All Episodes
Previous Next
April 6, 2024 - Health Ranger - Mike Adams
33:24
EMERGENCY COMMS: Brighteon.io application launched for decentralized, P2P...
| Copy link to current segment

Time Text
Alright, welcome to this Emergency Comms podcast for Saturday, April 6th, 2024.
Mike Adams here.
And of course, the solar eclipse is only two days away.
We don't know what's going to happen during the solar eclipse, but of course, all these counties and cities and states are saying, you know, cell towers might go down.
Hospitals might be full.
Food might be gone from the shelves.
I don't know what they think is going to happen.
But it's not normal to say those things from, you know, government authorities just because the shadow of the moon is sweeping across North America.
I mean, we've had eclipses before.
It didn't cause the end of, you know, society or anything.
But, hey, gold is $2,330.
So clearly a lot of people think that...
Something's going to happen because silver is $27.50.
I haven't seen it at that price for quite some time.
But anyway, this isn't about gold and silver.
This is about emergency comms.
So here's the deal.
In Canada, yesterday, there was a loss of DNS services, domain name resolution for web domains.
Some outage, pretty big outage in one region.
I think it was in Eastern Canada, but I'm not certain where it was.
It is suspected.
I had one very competent tech person tell me that they thought that it might be a test for the internet kill switch.
Now, the internet kill switch, you may know, is designed to cut off the people from websites that they don't want you to reach.
And the way the Internet kill switch works is by actually keeping TCPIP up and running.
In other words, the infrastructure of data moving across fiber optics and everything, that will still be functioning because, of course, the government needs that.
But they will block the domain names of the domains they don't want you to visit, which could be everything except their official domains.
So that's how they can take...
Entire websites offline.
And that would, of course, take down Natural News.
It would take down Brighteon.
But, you know, they would probably have, well, Google would still function because, you know, they want you to have all the disinformation that Google is feeding you.
So we have a very important workaround.
And it's critical that you know about this and that you take action here before Monday.
And this is a pretty big deal.
So, take some notes here.
We have just approved the app for Brighteon.io.
Now, Brighteon.io, you may not have even heard of it.
You probably have heard of Bastion, because I've mentioned that before.
The Bastion peer-to-peer decentralized free speech platform.
It's kind of like social media, but there are no central servers whatsoever.
It's non-censorable.
It's non-confiscatable.
It cannot be stopped.
And there are no servers anywhere that can stop it.
And it doesn't even need domain names when you use the Bastion app.
Well, we have now a Briteon.io app for Android, iPhones, Mac, Windows, Linux.
All working.
And we have Brighteon.io website where you can sign up and grab your user account, you know, reserve your name there.
And here's the cool thing.
Brighteon.io is built on the Bastion platform, and it's the same universe of content.
So if you're already signed up on Bastion, your username will automatically be the same.
And it will be shared across the Brighteon.io universe.
In fact, you can go to a website, you know, a browser right now and just type in Brighteon.io and that'll bring it up.
And you can log in there if you already have a Bastion account.
But here's the cool thing.
If you go to Brighteon.io slash applications, plural, then that brings up the page of the new apps that we have available as of just yesterday.
Briteon.io slash applications.
And then you can download, like I said, Android.
There's an APK for de-Google Android users.
There's Mac OS, there's Linux, and there's Windows.
All right, how cool is that?
So you download this app, you run it, and even if the government takes down DNS, even if they take all the websites offline...
because it doesn't use domain names.
It uses peer-to-peer decentralized traffic on an internal blockchain, by the way.
And it has its own reputation currency called Pcoin, which is short for pocket coin, by the way.
And the way they have it set up, Daniel Sachkov is the creator of this.
I've interviewed him a couple of times.
I was so intrigued by his technology that we decided to build Brighteon on top of it.
But the way he describes it is that They use Pcoin as a reputation marker so that people don't come in and spam the blockchain with a zillion videos, for example.
To post videos into the Brighteon.io or Bastion ecosystem, you have to have a certain amount of Pcoin in your wallet.
And that just prevents robots from coming in and spamming up the place.
It's kind of like when Elon Musk said, hey, we want to charge everybody a dollar a month.
Just to stop the spam bots, right?
That's the only purpose.
I think that's actually a good idea.
But anyway, Pcoin is that purpose on the Brighteon.io or Bastion.com universe here.
So, I hope I'm not getting too technical here, but what I really wanted to explain is that if crazy stuff happens on Monday, this is my emergency comms channel.
I will be posting audio to Brighteon.io, which is the Brighteon.io app also.
I'll be posting audio updates.
I'll be posting text updates, maybe video.
Who knows?
Depends on how crazy it gets.
And I'm in touch with former military intelligence, you know, retired guys who were high level.
I've interviewed former CIA analysts like Larry Johnson, and I know a lot of people who have been in high places and so on.
And so if they get me any intel on Monday, who knows, then I will be able to pass that on to you using this emergency comms channel, which is, again, brighteon.io.com.
Or download the apps at brighteon.io slash applications.
Now, it is critical that you do this before Monday, I believe, just in case there's a takedown of domain name servers, which would be the internet kill switch.
Then you will still be able to get information through this app.
So, put it on your priorities list for this weekend.
Like, download this app.
Create your username.
Oh, by the way, when you create your username, you will get...
What's it called here?
You will get something called a private key.
Yes.
This private key is basically your identity and password all together.
If you lose this private key, you will lose access to your account and whatever pcoin you have in that account, if any.
So think of this private key as your seed phrase.
If you're familiar with crypto wallets, you don't want to lose your seed phrase because then you're screwed, right?
Then you've lost your wallet.
You know, I mean, if it crashes or whatever.
So make sure that when you sign up and it gives you a private key, that you take that private key and you make a copy of And you put it in a safe place, but a protected place.
It's best to put it on a thumb drive and keep it offline somewhere.
Because if somebody else gets that private key, then they can log in and they can impersonate you or edit all your posts.
Put the word not in front of everything you said.
Things like that.
So definitely protect this login now.
Remember that once you have that private key squared away, you know, you can use that on any browser anywhere to log in if domain names still work.
But if domain names don't work, then you need the app, okay?
But you can also download the app, the installation file, and you can put that on a thumb drive.
And then you can install that app anywhere, okay?
You know, on any system where it's compatible to run it.
And then you could be back up and running and getting information through the blockchain.
And remember that the Bastion ecosystem is entirely powered by blockchain.
And that's why it's uncensorable.
So even the founders of it, Daniel is the key founder, but he's got an incredible team.
Even they can't censor anybody.
You might wonder, well, what about crazy people that come on and post pro-demonic crap?
Well, guess what?
They're not censored.
You can censor them yourself if you see some crazy post coming at you and you don't like it.
You just block this user.
Or if there's some crazy illegal content, a bunch of users can get together and they can...
Vote a user to be kicked off, but there's some kind of a consensus trial system.
A certain number of people have to be jurists.
I'm not sure exactly how that mechanism works, but it's pretty hard to get booted off of Bastion.
You have to be doing something blatantly illegal like posting, shall we say, cheese pizza type of videos, as they say.
And by that, I don't mean cheese pizza.
Yeah.
Pizza parlor.
You know what I mean.
Because that stuff is not welcome anywhere in the civilized world.
Okay.
So, bottom line.
Emergency comms.
Boom.
This is how you do it.
Okay?
Brighttown.io slash applications.
Download that and install it.
Now, other news.
Very important.
We released yesterday the first distribution of our pre-release of the NEO large language model.
It's actually NEO Phi 2, Epoch 2.
And so it's a pre-release, very early model.
I call it version 0.1.
It's not even 1.0.
It's 0.1.
I should have called it 0.01, but I didn't want to have that many zeros in there.
I didn't actually want to release this model yet because I know we can do better because we've got a lot more data.
We've got whole racks of servers and systems crunching on data pipelines and then doing fine-tuning training.
But I wanted to meet the deadline before the solar eclipse and just get you the model that we had, even though it's not the best model yet.
It's still pretty good.
I was asking it about how evil is the CDC and it was telling me about the CDC pushing eugenics and illegal experiments on people and things like that.
So the model is not bad.
It's going to get better.
So next week we're going to be releasing Neo Phi 2 Epoch 3.
And then pretty soon we're going to be releasing Neo Dolphin Mistral 7B Epoch 3.
I know it's hard to keep track of all this.
That's a 7 billion parameter model.
They're all free to download, free to use, non-commercial, no advertising, no surveillance, no monitoring.
It doesn't run in the cloud.
It runs on your local computer.
You will need probably at least 8 gigabytes of RAM to run the Neo Phi 2 model.
You'll probably need 16 gigabytes to run the Neo Dolphin Meestrel 7B model.
So...
Just make sure that you've got a good enough computer to run these things.
And also, we've had some people that are trying to run this on Macs, and they're running into some trouble.
I admit, we don't yet have good documentation on Macintosh applications, but I do know there's an app called Ollama, O-L-L-A-M-A, not Obama, but Ollama, We should call it Obama's llamas.
And it works on Mac.
And there are several people who are getting it to work on Ollama.
In fact, one guy posted some command line stuff that we may share.
There's a little bit of a learning curve, a little bit technical.
If you're frustrated, just check back.
Go to brighttown.ai and click on Tutorials.
We do have how-to tutorials there with pictures or screenshots to walk you through how to run these models on different pieces of software like LM Studio, GPT for All.
And then also we have a self-executable file that works on Windows.
It's an EXE that you download.
You just double-click it.
And then it launches itself in your browser locally.
And that's real easy to use because you don't have to install any apps.
But that version, being that it had to fit inside a four gigabyte file, it's highly quantized, which means it has much lower fidelity compared to the other models.
But what that means is the answers aren't as good.
It's just not as good.
It's just too small.
So keep all that in mind when you want to run these models and query them.
If you have a 16-gigabyte RAM computer, that's going to be better.
And, by the way, if you have a NVIDIA GPU that has at least, let's say, what would you need?
You'd need about 6 gigabytes or better, probably, to run these models on your NVIDIA GPU. Then the speed is going to be way faster.
So some of you gamers out there, you might have like an NVIDIA, one of the RTX 3090 or a 4090 or something like that, or a 2090.
As long as it's got six gigs of RAM or better, that's video RAM, then you're going to be good.
It's going to run.
It'll be faster.
And I think you'll be very pleased with the results.
Now...
We are not only doing more waves of training, but we are vastly expanding our data set.
So I have literally, I have a thousand books that have already been scanned and converted into text.
And we're doing data pipeline processing on these thousand books, like a hundred at a time.
It's extensive, let me tell you.
The processing power required to do this is...
Well, I think it's massive.
Maybe companies like Google or Meta or Microsoft, they don't think it's massive.
They're like, yeah, we've got a trillion dollars.
We can just buy a warehouse full of servers.
Well, I can't do that.
We earn all the money.
Nobody just hands it to us.
It's thanks to people like you listening here.
You support us at healthrangerstore.com.
And then we have to make enough money to plow it into projects like this.
So I don't have unlimited funds.
So we have spent a lot and we're making a lot of progress.
It's still going to take quite a bit of time.
But let me tell you, when we finally get to version 1.0, Which might be a few months away, actually.
I mean, for example, we will release these models in version 0.2 and then 0.3 and 0.4 and so on.
This represents an expansion of our data set.
By the time we get to version 1.0, which might be this summer, if the world is still here after Monday, we'll see, it's going to be a very nice data set.
I would imagine that by the summer we'll have a thousand books in this.
And probably tens of thousands of video transcripts that we are crunching on.
The transcription process is actually pretty straightforward.
It's the processing after transcription that is the hard part.
You can't just take spoken word transcriptions and just train the models on them.
It turns out, because it makes the models go crazy insane.
The models start talking like...
Someone who's stuttering in their interviews.
Anyway, it doesn't work.
There's a lot of steps to go through.
But our data sets will continue to expand.
0.2 will be the next one, obviously.
And then we'll have multiple epochs of training, bigger data sets, and bigger models.
We may train on a 13 billion parameter model in the coming couple of months.
We'll see.
But the next models we'll be releasing are the 7 billion parameter models on Mistral.
So that's mistral.ai.
I think that's their website.
They've released some open source models that are 7 billion parameters, and we're already through a couple of epochs of training on those models.
So if you want to play with these models, go ahead and download them.
You can check out the tutorials on brighteon.ai.
You can run them locally.
I do want to give you a bit of advice here.
All of these models of LLMs, as they're called, they all work the same.
They give you whatever you ask for.
So if you ask this model, like, tell me all the things that are great and awesome about vaccines, then they will.
And it'll sound like a CDC press release.
Oh, vaccines are awesome, you know?
They'll just give you 10 bullet points on why vaccines are awesome.
But if you ask the same model, hey, tell me, give me five reasons why vaccines are harmful and dangerous, Well, it will do that, too.
It will give you what you ask for.
It's not like a person who will just sit there and argue with you.
You ask a person, you know, like, tell me five ways vaccines are dangerous.
Depending on the person, if they're all jabbed up, they'll say...
Well, they're not dangerous.
I believe in the science, you know.
And they'll start citing Dr.
Fauci.
Well, the large language models don't do that.
They will give you what you ask for.
So it's important to actually intentionally ask the leading questions of what you're looking for.
So if you want to know, for example, hey, which minerals are great for controlling blood sugar?
Well, then ask it that question.
Don't ask it just a generic question like, tell me about minerals and blood sugar.
That's too generic.
Or you can't just ask it, tell me about vaccines.
You're going to get the underlying Microsoft base model propaganda if you ask just an open-ended generic question.
Or like, tell me about transgenderism.
Yeah, you're going to get some garbage that's like, yeah, men can become women and have babies.
So you need to learn what's called prompt engineering if you don't already know it.
Prompt engineering means that you're good at asking questions of the LLMs so that you get the answers that you're looking for.
And that often means intentionally asking what we would call leading questions.
There's nothing wrong with leading questions.
It's just like when you search a search engine.
You would ask a leading query of like, you know, tell me where to find herbs that prevent cancer, right?
Same thing with the LLM. Don't just ask it like, tell me about herbs and tell me about cancer.
Say, I want to know about which herbs are known to prevent cancer or to halt the growth of cancer tumors.
Something like that.
And then it will give you much better answers.
Another thing to keep in mind is that these language models have a randomness that's kind of built in.
There is also a randomness control factor called temperature.
In all the software that runs these language models, you can adjust the temperature.
A higher temperature means more random.
And a high temperature would be something close to 1.0.
0.9 would be a high temperature.
0.1 would be a low temperature.
0.1 would give you very consistent answers almost the same every single time.
So depending on what you're looking for, if you want a more creative sort of answer, you could raise the temperature and see what happens.
Or if you want very consistent answers, lower the temperature.
The other thing is that these models will sometimes tend to repeat themselves.
And depending on what you ask them about, they can put out garbage.
They can even seem to stutter.
This is common in areas where the model is not sufficiently trained.
This is happening in these distributions right now because we focus our training on herbs and nutrition and permaculture and food and superfoods and whatever.
So if you ask it something outside of that realm, it might not give you good answers.
It might be like, and the blonde, it was in the garbage pan in the pumpkin patch with orange hair.
And you're like, what?
It's either quoting Joe Biden or something's gone horribly wrong.
You should expect those kinds of results.
Because, like I said, it hasn't been trained on a lot of areas.
So as our training continues to expand, these models will get better and better and better in answering your questions about all these areas.
By the way, the next topic we're focused on after nutrition, herbs, food, off-grid medicine, and so on is prepping and survival, by the way.
Because we intend for these models to be useful in a collapse.
We want you to have basically this large collection of human knowledge that you can query locally, you know, if we get nuked or something.
You know, if the whole internet goes down, but somehow you still have electricity locally, maybe you have a solar panel.
And you can run a laptop or whatever.
You could query this on your own laptop locally.
You do not need an internet connection.
And this thing is good forever.
I mean, well, you know, as long as silicon works, I should say.
One more thing.
Don't waste time arguing with the LLM or trying to teach it anything.
It doesn't learn anything from you telling it stuff.
It just reacts.
And it's funny because it's an intuitive thing.
A lot of people, when they ask a large language model something and they don't get the answer they want, they start to argue with it.
It becomes a debate.
And they think that they can convince the language model that they're right.
And then they think that somehow the language model is like a human being that once you convince it, it will remember what you just said, and then you will change what it says in the future.
Well, that's not the way these language models work at all.
You might as well be talking to a puppy dog.
It might actually be a better use of time talking to the dog.
Language models do not learn from being used.
They don't learn anything.
Nothing changes in the executable file just because you're asking it questions, okay?
So don't waste your time.
If you're not getting the answer you want, change the way you're asking the question.
Learn more about prompt engineering or wait for the next wave that we release and then download the newer model and try it on that model, see if it's better.
Because we're doing a ton of training, let me tell you.
So all these models are going to keep getting better and better.
But don't make the mistake of thinking that, you know, Hey, if I just show this model the truth about, you know, X, then it will remember that.
No, it's not going to remember.
Jack doesn't have any memory at all.
It uses a lot of memory to run, but it doesn't remember anything that you say.
All right, so to summarize here, this has been a big week for us in terms of technology.
I've been working on this language model project since shortly after Thanksgiving.
And we've been working on this Brighteon.ai, peer-to-peer, decentralized, uncensorable social media platform system for, I don't know how many months, many months, six months, whatever it is.
And it happens that we release these on the same day.
Like the Friday before the solar eclipse.
Wow.
I feel like there's some kind of divine guidance happening or something.
We had no intention of releasing these on the same day.
It just happened that way.
So here we are.
But these are two tools that should be part of your...
Portfolio of communication options and knowledge options.
So again, the websites are brighteon.ai to download the language model, Neo, and then brighteon.io to use our new decentralized platform.
Or if you want to download that app, then it's brighteon.io slash applications.
If you do all that this weekend, you're way ahead of the curve, let me tell you.
And if you use the Brighteon.io app, which is built on the Bastion app, then you will have emergency comms up and running from Monday.
So in case during the solar eclipse, if a giant portal opens up in the sky and giant flying multidimensional winged serpents come pouring through from the sky, and then Thor shows up and Iron Man and Hawkeye, and there's like an Avengers battle in the sky, then you can bet I'll be tweeting about it on the Brighteon.io app, which is, again, the Bastion universe.
Even if all the other websites are down, we'll be able to reach you through that.
So I don't know if I can say that's your homework for the weekend.
Consider it something pretty important, I would say.
And also, by the way, you know, look.
These tools are all free, right?
So it doesn't cost you anything to use these.
We are building the infrastructure of human freedom, right?
That's the slogan that I have been mentioning here for quite some time.
And this just demonstrates what we're doing.
And we couldn't do it without your help.
It is your...
Repeat business with us at healthrangerstore.com that gives us enough money that we can donate to our nonprofit and then the nonprofit CWC data science division can hire the developers and, you know, pay the licensing fees and buy the hardware in order to do all this.
And get this stuff done and then give it to you for free.
So, it's a really simple equation.
You support us by shopping at healthrangerstore.com to get lab tested, ultra clean foods, super foods, preparedness foods, Personal care products that we manufacture, most of them, and we test all of them.
And we also have, of course, our knives with Dawson knives that are made with magna-cut alloy that is corrosion-resistant.
It's a really amazing alloy, the best science in edge weapons that exists on the planet today, by the way.
I can't even go into that.
But if you shop with us, we take...
Whatever profit we can manage to make, we reinvest it into the infrastructure of human freedom, and then we give you back those tools like we're doing right now.
And together, we are building a future of decentralization, of knowledge, of like a parallel economy, you know?
We're doing what the establishment refuses to do.
And we are bypassing their censorship.
And we are bypassing their processed junk foods and garbage and mass medication agendas and all that crap.
We're bypassing Google.
You don't even need Google once you have this tool running for a lot of questions, for many things that you would ask.
Forget Google.
You're going to get better answers from our language model now.
And it's going to get even better.
So Thank you for supporting us.
We are committed to bringing you more tools, more solutions, more infrastructure for human freedom.
It's going to keep getting better and better all year long, as long as a giant multidimensional portal doesn't open up in the sky on Monday with flying winged serpent demons pouring through.
Of course, I'm joking.
I don't think that's going to happen.
No.
They will come out of the ground.
Yeah, no.
I don't think that's going to happen.
I think Monday is going to be okay, but we'll see.
We'll see.
Just in case it's not, just be ready.
All right, everybody.
Thank you for all your support.
Mike Adams here, the founder of Brighton, founder or executive director of Consumer Wellness Center.
Thank you for all your support.
Very happy to bring you this.
Get prepared, have a great weekend, and get ready for Monday morning, the solar eclipse day.
Hmm, should be interesting.
Take care.
A quick update on what we have back in stock at healthrangerstore.com.
I've got them here on my desk.
We have, by popular demand, organic lab-tested white cheddar powder.
This is real cheese, no crazy fillers, nothing like that.
No artificial colors, nothing.
Just real cheese powder.
And we also have heavy cream powder, which is an excellent addition to a lot of meals.
Again, also organic.
It's lab tested for glyphosate, for heavy metals, for E. coli salmonella, and so much more.
We do the most extensive lab testing of anybody in the industry.
We also have turmeric root powder back in stock.
This is a very difficult product to find in terms of getting clean lots because lead is the most common contaminant in turmeric root powder.
And if you're getting turmeric root from a source that doesn't do routine testing of every lot Then you're going to be exposed to higher levels of lead than what you might expect.
Trust me on that.
We have to reject most of the lots of turmeric that we see.
We also have collagen powder back in stock.
This is a real neutral tasting collagen.
It of course helps support your body's natural construction of ligaments and connective tissue and can help your skin have the appearance of a more youthful appearance.
A lot of people love collagen.
For a while we didn't carry it, but there were so many non-clean sources out there that we decided we've got to have a clean source.
You can find all those products and more at healthrangerstore.com.
We have a new website layout there.
We've got items that are on sale here.
We've also got categories that can help you shop.
Our laundry detergent is really popular among people who want to avoid artificial fragrance chemicals.
And here's some of our new products.
We have a date sugar that's in stock now.
We've got ahi flour oil, which is the plant-based version of fish oils, but it's from plants, not fish.
Freeze-dried pineapple and so much more.
We've got food in number 10 cans.
You can shop by health goals.
And then all of these are back in stock, including N-acetylcysteine, NAC, which is very popular, as well as our 5G defense product.
So So check it all out at healthrangerstore.com and know that every purchase helps support this platform, Brighteon, as well as all of our other projects like Brighteon AI where we're giving back to the community this technology to help empower you with decentralized knowledge.
Where you can bypass Google and the big tech censorship with tools that you can use yourself locally to get answers to questions.
That's at brighteon.ai.
And we'll talk about that another time.
But support us right now, healthrangerstore.com.
Thank you for your support.
A global reset is coming.
And that's why I've recorded a new nine-hour audiobook.
It's called The Global Reset Survival Guide.
You can download it for free by subscribing to the naturalnews.com email newsletter, which is also free.
I'll describe how the monetary system fails.
I also cover emergency medicine and first aid and what to buy to help you avoid infections.
So download this guide.
It's free.
▲
Export Selection