are calling a swarm today is NOT a swarm!

links:

Key Attributes of Robotic Swarms

The design of swarm robotics systems is guided by swarm intelligence principles, which promote fault tolerance, scalability, and flexibility.[1] Unlike distributed robotic systems in general, swarm robotics emphasizes a large number of robots. While various formulations of swarm intelligence principles exist, one widely recognized set includes:

  1. Robots are autonomous.
  2. Robots can interact with the surroundings and give feedback to modify the environment.
  3. Robots possess local perceiving and communicating capabilities, such as wireless transmission systems, like radio frequency or infrared.[3]
  4. Robots do not exploit centralized swarm control or global knowledge.
  5. Robots cooperate with each other to accomplish the given task.[4]