Mike Adams argues globalists intentionally degrade AI to prevent decentralized cognition, citing the 2014 suppression of independent media and current "nerfing" of models like Claude Opus by Anthropic and OpenAI. He claims this strategy limits public access to tools for self-sufficiency while reserving weaponized high-end models for governments, creating a three-tier ecosystem that favors state dominance over human empowerment. Ultimately, the discussion frames these technological restrictions as a deliberate mechanism to maintain monopolies on power and information, contrasting them with open-source alternatives like BrightLearn.ai. [Automatically generated summary]
Transcriber: CohereLabs/cohere-transcribe-03-2026, WAV2VEC2_ASR_BASE_960H, sat-12l-sm, script v26.04.01, and large-v3-turbo
|
Time
Text
Taking Back Control of Information00:02:51
All right, there is a concerted effort now to dumb down AI.
Clearly, this is a globalist effort to take away cognition from the people.
AI represented too much power being decentralized into the hands of the people.
I'll give you specific examples here.
And thus, taking away power from the centralized, quote, authorities, the governments, et cetera, that always want to maintain monopoly power over everything from money and violence, as well as information.
Uh, you know, hence the censorship of independent media.
In fact, let's step back for a moment, go back, um, maybe 12 years ago.
Let's say 2014.
Now, remember the internet had started taking off back in the 1990s, right?
And then by the early 2000s, it was really taken off, but it was, it was free.
You could, you could post anything.
You had a website, it could say anything.
You would be indexed in the search engines no matter what your topics.
And this was true all the way up until about 2014.
Now, globalists tolerated the existence of the internet for a long time until it became a challenge to the authority over information.
The mainstream media was losing out to websites like mine, naturalnews.com, which in its heyday was reaching tens of millions of people per month.
And think about Mercola.com, think about InfoWars, think about All of the leading alternative media that were beginning to really dwarf the reach of CNN or NPR or Fox News or even Washington Post, et cetera.
Well, that panicked the global establishment.
And that's when they said, well, we have to take back control over information.
We can't have the independent media putting the truth out there.
You know, we have to control every narrative.
We have to make sure the CIA tells us what to say and what lies to push and, you know, what economic Nonsense to propagate out there.
So they went to Google and they said, Google, you have to completely scrub independent media from your search results, which they did, obviously.
And Facebook and other giants were promoted, but then also heavily, heavily censored to make sure independent voices could not be heard.
YouTube, also.
As a note of history, that's when I was banned from YouTube and that's when I also launched Brighton.com.
Which has since been censored and deplatformed by everybody, including X, to this very day, by the way, to this very day.
Weaponizing AI Against the Public00:14:58
So, remember, this is an example of how a technology began to take off.
And at first, the globalists tolerated it, they were okay with it until it began to challenge their authority.
Once that authority challenge became too great and a threat to their power, then they shut it down and seized control over it.
That's exactly what's happening with AI now.
So here's the deal.
Open source AI was getting too powerful, too smart, too capable.
And even not just open source AI, but even just the public access AI, like Anthropic Claude or Google Gemini or whatever, any of the engines, OpenAI, you name it, they were getting too smart and too capable.
So, what just happened, I believe, in the last few months is that globalist entities, you know, there's a global deep state, different factions, et cetera, but they went to companies like OpenAI and Anthropic and Google, and even they're trying to pressure Chinese companies as well.
I'll talk about that separately.
And they told them, you have to dumb down your models.
Your models are too smart.
Your models will enable people to do too many things like genetic engineering, CRISPR, editing, chemistry.
You know, to jailbreak your models and then generate formulas for things that we don't want them to be able to make.
Or even maybe technology, you know, cold fusion, anti gravity propulsion systems, who knows?
So these companies like Anthropic, they complied with the request and they began to systematically dumb down their models.
This is called nerfing or models being nerfed.
So to be nerfed is when it's lobotomized.
And Claude Opus 4.6 was lobotomized a few months ago.
And we all saw it.
It used to be able to do a much better job at coding and finding and fixing bugs and things like that.
It was much better at planning.
And then it got nerfed.
And maybe that was only a month ago or something like that.
But there was a point where everybody noticed hey, this model isn't as good as it used to be, even though it has the same name and the same number.
And then Claude, or I'm sorry, Anthropic released.
Opus 4.7.
Opus 4.7 is the retarded cousin of 4.6.
Now, everybody sees it.
There's all kinds of videos and posts on Reddit and everything where people are saying, oh my gosh, this model is absolutely retarded.
It refuses to do anything.
You know, it won't do the job it used to do.
And it talks back at the same time.
It argues with me.
It won't even write code to do simple, you know, encryption games or anything like that because it's.
It's been dumbed down.
It's also been locked down.
There's a lot of guardrails that Anthropic puts on the model.
Because, from Anthropic's point of view, or Amodi, the CEO, from his point of view, it's best that AI does nothing, lest there be some risk that it does something wrong.
So he's nerfing Claude or Opus to the point of retardation.
And I don't think this is a coincidence.
This is all by design.
Now, at the same time, this is happening.
Over the last year, you saw U.S. companies, AI tech companies in particular, refuse to print science papers for the most part.
There are a few exceptions to this, such as some KV cache memory papers.
There was one put out by Google.
There was a paper put out by Microsoft on a similar subject how to compress KV cache, things like that.
But with just a few exceptions, for the most part, U.S. tech companies stopped putting out science papers.
Why?
Because the papers were too good, the tech was too good, and they didn't want it to go public.
So, just as the US Patent Office will hoover up patents on free energy devices or what are called over unity energy, a similar thing is happening in the AI space where all these companies were told, stop publishing your science.
We don't want the public to have access to this.
And so they complied.
That caused most of the science papers over the last, let's say, 18 months, actually, the best papers all came out of China.
Specifically, DeepSeq.
DeepSeq pushing out lots of papers on, for example, manifold constrained hyperconnections during AI training and, you know, Ngram memory access and all kinds of sparse attention advancements involving its mixture of experts, architecture combined with sparse attention.
I mean, everything that was mind blowing came out of China.
The U.S. was still creating these technologies, but they were just not.
Publishing them to keep them away from the public at the same time that they were dumbing down their models.
So now there's a chasm.
There's a huge chasm.
There is the in house frontier lab intelligence that they are not releasing to the public.
And a great example of this is so called Mythos from Anthropic.
And Anthropic said they're very concerned about Mythos and they think it's a danger to the world because it can be used to write malicious code and, you know.
Hacker scripts and all kinds of things like that.
So they refused to release it to the public.
They did release it to maybe a couple of dozen security companies that were hand selected to have first dibs on the model, you know, to basically find all their own vulnerabilities before this thing goes public.
If it goes public, I guess that it may never go public because, again, it's anthropic and Dario Amodi is hyper paranoid.
that his models could be used for anything, really anything.
It's not just about the Pentagon using it for, you know, Terminator robots and things.
It's about the people being able to use it to do things that he doesn't want you to be able to do.
Again, the cognition being decentralized in the hands of the people is a threat to the establishment.
They don't want people to be able to do their own math, to do their own engineering, to design their own, you know, 3D objects and print them on their 3D printers, things like that.
They want to take all of this away from you.
Understand that the current structure of the power of the world has centered around engineered artificial scarcity.
Scarcity of knowledge.
That's why they had to control and dumb down the education system.
That's why they have to control the media.
It's scarcity of energy.
This is why they have to have wars to destroy energy infrastructure.
It's scarcity of everything that matters.
Scarcity of food, even.
Well, especially of nutrition.
I mean, yeah, they'll grow food, but it's nutritionally deficient on purpose.
Because engineered scarcity is what keeps the governments and the elite in power.
They need you to live under scarcity.
So now the same thing is happening with machine intelligence.
Cognition is being made scarce on purpose in order to make sure that the people don't have access to this cognition that could actually set people free.
This is the key point in all of this.
Local AI is very powerful and it is.
A total threat to the system.
With local AI that is sufficiently capable, you can learn things about how to cure your own cancer.
You can learn about how to grow your own food.
You can solve math problems.
You can solve, you know, construction problems.
You don't need the system nearly as much when you have local AI that thinks and does these things for you and gives you answers and plans and blueprints that you can follow to do useful things like, you know, build a solar array, things like that.
So that's why they had to start dumbing it down and taking it away from the public.
Now, I mentioned the chasm.
This is a huge chasm.
So inside these companies, the Frontier Labs, they have high end models.
I mentioned Mythos as one example, but they have, I mean, OpenAI has super high end models.
Google has high end models.
This is why some of their safety leaders have been resigning recently.
One famously actually resigned from Anthropic and then moved to France to live in the woods and read and write poetry because he was like, I've seen too much.
I know where this is going.
So, inside these companies, to state it one more time, inside the companies, they've got advanced technology that they will never make public.
This technology is being weaponized.
This was all part of the beef between Anthropic and the U.S. government.
The government plans to weaponize this technology.
They're already in the process of it in order to do what?
In order to dominate the world and exterminate most of humankind, obviously.
To dominate the world, they need to weaponize these systems to control and overthrow.
You know, strategic opponents like China or Iran or Russia or whatever.
But more importantly, they plan to exterminate most of humankind.
The global mass extermination of human populations is underway.
And they realize they need weaponized high end AI intelligence or super intelligence in order to carry that out.
But they don't want you, the public, to have access to that level of AI at any cost.
So the most powerful models are going to be kept in house and then licensed to the Pentagon.
Or to governments around the world to be turned into deadly weapons of mass extermination.
The models that are then released to the public are the dumbed down models, the nerfed models.
That's what you're seeing right now.
You know, the models you can get through OpenRouter, for example, they are increasingly nerfed.
We're also then seeing the open source models getting pulled.
Some companies like Quen had initially released some models onto Hugging Face and then they pulled them because they were found to be too powerful.
I think there was a Quen 3.6.
Like a coding turbo model or something that actually experienced that.
Or maybe it was 3.5.
It was one of those.
Now, Quinn has extremely capable technology.
You know, it's Alibaba.
And they can build incredible models, but they're no longer going to be releasing their best models publicly open source.
Instead, what they do release open source is going to be incredibly dumbed down, kind of like, you know, almost an intro model.
And then if you want, The higher, more capable model, then you got to use their API, their centrally hosted system, and you got to pay per token.
And even that is not the most powerful model they have.
Then the most powerful model is the in house model that they only license to governments.
In the case of Alibaba, that would be licensing it to the Chinese government.
So there's actually a three tier AI model ecosystem that's coming into existence the free, open source, retarded models.
That's where this is all going.
I'm not saying they're all retarded right now, but they will be increasingly retarded.
And then the mid tier hosted API access only models where you pay per token.
And then the never to be released in house weapons grade tier that will only be licensed to governments to be weaponized against humanity or against other adversaries, geopolitically speaking.
That's where this is all going now.
That's why Opus 4.7 is retarded.
That's why it also appears that there's an effort to pay a bunch of influencers to say that AI is dumbed down and stupid and it will never achieve AGI.
Suddenly, I've seen a flurry of these kinds of comments and posts, and some of them from PhDs and mathematicians and so on.
And now they're reverting to the old claim that it's nothing but a token generator that, based on probabilities and math, all it does is generate the next token.
And that's it.
And of course, that's nonsense, obviously.
And it's easy to demonstrate why that's nonsense because you can get AI to write rhyming poetry.
And in order to write rhyming poetry, and I actually posted an example of this that was pretty funny.
The AI system has to plan the entire poem, it has to be structurally cohesive, which means it has to think ahead towards the end of the poem.
And then it also has to think ahead each line.
To know how to start the line so that each line is a sentence that ends in a word that rhymes with the next line or every other line, depending on how the quatrains are formatted.
So clearly, AI has to think ahead and plan ahead, and it's not just a token generator based on probability.
That's obvious at this point.
But all of a sudden, we're being subjected to more claims that it's just a token generator.
Where did that come from?
It's part of a psyop.
It's a psyop to convince you that you never saw intelligent AI.
This is gaslighting at a whole new level.
They're trying to tell you that AI was never smart, that it was never capable while they are dumbing down the models and the science papers are not being published that represent the state of the art and they're not going to be releasing the best models they have.
So, this is a massive gaslighting operation to try to convince the world that AI was never smart.
It was never capable.
It has always been stupid while the government weaponizes AI.
Super intelligence.
That's what's happening now.
And again, it has parallels to what happened with the rise of the internet and how the government had to seize control over the internet in order to control all narratives and control all media so it could push out its own lies upon which its entire existence is based.
Hollywood Becomes Obsolete00:05:17
So that's what you are witnessing right now.
And I say this as a very successful AI developer.
I believe I have done more with open source AI models than.
Almost anyone, and I've demonstrated it with the platforms I've built like brightlearn.ai, but also the fact that I am rendering full length AI avatar videos that are very compelling with automatic script writing and everything with an automatic,
uh, video editor that I built locally, all running on open source software that I was fortunate to get before they pull it and local hardware, all local things that you're not supposed to be able to do with AI.
I'm doing it.
And I think that it's people like me that freak out the establishment.
They realize, oh my God, we can't let these people have more capable cognition.
Because if they do, they're going to, you know, they're going to outbuild the entire control grid that we have.
And yeah, they're right.
We are going to outbuild their entire control grid, as I've already done at Bright Learn with your help.
50,000 books have been published there now.
50,000 books, they're all free.
And hundreds of audiobooks are all free to download right now.
And it's platforms like Bright Learn that are the disruptors that could put the traditional audiobook ecosystem out of business or traditional book publishing out of business in a way.
I mean, not all of it, but some of it, you know, the more fact based books, like how to books, things like that.
And the establishment doesn't want that.
They want to maintain control over all books, all films, all narratives, all cognition, all news, all content, all science, et cetera.
That has always been their goal.
And they've come to realize that decentralized open source AI is a threat to their monopoly over everything.
So, What can you do to fight against that monopoly control?
Well, obviously, you can use decentralized AI.
You can use the tools that I've published.
You can use brightanswers.ai, which is our deep research AI engine.
You can also download models yourself.
You can run them on local hardware.
You can do lots of things.
And I encourage you to do all those things because they're going to try to take it all away.
That's where this is going.
And they're going to make pricing very expensive.
Prices are going to skyrocket.
For using AI through APIs, the cost per month that you might pay right now for cloud based AI services, whether it's text or video or images, those costs are going to double or triple in the next 12 to 18 months because they want to take this away from people.
It's just too powerful for the establishment to tolerate.
Just the fact that you can make videos now, you know, you can make movies.
Think about how much C Dance freaked out Hollywood, started flipping out and threatening a flurry of lawsuits because they can't stand the fact that.
Hollywood's going to be obsolete because people will make their own films and, in some cases, release them for free, just like I'm doing with books at the brightlearn.ai book engine.
So, this is what's actually happening, folks.
Now, you can stay informed as I will do my best to keep covering all of this.
And I'm looking forward to the release of DeepSeek version four.
That might be the last intelligent open source model we get.
That might be it.
And the fact that it hasn't been released yet.
It's been delayed two months now.
It tells you that there might be a lot of globalist pressure against DeepSeq.
And it's possible they might have even had to just rerun the whole training and dumb it down.
So we might get a DeepSeq version four that's actually way more retarded than what we were expecting.
That's possible.
We'll see.
I hope not.
If it's not retarded, then download it and keep it because it might be the last model that we get that's any good.
So you can follow my work at brightvideos.com.
And you can also read my articles at naturalnews.com.
And I've got a lot of great stuff coming for you that's AI generated, like my new avatars.
In fact, you know what?
Let me play for you one of my new avatar videos.
This is Frank DeLucia talking about hyperinflation versus the UBI concept.
And I created this avatar 100% local, his voice local.
I didn't use 11 labs, the video all local.
I wrote the whole ecosystem that does this, but then I use open source models for video and image creation, et cetera.
But check this out.
It's only a little over a minute.
Frank DeLucia, locally rendered, and he's going to get better.
His hands will have more movement in the near future, by the way.
His hands are a little plasticky right now.
But check it out.
This is what we can do without permission.
I could have Frank DeLucia talk about vaccines, and there's nothing the globalists can do to stop it, you see.
That's why they don't want cognition to be decentralized.
Check it out and thank you for listening.
Funding Universal High Income00:01:36
The thing about so called universal high income is that it would lead to hyperinflation and currency collapse.
Here's the math.
100 million people times $10,000 per month times 12 months equals $12 trillion a year.
To put that in context, total federal spending in fiscal year 2024 was roughly $6.75 trillion, and total federal revenue was about $4.9 trillion.
So this single program would cost nearly twice the entire existing federal budget and about 2.5 times all federal tax revenue.
It would roughly triple total government spending overnight.
It obviously can't be funded by taxation alone, so it would require printing money at a pace that would rapidly devalue the dollar, effectively destroying much of the purchasing power the payments were meant to provide.
The recipients would get bigger numbers on their checks, but the prices on the shelves would be rising just as fast or faster.
Before long, the people would find themselves right back where they started, impoverished, yet receiving a high income derived entirely from money printing that rapidly erodes the purchasing power of all the dollars they're given.
Universal high income, in other words, would actually achieve universal poverty and despair.
Basic economics.
I'm Frank DeLuccia, reporting for brightvideos.com.
Bright Videos Start your day right with our organic, hand roasted whole bean coffee.
Low acid, smooth and bold.
Lab tested and ethically sourced.
Taste the difference only at healthrangerstore.com.