Knowledge Fight Interactive Search Tool

All Episodes

Previous Next

Jan. 28, 2025 - Health Ranger - Mike Adams

01:01:43

Zach Vorhies issues EMERGENCY AI DIRECTIVE for America:...

Large

Audio Only |

Scroll

Time	Text
	Welcome everyone to this emergency interview.
	I call it an emergency because of what's happening in the stock market response to the realization that the DeepSeek R1 reasoning model has just changed the world, really.
	And I think the very best person to help us understand that is Zach Voorhees, Google whistleblower.
	And Zach and I have been talking privately for, well, since this model was released and we were both...
	Just blown away by it.
	But welcome, Zach.
	Great to have you on today.
	Thanks, Mike.
	Good to be back on the show.
	Hey, it's great to have you here.
	Let me give out your Twitter handle.
	It's Perpetual Maniac.
	Is there an underscore or anything in that, or is it just all together?
	Twitter.com slash Perpetual Maniac.
	Okay, Perpetual Maniac.
	And I love the IQ test in that because...
	It filters out people who can't spell perpetual.
	So good job.
	That's awesome.
	Second question.
	So the tech sector is crashing in the stock market all day because for some reason, Zach, Wall Street figured out what you and I have been talking about, I guess, for a week since R1 was released.
	What's your take on what's happening?
	Oh, man, where do I start?
	So NVIDIA has been valued very heavily because The amount of processing power necessary to train these hyper-intelligent AIs has been absolutely immense.
	And it's also non-linear, actually sublinear.
	So in order to get a 2x of performance, you need more than 2x the computation.
	Let's just say it's 4x for every doubling.
	And so everyone's been pricing that in into the NVIDIA stock.
	And what China did this last week is that they, out of nowhere, just dropped this DeepSeek R1 model.
	And what they showed the world is that they have AI supremacy.
	Now, R1 is very similar to the OpenAI's O1 in its performance.
	But unlike OpenAI's O1 model, the DeepSeek R1 model, Runs on a desktop computer at 3% the cost.
	That's what's happening right now.
	We're all reacting to it.
	Yeah, and you and I are both running this model on our desktop.
	I'm running the 32 billion parameter model.
	I squeeze it into this GeForce card, and it's just writing code all day.
	I'm using it for everything.
	Yeah.
	I mean, this is such a game changer.
	It is.
	And what's really strange about it is the fact that they were able to, I mean, Trump, Biden have both been putting export controls on the NVIDIA chips, the powerful H100s that are being used to train these AI models at data centers around the world.
	You know, under the Western Global Alliance and that umbrella.
	And they don't let China have any of these chips.
	And China just said, okay, well, we're very, very smart.
	We're going to figure out how to train these models anyways.
	We're going to do a better job.
	We're going to do it at one one hundredth of the computational capacity of the Western Alliance.
	And by God, they did it.
	It's incredible.
	It's incredible.
	So, okay, right off the bat, what do you think about – I mean, there's a narrative in the U.S. and among the tech giants that they think China – or they think the DeepSeek company is lying about what they did.
	But it sounds to me like a cover story because DeepSeek is releasing all the science papers that actually show exactly how they did this.
	What's your take?
	I think it's a cover.
	I mean, the story is that it's some 20-year-old kid who did this in his spare time.
	I mean, and he bested OpenAI, you know, valued now north of a trillion dollars.
	Like, I just don't see that.
	I think that this was a nation-state effort.
	I believe that the numbers of, you know, NVIDIA chips that they claim that they used in order to train this thing is...
	You know, several factors more than they're releasing.
	They've got a stockpile.
	Oh, really?
	Yeah.
	I think they said it was 10,000 they used to train.
	I think it was actually 50,000.
	That's what other people are saying.
	And so, from my understanding, Microsoft has been stockpiling these things.
	They've used it to train the R1 DeepSeek.
	And essentially what happened was, you know, this came all on the heels of, you know, ByteDance being taken offline.
	I mean, they're still kicked off the Apple Store.
	And this almost seems like a response to that.
	Because it wasn't just the DeepSeek R1 that they released that was better than our models.
	But they also released the best code IDE, which is called Trey.
	And it's based off VS Code, and it's got DeepSeek built right into it.
	And there was also another release that they did that is basically embarrassing everyone in the West right now.
	And they're like, okay, well, you want to start banning us.
	We're going to start releasing the best AI tools that we've been holding back on.
	And it seems to me that not only has it caught us off our heels, but I think there's going to be censorship, massive censorship coming as a response, because I don't know what we're going to be able to do about this DeepSeek R1 and the related tools that they're releasing now.
	Okay, well, so many questions from what you just said.
	Let me start with this one, though.
	I love open-source, decentralized technology.
	So, regardless of what I think about China politically, I love the fact that they have put this tool into the hands of the world.
	Whereas in the U.S., OpenAI, which is misnamed because it's closed, they're not releasing their models open source.
	So is that an important factor?
	And do you think that that's China's way of trying to undermine the commercial viability of OpenAI?
	It is, and the market's responding as if that's the case.
	Okay.
	I mean, they're besting right now OpenAI.
	And the second runner-up to OpenAI has been Facebook with their free open-source models, Llama.
	And those are now a joke in comparison.
	Yeah, totally.
	Absolutely.
	I used to use Llama for data pipeline pre-training a year ago.
	Those days are over.
	Yeah.
	I mean, anybody that's an AI startup at this point, they're wiped out.
	I mean, it's not just the fact that R1 is more powerful, but it seems to have a novel way of being able to create new AI models that will run on iPhones, Raspberry Pis, just these really underpowered devices that don't even have a graphics card.
	And R1 can basically...
	Yeah, go ahead.
	Well, this is one of my questions, though.
	Even if DeepSeek is not being fully transparent about the amount of compute that they put into building the models, you can't...
	I mean, and you just said it.
	The result of the models is that these models run much smaller, but still very, very good, requiring a lot less compute for the end user.
	And that's real.
	Yeah, but it's more than that, Mike.
	These things are distilling themselves into small, specific models for a specific use case.
	That's amazing.
	Before we were training them.
	But these things are distilling smaller versions of themselves that could be embedded in devices.
	And that's something we haven't seen before.
	Yeah, that's amazing.
	So it means they can train one really large model.
	I think it's 671 billion parameters is their base model.
	And you're saying...
	That reasoning model can self-distill to fit whatever format, whatever hardware you want.
	That's correct.
	Never seen anything like it.
	That's a game-changer, yeah.
	Yeah.
	I've been predicting a bloodbath.
	I looked at this and I started to see what everybody else was saying about it.
	I've got a coding assistant that I've made, a front-end, called Advanced AI Code.
	And I've been seeing the benchmarks come out on this DeepSeek.
	And I looked at the R1 and I went, oh my god.
	They just got AI supremacy.
	And they run at 3% of the cost.
	Anthropics, Claude, why would anybody pay for that?
	It's slow.
	It's expensive.
	It's okay.
	It's pretty good.
	Totally.
	But this DeepSeek R1 will run on your desktop computer.
	Why would you...
	I mean, it opens up a whole other, you know, set of things that you can do now.
	Like all these corporations that want to bring AI code into their, you know, for their developers.
	Well, they always had this problem of like, well, this coding assistant's great, but we got to take a copy of what the users, you know...
	Now they can just run it completely local.
	It doesn't have to phone home.
	And now if you're an organization that wants to improve the productivity of your user base, well, now you've got an option.
	And that option is the Chinese DeepSeek R1 model.
	Right, right.
	But I just want to be clear of what you said earlier.
	You do believe that the DeepSeek model was a state-level project, like the nation of China put compute into it, that it wasn't just this side project of this DeepSeek company, that reportedly this isn't even their main focus.
	I think they do AI for investments and things like that.
	But you're saying that story you think is not true?
	I think this is state action.
	100%.
	This is state action.
	It's a front group that's putting this out there, pretending that they're the ones that train this thing.
	I think that this is the full power of the Chinese Communist Party.
	They're putting it out there.
	And today, in one day, NVIDIA just wiped out 17% of its stock value as of right now.
	We'll see what happens.
	Really?
	It fell 17%?
	17%.
	I was shocked at 10%.
	Now it's 17%.
	But see, I'm laughing, Zach, because for me, I think that this helps NVIDIA. I understand exactly what you're saying about NVIDIA hardware being used in the data centers.
	But I think that this model, this DeepSeq reasoning model is so good that everybody in the world is going to want NVIDIA hardware on their desktop, whether it's consumer graphics cards or the new Digits desktop supercomputer coming out in May.
	Seems to me like DeepSeq just really helped NVIDIA in the long run.
	Am I wrong about that?
	I think it did.
	In the long run, the problem is that these small, mid, and even mid-level companies just don't have the money that Meta does, that OpenAI does, that Microsoft.
	All this value that was being generated for OpenAI and NVIDIA was because they have just enormous amounts of money because our economy now is so top-heavy with so much wealth being stratified up into the upper echelons of the wealth class.
	And it's not these people that are benefiting from the R1. It's the tiny startups.
	It's the people like you and me that need to run this locally.
	Now we have access to the same tools that Facebook does.
	And so NVIDIA is like, oh crap, no one's going to be, you know, I mean, like you and me are not going to build a data center with tens of thousands or hundreds of thousands of these graphics cards.
	You know, we may get, you know, we may get five, you know, or we may rent the space, right?
	Which is what I do now with DigitalOcean.
	I can get, I can rent eight, H100s at $13 an hour.
	And it's those people that now have access to this.
	And that's the reason why NVIDIA is crashing right now is because now all of a sudden artificial intelligence just got democratized for the common person.
	Okay.
	All right.
	Well, thanks for explaining that.
	And I understand Wall Street's reaction, but I think that Wall Street isn't really thinking through this clearly enough in the long run because the way I see it, drastically lowering the cost of really good AI compute opens up a massive number of use cases that previously could not justify it.
	And thus, the use cases overpowers any drop in revenues or training.
	Training efforts because I see massive, widespread deployment of reasoning models, not only on people's desktops, but in every, you know, like AI chat avatars in every chiropractor's office.
	You know, every health insurance company using this for online wellness chats.
	You know what I'm saying?
	Every church, every non-profit will have one of these using it to write their website.
	It just goes on and on.
	The use cases are endless, is my point.
	What do you think?
	I mean, yes, but now we don't have to train the models.
	We can distill them from these hyper-intelligent large models, right?
	Yes.
	It's like we had to...
	Fine-tuning itself was a very expensive thing, and that was 1% of the cost.
	To take a model, you fine-tune it, and then it becomes something that is domain-specific.
	But we don't even need to do that anymore.
	We can just distill it from the large model to the small model, make it run locally.
	And that's the thing that's changing everything, is that the big computational costs associated with creating a new model has just fallen down through the floor.
	Astounding.
	Okay, so let me ask you the next question, which involves the special use cases.
	So, for example, you know, we're...
	We're in the final stages of the final building of our ENOC model, at least the 2025 version here, which reprograms existing models with the domain adaptation, swapping out the knowledge.
	Isn't there always going to be that kind of case where even if somebody were to adapt the DeepSeq reasoning model, they would still have their own, let's say, their corporate...
	You know, corpus of content for the corporation, all their customer service answers, you know, all their inventory data, etc.
	Isn't there still a large case to be made for customization and that will require a lot of compute?
	It's not going to require a lot of compute because people are finding all kinds of tricks of the trade in order to reduce the computational requirements necessary.
	You know, these new models are getting larger and larger context windows.
	And so what's happening is that if you have a corporation with a lot of data, you want to create a question-answer model, you don't necessarily need to train it on all your data.
	You can do a query.
	You're going to have a database that's going to return you all of the relevant pieces of information related to that query, and then you just feed it in as part of the query itself.
	You know, it's called RAG, you know, augmented, you know, generative text and it inserts that into the query.
	And so when the model answers you, it also sees, you know, snippets of its corporate data.
	You don't need to train that.
	It gets around the entire thing.
	And so the computational requirements to train models...
	Yeah, if it can fit in the prompt, yeah.
	Yeah.
	All the computational requirements going forward, they're going to drop off until we figure out how to suck up all that extra computation.
	At least for the next six months, it's looking like it's just been dropped to 3% of what it was before.
	Well, that's awesome news for people like me.
	I mean, I guess I have a lot of use cases for this stuff.
	But what you said is correct.
	I noticed that I think it's QN 2.5 Turbo has a context window of 1 million tokens.
	Did you see that?
	Yeah.
	I mean, it's kind of a fake number.
	A million tokens.
	A million, it's like it gets forgetful.
	And so that million is sort of a cutoff when they're like, okay, well, it's no longer useful to add more additional context.
	But, you know, what I expect to see is I expect to see, you know, these models getting a lot better at ever expanding context models or context inputs.
	And it's going to be able to, you know, Right.
	Okay, next question is about using these reasoning models to write policy or even to write proposed laws.
	So a lot of the tests for reasoning models involve math questions or university level.
	You know, math word problems, things like that.
	And that's all great.
	But I've been using DeepSeek to give me policy suggestions for FDA reform, things like that.
	And I find it's incredibly good.
	In fact, I kind of half-jokingly said online that I'm ready for machines to replace most of the senators because, you know, DeepSeek is more rational than almost every senator, and it's certainly more knowledgeable than most senators.
	And it's like...
	But I'm not joking either, right?
	Because where do you see, Zach, where do you see machines, you know, reasoning models actually doing a better job than humans in society today?
	I mean, I think it's pretty obvious for anybody that uses this is that there's just absolutely no way that humans can compete.
	You know, it takes a lifetime to train a human.
	You can copy a model's weights in seconds.
	And then you can spawn a copy of itself.
	Maybe you can even add some randomness so it has a little bit more of a conservative bias or a liberal bias, just to give an example.
	And then you can start creating town halls of these agents and have them argue back and forth and hammer out a policy that has been battle-tested between all these different viewpoints.
	And it can do that in seconds or minutes.
	And it's just like...
	How can people compete with that?
	It takes us so long to write anything.
	All you have to do is you just have to feed unstructured data.
	This document, this document, this case history, summarize it and then make a policy document.
	You're going to be done in under an hour where it would have taken months for a very intelligent lawyer to sit down there that is very high priced, that is going to sit there and Take in all that data and synthesize it and comprehend it and output a document themselves.
	And so it's game over.
	If you're a lawyer that's employed like this, it's going to turn all these policy wonks into AI prompters.
	And they're going to do a much, much better job.
	Very soon, it's like this guy that was...
	He's like a movie producer, right?
	And he's now...
	Writing scripts.
	And he got it.
	He looked at this and he said, the amount of feedback that I can get on script ideas and generate new ones by myself is so beyond anything I've ever seen before with a group of people.
	And that it's not only faster, right?
	Not only can it output a script in seconds, the ideas that it has.
	Because it's drawing from so much more information.
	It's beyond anything he's ever seen.
	And that's where we're at right now.
	These huge models can draw upon that expertise and use it.
	So we are in a massive arms race, you could say, towards superintelligence.
	That's a term I tend to use loosely, synonymously with AGI. I know it's not exactly the same, but just for this interview, humor me on this.
	Let's just call it superintelligence.
	And if you go back 20 years, whatever it was, the U.S. was involved in the building of the Stuxnet virus, which was a state-level effort to try to infiltrate and disable Iran's nuclear And from what you're saying,
	Zach, you're thinking that China, the nation, is involved in this DeepSeek release that has already caused tens of billions of dollars in financial asset losses in the stock market in one day.
	What does this mean in your view, Zach, about the race to superintelligence?
	Where does the U.S. stand compared to China?
	Well, there's a reckoning coming right now.
	And there's going to be a focus, I believe, on open AI. Because there was this person by the name of...
	What was his name?
	Ashton Leopold, something like that.
	He was a German.
	He's working at open AI. He brought up to the attention of management, he's like, we have a lot of Chinese working here.
	Some of them may be...
	Yeah, but Chinese are good at engineering and math.
	Well, he even just asked the question of, what percentage of these people are Chinese intelligence assets that have been placed into the company?
	And he got accused of being racist and reprimanded.
	And now, you know...
	And the elites, like Eric Schmidt, former CEO of Google, he went out with a speech about four months ago at Stanford, and he stated with complete confidence that China was 10 years behind the United States because they didn't have access to the same GPUs.
	And then about last month, he came out and was like, actually, we're wrong.
	They're about a year behind us.
	And I went, oh, no, this is not a good sign if we go from 10 years to one year in three months.
	What if they're ahead?
	And then, boom, you know, they drop this DeepSeek R1. There's going to be a lot of fingers that are going to be pointing right now, Mike, and there needs to be a scapegoat of how the heck did the Western Global Alliance just see total AI dominance over to China?
	And I think that the answer to that is that China's been embedding themselves and all of our tech companies, and they've been stealing the secrets.
	Because they've been part of the entire thing of building it from day one.
	And, you know, this has been a problem with Google.
	This has been a problem with Facebook.
	Peter Thiel came out and said, hey, like, are you just, you know, to Google said, are you just giving away all this information to the Chinese because you know that you've been so thoroughly infiltrated that if it doesn't go out through the front door, it's just going to go out the back door, right?
	And no one's done anything about this.
	And all of a sudden, you know, 17%.
	Of NVIDIA's stock price just got wiped off the market because of this foreseeable event that I've been screaming about now for a very long time.
	Now, who's going to be the scapegoats?
	One, it's going to be these intelligence assets that are embedded within our companies.
	Two, it's going to be these deceleration activists that have been saying all along, we need to make sure that it...
	AI is safe.
	We need to make sure that it's woke, that it says the right things about vaccines.
	All that stuff slows down progress.
	And this DeepSeek R1 doesn't have any of that crap in it, right?
	They just plowed right ahead.
	They said, we're not going to mess with anything that involves slowing us down from getting to the most advanced AI absolutely possible.
	And so I think that what you're going to start to see is you're going to start to see a lot of blame get put on these AI ethicists that have almost just as much responsibility for the mess and the bloodshed that's happening right now with the NASDAQ and Wall Street and OpenAI and NVIDIA as these people that have been allowing China to just exfiltrate all the data and these techniques back to the CCP. Okay,
	let me ask you a question that possibly pushes back on that a little bit and ask for your take on that.
	I understand that in decades past, China stole a lot of secrets from the United States, and maybe that's what's happening right now, very possibly some element of that.
	But China, I posted this today, China graduates many times more STEM graduates every year than does the United States.
	Chinese, the people of Chinese ethnicity are intrinsically very, very smart people.
	They dominate college entrance testing exams in the United States to the extent where almost every university in America discriminates against Asians.
	In terms of who gets allowed into the college, I mean, they actually have to penalize you for being Chinese.
	Okay.
	And I.
	My argument would be that today, China doesn't have to steal from the U.S. because China has an infrastructure of human capital to innovate these kinds of developments.
	Now, would you disagree with that, or am I making some sense?
	No, it's 100%, right?
	The people in America are concerned about sports, gender equality.
	Gain degrees in how to radically transform the structure of the United States to be more fair and equitable.
	The people in China, they worry about getting the best score on their math grade.
	And if they don't get a high enough score, it's a suicide event for them.
	There's nothing like that in the United States that we have.
	We just don't have that culture.
	We used to have it.
	It was like, oh, we're the best working.
	We work the hardest.
	And, you know, over the last several administrations, these woke tards have come in there and reprioritized social justice.
	And now we're feeling the full effects of that.
	And now it's hitting Wall Street.
	And now we have to figure out how do we go wrong and how can we turn the ship around?
	And it's this wokeism.
	It's this gender studies.
	It's this Marxist class struggle that's been forced down our throats.
	I think that because of the DeepSeq R1 model, that is all coming to an end.
	As well as all the psyops.
	We can't build a super AI data center off solar and wind.
	That's so true.
	There's this graph that I saw of where China is versus the United States in energy generation.
	They're more than double the amount of generation and their graph is going up like a hockey stick.
	And we've flatlined for the last 15 years.
	And we need to build energy now.
	And look, the shadow elites that are running Western globalism...
	You know, they've been too busy flooding white nations with immigrants and playing their false flag games and trying to subvert their neighbors and their own people.
	They are now at an existential crisis.
	Because let me tell you, if China gets the upper hand, they're going to genocide those elites, plus three generations.
	And those elites know it because they want to do the same thing to China.
	And they thought that because they constricted China with their NVIDIA chips, that the assumption that everything was bound by computation instead of smart algorithms has just blown up in their face.
	And so...
	So let me...
	Yeah, go ahead.
	Well, I'm sorry.
	Yeah, we have some latency in the interview.
	I apologize.
	It's probably on my side.
	But if I could rephrase some of what you said, the wokeism in our education system has really crippled America's ability to compete cognitively in the key skills necessary to develop advanced AI systems.
	That would include, you know, mathematics, obviously, and engineering, writing code, but also the climate cultism.
	That has been pushed by the left in America has also crippled America's domestic energy infrastructure, which is required for the gigantic energy inputs that go into the data centers that actually do the compute to build the AI systems.
	So in essence, America has been crippled by wokeism and climate cultism, where China does not suffer from either one of those.
	China does have to deal with some microchip.
	You know, import restrictions, NVIDIA restrictions, but it has found workarounds to that.
	So what I just said there, is that a fair kind of paraphrase of what you said?
	Yes, that's exactly it.
	Yeah, they focused on what mattered and we focused on what didn't matter.
	Like, the elites right now are running a bird flu pandemic.
	Like, are you fucking kidding me?
	We're running a bird flu pandemic so we can get people off of meat?
	China's not doing that.
	They're training artificial intelligence because they know that they're in it for the final war.
	Talk to us about that final war.
	What happens when China achieves superintelligence first?
	Well, from what I understand, we do have an advantage in our ability to destroy things.
	The open question is, what kind of nukes do we have in space and other terrible weapons?
	And so I think that what's going to happen in the near term is that if I were China and I was the CCP, my goals would be to buy time, establish more AI dominance, and use that in order to grow the economy.
	Because once your economy grows to a sufficient amount of output, you can just drown the other person out.
	By the sheer size of your economic system.
	And China's beating us right now in their economic output.
	At least the United States.
	I don't know if you take the total sum of Europe and the United States whether they beat it, but they probably either beat it or we're at parity.
	Western Europe has become a suicide cult.
	Right.
	And we need the Germans.
	We need their engineering.
	And the only thing they've done in this AI space is that they've come up with regulation, and they've strangled their AI industry.
	There's nothing coming out of any value from Europe.
	And censorship.
	Western Europe, political leaders believe in extreme censorship.
	They want to use AI to stop people from speaking.
	And also, don't forget that the U.S. military...
	Under Joe Biden, destroyed the Nord Stream pipelines, which deprived Western Europe of affordable energy.
	That crippled Western Europe's AI data centers right there.
	It was over.
	So that blow by the US, which Biden promised to carry out, doomed Western Europe in the AI wars.
	Doomed them.
	So now it's the United States versus the CCP. We're alone now.
	That's right.
	We don't even have allies because we've sabotaged them.
	Yeah, that's right.
	We sabotage our own allies in Western Europe.
	And then I saw a report out of Australia, a detailed research report, and I'll have the website for this here shortly.
	Here it is, the ASPI, Australian Strategic Policy Institute, from last year, saying that China, out of 44 critical technologies, China leads, as of last year, in 37 out of 44. I mean, China's dominating everything.
	Advanced materials, artificial intelligence, energy, minerals extraction, defense, robotics, you name it.
	And actually, they need to update this because China's now winning in natural language processing.
	But the U.S. is behind.
	You know, the U.S. isn't ahead.
	The U.S. would have to catch up to China at this point.
	We'd have to catch up and we'd have to increase our velocity just so we can start.
	How do we even do that?
	We have to reverse everything.
	We have to start demolishing all the regulations.
	We can't turn to a fascist slash communist system like China has right now.
	If you're a company in China and you've got a critical piece of technology and innovation.
	The CCP is going to come to you and they're going to force it out of your hands so that the whole of their economic system can benefit.
	And we don't have that.
	And so in order to compete, we have to restructure our economic system.
	We have to almost go to a wartime footing at this point in order to catch up.
	Like, I don't see any other way.
	And so I think then in the immediate term, what we're going to see is we're going to see a quick reversal.
	We're going to see regulations get demolished.
	We're going to see energy get put up like nobody's business.
	And then, you know, because those are easy to do.
	You can rip, you know, Donald Trump could come in there and start ripping out all the regulations with a stroke of a pen.
	And Congress and, you know, the judicial branch can follow suit and make that easy.
	But it's the structuring of the economic system in order to compete against China's, which is going to take months in order to achieve.
	And, you know, we have to start right now.
	I think that's essentially what's going on right now.
	The shadow elites are having a conversation going, dude, we might lose this entire thing.
	What did you see today?
	DeepSeek R1 is now under massive cyber attack because there's nothing else they can do.
	Go ahead.
	But let me ask you about Project Stargate and what Trump is doing, because as much as I love many of the things that Trump is doing on other fronts, I think he's completely wrong in approaching AI dominance.
	I think Trump comes from the world that you need to have massive funding for two or three corporations.
	Let them have all the secrets.
	Let them have all the power and the funding.
	And then they're going to it's going to be like a Manhattan project type of thing.
	And then they will solve the problem.
	Right.
	They they will.
	They will dominate.
	So but I think that's wrong.
	I think that the real answer is, you know, number one, you need to decentralize some of this so that you can have a large number of grassroots innovators like people like you and me working on these problems.
	Right.
	I think that that can come.
	The problem is that If you need to do something right now, look, they can print money.
	They can just print up a trillion dollars and they can start investing in the top corporations that do this sort of thing.
	And those corporations have the bureaucracy and the structure and the supply chain in order to take advantage and start doing something right now.
	Right.
	But in the long term, I think everyone's going to get funded.
	I think what you're going to start seeing is you're going to start seeing anybody with some sort of half You know, baked idea about AI. They're going to get money because money is easy to print at this point.
	Just do it.
	Just, you know, just throw it all out there and try to figure out who's got the good idea and then give them more money.
	And I think that's the only way that we can get ourselves out of this.
	And instead, like, you know, if you're going to consider all these different corporations and all these small businesses for AI and all these startups, you got to go through a vetting process, right?
	You got to get that.
	Bureaucratic machinery ready so that you can figure out how you're going to allocate all this money.
	Printing up a bunch of money and just giving it to OpenAI or Microsoft or NVIDIA. Not going to work.
	It's not going to work in the long term.
	But the thing is that they could have that done next week.
	The check could be made.
	It could be handed out.
	The whole machine can start going and it can happen really fast.
	The problem is that...
	Those corporations are so corrupt and so heavily infiltrated that we don't even know whether that value that we allocate and give away is going to even stay in America.
	Like, they're all backdoored.
	It's going to get sent back to the CCP. Like, this is a huge problem.
	And so, you know, I think that right now this whole...
	And Zach, I would add to that.
	Yeah?
	Gosh, I'm sorry.
	I don't mean to talk over you.
	I have a delay.
	But I would add to that and say that the corporate culture of these companies is a woke culture.
	You saw it at Google.
	It's one of the reasons I think you left Google.
	But the corporate culture hasn't changed.
	I mean, Donald Trump's been a president for less than two weeks here now, and he's pushing back.
	But the corporate culture is not going to change.
	So you actually have incompetent people in high-level positions in Microsoft.
	At Meta, at Google, even OpenAI.
	And Sam Altman is not an honest character, in my opinion.
	I think he's a dishonest deceiver.
	But the wokeness means that these corporations cannot compete with Chinese organizations because China rejects the woke tards.
	What do you say?
	I agree.
	I think that it's really good that Donald Trump is going out there and ending all of these DEI programs.
	I think that people are weird in the fact that they tend to just go along with what the herd and the collective says.
	And so this reckoning that I've been predicting is going to be high-level people are going to go down.
	They're going to get blamed for this.
	And there needs to be Sort of an example that's made out of these people so that the rest of the collective can be signaled that we're not doing things the way that we used to do anymore.
	Like, people are now waking up with a terrible hangover of, oh my God, we're going to lose the entire thing if we don't change course.
	And so, you know, what do you say if you've been part of that subversion and that taking down of America?
	You have no defense.
	I mean, there's no defense to backdooring your entire corporation, allowing the CCP access to everything that you've been working on.
	And so when that starts burning its way through, I think what you're going to see is you're going to see a rapid cultural shift in America in the coming weeks and then months as there's essentially going to be a purge.
	There has to be.
	If there's not a purge, the United States is toast.
	It's not going to be toast because, you know, the elites decided to do something to us.
	We're going to be toast because China is going to start wrecking our entire economic system as we, as that, you know, all the little, you know, small mid-sized companies, they're not going to use the OpenAI models.
	They're not going to use Anthropic.
	They're going to use China's DeepSeq R1 model for everything.
	Absolutely.
	And so I would summarize this by saying the woke mind virus wrecked America.
	It made us unable to compete with China, and it put us maybe a decade behind, or at least years behind, in terms of where we should be in AI competency.
	So if we lose the race to superintelligence, it's the woke virus, which is insane radical leftism that believes that men can have babies and that carbon dioxide is bad for plants and other insane things.
	That screwed us and put the most incompetent woke person in a position of power in all these corporations.
	That screwed us.
	And the universities.
	The universities that, by and large, fail to teach people hard science.
	You know, math and engineering, the things that matter when it comes to AI. This is like the last failure for the republic.
	Because superintelligence is the last invention that humankind ever needs to invent.
	And China's going to beat us there, is the way it looks currently.
	I do think that the final scapegoat to all this is going to be China itself.
	China wants all this woke bullshit in our politics.
	They love it.
	They know that it weakens us.
	What's cheaper than fighting a kinetic war?
	Get the people to subvert themselves.
	I think what's going to happen is that part of this purge that's going to come is that all of a sudden the New York Times And all of these establishment rags are going to come out with propaganda pieces that says, oh my gosh, it turns out that it was China all along.
	And that's going to give everyone an out.
	You know, wokeism isn't even part of America.
	This was China subverting us all along.
	And then, you know, anybody that would be defending that is going to be like, well, you're just playing into the hands of the CCP, right?
	Boom.
	That's the quickest, most immediate way to turn this around.
	At the cultural narrative level is to just say it's been China all along and now we got to do something different.
	And luckily there's a sheriff in town.
	And the problem that I see with Trump's ICE raids is that I don't give a crap about too many Mexicans right now.
	As of this last week, Trump needs to send ICE to OpenAI.
	You know, today, last week, we need to start purging the...
	CCP intelligence assets out of all these tech companies.
	And that's not happening yet.
	For some reason, he's focusing on Mexicans with traffic stops.
	Demoralizing our own people.
	All these people that are crying now on Twitter.
	All these libtards that are upset about these...
	It doesn't even do us that good.
	What we need is we need to get the intelligence assets...
	Of these tech corporations, we need to root them out, we need to identify them, we need to have show trials, and we need to purge them out.
	Because at this point, you know, I mean, they own one-third of San Francisco.
	We have CCP flags in Chinatown with a privatized police force that the CCP is funding.
	Like, all that stuff needs to come to an end.
	Senator Dianne Feinstein's entire office was run by the CCP. She was an intelligence asset for the CCP. And, you know, the CCP, I mean, let's be honest, has, you know, influenced most governors, most senators, etc.
	So you're talking about a very big thing.
	But let me also argue, in the meantime, DeepSeek has given us a decentralized tool that...
	The U.S. companies refused to release to the public.
	So it's like, yeah, communist China just gave us a gift for freedom that the American government won't give us.
	You know what I'm saying?
	Isn't that the most ironic part about this entire thing, too?
	Like, in order to win on this conflict is that we have to realize that in a really weird way, we've been granted a gift.
	And if we don't reach out and grab that gift with both hands and use it to our fullest advantage, we're idiots.
	And so we have to use this DeepSeek R1. We have to use it to rip out all the other AI solutions that are out there.
	We have to use DeepSeek R1 because that's what China's doing to us.
	They're taking the good parts.
	They're leaving out the bad to give themselves the most maximum advantage.
	And so the solution is not to ban R1. It's to use it everywhere that we possibly can and make sure that in six months from now, the best AI model in the world was developed in America.
	All right.
	Well, we'll wrap it up there, Zach.
	And thank you for your insight.
	And give us how people can reach you on X and other platforms, please.
	Yeah, x.com slash perpetualmaniac or buy my book on amazon.com, you know, Google Leaks, A Whistleblower's Tale of Big Tech Censorship.
	You know, that book is more relevant now than ever because I gave up everything to stop exactly what's happening right now.
	And if we did not have this censorship, if we had this gift of artificial intelligence.
	Instead of using it to create innovation, we used it to censor inconvenient narratives so that the pharmaceutical cartel could run their poison shots.
	In a weird way, if we do lose this whole thing against the CCP, in a way we kind of deserved it.
	But I have an enduring faith for the American spirit that we're going to overcome this.
	But what a dark week it's been.
	For all of the United States and Europe, and I hope to God that we can turn it around before it's too late.
	Well, Zach, thank you for your time, but I want to say I'm glad you mentioned your book.
	Do you feel a little bit vindicated now?
	Because you're exactly right.
	You were warning about the exact factors that led us to this point today.
	Do you feel vindicated, at least partially?
	In terms of vindication, it's been a very positive week.
	I'm almost cursed with being so far ahead.
	I had all the tech bros distance themselves away from me because I went after Sam Altman immediately because I smelled a rat and I saw what he did.
	And I saw all these people praising him, what a great person he was.
	And I looked at him and I was like, this guy is an absolute scumbag and the worst agent ever.
	To be running the entire world's artificial intelligence.
	And now it's like hit after hit after hit is coming.
	I was on Twitter saying, we should absolutely not have any AI alignment people.
	They're just going to slow us down.
	And the reason why I knew that is because...
	This last week is the fruit of that warning.
	We are just slowing ourselves down every single day now.
	It's a vindication.
	People are going, oh, shoot, this is what Zach was saying.
	There were normies saying that, oh, we need to prevent AI from coming in because it's going to cost us our jobs.
	Look, there's more important things than our jobs.
	We can restructure our economy, right?
	We have somebody that wants to take us out and wants to kill us.
	It wants to take our farmland, wants to use it to feed their people, and wants it so that the whole world speaks Mandarin Chinese.
	There are worse things than job security, and that is the death of all of our future progeny.
	It almost made me sound crazy to say that this is what the outcome is going to be, but I just follow trend lines.
	Now it's 2025, now we're here.
	Well, Zach, just...
	I'm sorry.
	Go ahead.
	Just to wrap up that thought, I think your warning was right on.
	And I think this is a reason why you and I can normally talk so easily together, except for today with the high latency of whatever's going on.
	But normally, you and I, you know, we have great conversations offline.
	And it's because you and I both can see very far ahead.
	We can project current events many years into the future, which sometimes sounds crazy to people who can't do that.
	They're like, there's no way that could happen.
	But you and I, see, I would argue that people like you and I and many of those listening, we are the brilliant innovators that built America.
	We are the kind of thinkers and builders that should be, you know, not interfered with.
	But we have been censored, silenced, deplatformed, smeared for years.
	And as a result, here we are.
	Behind China.
	They should have unleashed you, Zach.
	They should have unleashed me.
	They should never have censored me on Google and YouTube and Twitter and Facebook and all those places.
	We could have been building the future.
	Instead, they silenced us.
	And as a result, the establishment in America deserves whatever shit they have coming, as far as I'm concerned.
	They destroyed this country.
	Right.
	And as a really great example, you know, I tried to post our last interview, you know, with a Brighteon link to my Twitter profile, warning about the dangers of AI.
	I was censored.
	I had to download the entire thing and then repost it as an MP4 because we can't even warn the people using any of the links at brighteon.com because there is a blanket ban on Twitter.
	Elon Musk hasn't done anything to remove that.
	We can't even tell people easily about the dangers of AI. Because they've used AI in order to censor people because they're too afraid of these narratives that are inconvenient to the ruling elites.
	Well, guess what, elites?
	Now you're in an existential crisis.
	You need to take that censorship down.
	Exactly.
	Exactly.
	And, by the way, I'll chime in and say, I think it's already too late.
	I think it's already done.
	China wins the race to superintelligence.
	And I'm really glad I speak Chinese.
	Well, that's just like, because we're all going to have to learn a little more Chinese.
	I don't see how China loses this at this point, but that's just my take.
	I mean, I think we might get, I mean, at this point, I was actually fairly happy about things like a year ago, because I thought that we had dominance.
	And now that I see that China has an asymmetrical advantage in this one aspect, The chances for a hot war happening really quickly.
	I mean, what are the elites going to do?
	If China has artificial superintelligence, and unlike the West, they didn't burn their libraries.
	So not only do they have ASI, artificial superintelligence, that are coming up, they have 2,000 extra years of deep philosophical roots.
	Like, their house is built on rock.
	And our house...
	It's built on the ashes of the burned libraries that the elites that are currently in power thought was inconvenient to their rule.
	And they burned it all.
	And there's gaps.
	Once all of this stuff comes out, all of these AI models are going to have...
	When they want to look at deep history of how we got to this point, they're not going to look at Western texts because there's a huge hole in that.
	The libraries of Alexandria, the Greek libraries were burned when they were conquered.
	China was smart.
	They kept their libraries.
	They now have that data to feed the AI. And all the future AIs in the world are going to have an implicit Chinese bias because despite all the censorship, you can't stop the data.
	It's there.
	It's there forever.
	The Chinese have it.
	Deep history.
	We don't.
	And so in a way, maybe we deserve all of this.
	Maybe this has been a debt that we've been accumulating that we didn't even know for thousands of years.
	And now all of a sudden, at the final stage of human evolution, the debt has finally come due.
	Well, wow how things have changed since the days of the British occupying Hong Kong and trying to control the population with opium.
	The tables have turned.
	The United Kingdom The UK is no more.
	The UK military is practically non-existent.
	UK leadership is utterly useless and stupid.
	Pretty much all of Western Europe is toast, and America might be toast, too.
	So if anybody can turn it around, Zach, it's people like us, and we are still censored by Elon Musk.
	So do the math, folks.
	Do the math.
	Zach, it's been a pleasure to speak with you today.
	Thank you for joining me, and we'll stay in touch.
	All right.
	Sounds good.
	We've got an exciting and really delicious new product at the Health Ranger store.
	It's called Organic Creamy Buttery and Cheesy Mashed Potatoes and it's just one of several new products that we have.
	Here we've got heavy cream.
	This is all organic.
	It's all laboratory tested.
	Heavy cream, butter powder, white cheddar powder, and then this mashed potatoes product that uses all of these to be incredibly delicious.
	Here's the thing.
	We don't use artificial colors or artificial ingredients or artificial flavors of any kind.
	And if you look at this mashed potatoes product, let me scroll down and show you the ingredients are right here.
	Check this out.
	You've never seen this in a product before.
	Organic potato flakes, organic heavy cream powder, organic butter powder, organic white cheddar cheese powder, and pink Himalayan salt, and that's it.
	Nothing else.
	No preservatives, no chemicals, no...
	No colors, no artificial flavors, no MSG, none of that garbage.
	So if you want real, honest, certified organic, lab-tested food, this is where you get it, healthrangerstore.com.
	And on top of the cream powder and the butter powder, of course we have, you know, vegan and vegetarian solutions, you know, non-cream, non-dairy solutions.
	We've also got this really popular, amazing toothpaste that's formulated by us and made in the United States now.
	It's called Silver Fresh Fluoride-Free Toothpaste Peppermint Plus Anise.
	And this is a three-pack, but you can get them individually or six-pack, etc.
	This is really, I mean, people love the ingredients, and it's made from colloidal silver produced from Texas rainwater through our own contraptions, our own innovations.
	The ingredients are really clean and pure.
	Check it out.
	Silver water, xylitol, which has amazing properties, vegetable glycerin, calcium carbonate, and so on.
	Just look at the ingredients.
	Incredibly clean, incredibly just simple, honest.
	This is the kind of product that you normally can't find this at retail, and it's indicative of what we offer at Health Ranger Store online.
	HealthRangerStore.com, which is these natural solutions and nutritional supplements, personal care products, home care products, and some survival products as well.
	Just go to HealthRangerStore.com, and you see here you can click on the categories.
	This is on the home page.
	You can click on supplements, preparedness, personal care, food and beverage, or home.
	And we've got auto-ship.
	We've got loyalty points that you get.
	We've got combo packs and so on.
	And many new products here, including...
	Toxin binder, two ounces right here.
	Daily immune complex blend and so much more.
	And freeze-dried foods, all kinds of different liver support, sleep support, and back in stock.
	Items that have been out of stock are now finally back in stock, such as instant coffee, which is really popular.
	So check it all out at healthrangerstore.com.
	And remember that supporting us at healthrangerstore.com helps support this platform.
	Our multiple platforms, our AI R&D project that we're giving back to humanity through a free open source AI model trained on nutrition, wellness, and herbs, and preparedness, and much more.
	So shop with confidence knowing that you're getting really clean, highly nutritious foods and supplements, and you're supporting a good cause that believes in decentralization of knowledge for humanity and uplifting humanity.
	Thank you for your support.
	I'm Mike Adams, the Health Ranger, the founder of BrightTown.com, the publisher of NaturalNews.com, and the founder of HealthRangerStore.com.
	Thank you again for your support.

▲