Mike Adams Announces the BrightAnswers.ai Deep Research AI Engine
|
Time
Text
All right, welcome.
I'm Mike Adams, an AI developer, the founder of BrightLearn.ai, and I've got news for you here.
We've just launched a new website called BrightAnswers.ai.
Brightanswers.ai is our new upgraded AI research engine.
It's not a chatbot.
It's something much more.
I call it the uncensored deep research with verified document citations engine.
And that's exactly what it does.
So if you've been using BrightU.ai, which has been a very capable engine, by the way, uncensored, you know, tells the truth about big pharma, tells the truth about history, tells the truth about, you know, economics, all kinds of things, then you're going to love the new site.
We've just launched, Brightanswers.ai.
Now, if you go there, it works just like BrightU.ai.
In fact, it is the same code at the moment.
And, oh, by the way, it might be slightly buggy for the next day or two or three because of the transition.
But if you ask it a question, now it will do deep research into our curated indexed document library, which contains tens of thousands of books, tens of thousands of science papers now.
Those are launched, and millions of articles from websites like naturalnews.com and others.
Now, then the engine takes into account all of these documents in formulating its answer, and it cites them properly now, where we were unable to achieve well-formed citations previously.
Now it's perfect.
So now all of the science papers are cited with their actual science journal citations, you know, the year, the journal name, the volume number, the page number, etc.
Title, author, the whole deal.
So you can ask questions about herbs, about diabetes, about immune systems, about the brain.
I mean, anything you want.
In fact, we're covering economics and metallurgy and mining and agriculture, fusion and physics.
I haven't yet put chemistry in there, but I will soon.
And I had been promising you for many months that the science papers were coming.
Remember that?
If you've heard my podcast?
And now they're here.
They're here.
So I've got well over 10 million science papers that have been summarized and well, keywords have been created for them.
They've been cleaned, normalized, as we say, ready for ingestion into our document system.
Over 10 million ready.
I have so far selected and put in a number smaller than 10 million.
And I don't even know the actual number of unique titles because of the nature of the system, but I can tell you there are over 200 million words that are in the system right now.
Over 200 million words of science papers.
And on top of that, we currently have over 10,000 books.
That number is about to leap into 20,000, 30,000, probably 50,000 soon.
So I know that we have over 50,000 science papers in the system right now.
And soon we'll have over 50,000 books.
We have way over a million articles, actually, in the system right now.
And these are all accurately cited sentence by sentence or paragraph by paragraph in the answer that is generated for you.
So that's available to you right now.
There's only one little caveat in all of this.
Because of the capabilities of this, we are splitting brightanswers.ai into a free tier and then a token tier, just like we've done at brightlearn.ai.
So if you haven't used brightlearn.ai yet, that's the book creation engine, and there are like 15,000 books generated there right now.
Oh, I forgot to tell you that when you use the AI engine at brightanswers.ai, at the end of every prompt, or at the end of every answer, it will recommend five books from brightlearn.ai that are related to the query.
So in other words, the bright learn books are now appearing in brightanswers.ai, all those AI answers.
And that's pretty cool.
But anyway, the caveat is that, you know how on brightlearn.ai there's a free tier where you don't need a token and I'll describe the tokens in a second.
You don't need a token to create a short book of three chapters.
So that's called the free tier.
But if you want to create a longer book with better cover art, with more chapters, many more chapters, then you use what I'm going to call a Bright token.
And a Bright token, or as I previously said, just tokens, these can be acquired at healthrangerstore.com.
When you purchase products at healthrangerstore.com, you get loyalty points equal to about 5% of your purchase.
And those loyalty points can be used as credit on future purchases there, or you can trade them for bright tokens.
And these tokens are really becoming the digital, I wouldn't quite call it a currency, but like a digital voucher across our ecosystem.
So you can use those tokens to generate long books.
And then on our new AI engine, Brightanswers.ai, you'll be using those tokens to get the answers that use the book engine and the science papers engine with much more research and much longer answers.
Whereas the free tier will use only articles.
And again, there's over a million articles, so it's still awesome.
But it will create answers using articles as reference items, but not science papers and not books.
Now, here's the thing.
I haven't put in the token system yet.
That will happen in the next couple of days.
So as of right now, everybody gets all the features for free until I put in the token feature, at which point the science papers and the books will not be available to the free tier users.
But then again, free is awesome.
You know, free is free, and the answers are still amazing, even in the free tier.
But the token tier brings in a whole different level of authoritative research and also longer, more detailed answers, by the way.
Significantly longer answers.
So it's equivalent of having 100 hours of deep research and then writing up an executive report on your prompt.
Even if it's a simple prompt, like, hey, tell me about turmeric versus cancer or something.
Or tell me about foods that help prevent type 2 diabetes.
It will do extensive research on that.
And it will give you extensive citations.
Again, books and science papers and articles.
And it will give you that full report.
But that will require a token for all of that.
So that's the only difference that's coming up.
But anyway, we decided that the old website, BrightU.ai, wasn't that great of a name, which I understand.
And then also, Brighteon.ai wasn't available because the word Brighteon has been censored by so many tech platforms, including X, that Elon Musk, his team, is still censoring links to Brighteon URLs, Brighteon.com, Brighteon.ai, you name it, right?
And that's why we've sued X, and they still maintain their censorship, despite the fact their legal team is well aware that we're going to win that lawsuit eventually.
And they'll have to, you know, have to compensate us for all the years of extreme censorship, or they'll have to turn the channels back on or whatever.
You know, they're going to lose.
But until then, they can delay and delay and they can keep censoring because X is not a free speech platform.
It's a censorship platform, as you are probably well aware.
So we had to move away from the Brighton domains.
And so we went with brightanswers.ai.
And that's why we also have Brightlearn.ai.
And if they're going to censor every domain with the word Bright in it, well, then that's going to be a million domain names belonging to all kinds of people who have the word Bright in their domain names because it's actually a very common prefix for domain names.
So hopefully we won't be as extensively censored because the whole point of this engine is knowledge without censorship.
We want to give you the tools to be able to dig up good quality information about pharmaceuticals and psychiatric drugs and vaccines and fiat currency, the history of the Federal Reserve, false flag operations, whatever.
Anything that's important, we want to be able to have access to that.
So that's why this engine, that's why we moved it over to brightanswers.ai.
So get used to that new domain name.
I'll mention it a few times.
And again, for right now, it's all free.
And it also has the coaches.
It has the wellness coach, the financial coach, the survival coach, and the ingredients analyzer.
Those might be a little glitchy at the moment just because of, well, we've made a lot of changes under the hood also.
But you're going to find that when it does work correctly, the answers are freaking amazing.
The other thing is this puts a much heavier load on the LLMs that we use.
And the heavier load is resulting sometimes in like errors of too many requests.
So if you get a too many requests error, understand that we've already implemented a fallback language model.
And then I'll probably put a third tier in there as well.
So I'm trying to work with that so you don't get too many requests errors.
But scaling this up is possible, but difficult because there's a scarcity of compute in our world right now.
There's a scarcity of compute.
That is in the Western world.
In China, they've got plenty of electricity and it's dirt cheap.
But in America, electricity rates are skyrocketing, especially along the eastern 13 states.
And this is partly because of some of Trump's tariffs and things like that and geopolitics and also reciprocating trade wars.
There's more commodities scarcity in metals like copper and nickel and cobalt and whatever else that is necessary to manufacture microchips.
So as a result, just making the GPUs is becoming more and more difficult.
And that's why NVIDIA just announced, well, a week ago, that they're going to drastically raise their prices for all their GPUs this year.
They're going to more than double their consumer-grade hardware price.
So the 590 or I'm sorry, 5090 cards, the GeForce RTX 5090s, I own quite a few of those.
I was buying them at $2,400.
They're going up to $5,000 in February.
And out of curiosity, I checked the price over the weekend.
On Amazon, they were already $3,500.
So they went up $1,000 just from the announcement.
Yeah, same exact card.
It was, you know, it was $2,400 last week.
Now it's $3,400, $3,500, $3,600, something like that.
And it's going to go to $5,000.
Think about that.
So the hardware cost of compute is going through the roof.
And that's because of scarcity of commodities.
And also inflation, which stems from money printing.
Yeah, you got it.
So in this world where hardware is getting more and more expensive, isn't it great that we have services that are less expensive, that are free?
Like Bright Learn, where you can create your own book free of charge.
And Brightanswers.ai, where you can conduct these deep research queries now based on all these indexed curated documents.
And that's free.
But there's going to be a limit.
Like if a million people try to use it, it's not going to work.
It'll overload the thing.
And then at that point, I don't know.
I don't know what we would do.
But fortunately, it's not that well known.
I mean, not on the mainstream basis, but all of you, you'll be able to use it.
Just don't tell everybody about it.
We can't let everybody use the engine.
It'll overload the thing.
The other thing worth mentioning, by the way, is that for me to start bringing these science papers into the system, which is now happening, I spent two years.
I spent two years building an in-house infrastructure to acquire all the science papers, or nearly every paper that's ever been published in the world in every language, and then to process those science papers.
I've had, I mean, I think I've told you the story.
I have 48 workstations.
Each one has a GPU.
Some of those are the 50-90 cards.
Some are 40-90s, 4080s, 4070s, all kinds of cards, right?
And these 48 workstations have been running 24-7 for at least a year and a half, I think.
They're doing the data pipeline processing for our AI engine that we released that you can download at brightanswers.ai, by the way.
And also for prepping all the science articles and prepping all the millions of books that we're working through right now, getting everything cleaned up and ready for ingestion.
So if you're wondering, like, how did I get 50,000 science papers indexed into this system?
And by the way, that's going to be 100,000 in a week, and it'll be a million in a month, if not sooner.
The answer is that that's actually a multi-year process.
So it didn't just happen overnight.
You know, it took a long time to acquire them, a long time to process them.
And then classification prompts and all kinds of things.
And only then can you do the ingestion and indexing into an index system that is queried by the AI engine to say, hey, here's a prompt about whatever, vitamin D and cancer.
And tell me all, give me all the science papers related to vitamin D and cancer.
And then it retrieves all those and then uses those to help formulate the answer.
And for those of you in the AI space, you might call it rag or retrieval augmented generation, but it's actually, this doesn't exist as a rag layer.
It's more like a research augmented generation because it's way more extensive than rag.
And we're not using a typical rag vector database for this purpose for a number of reasons.
So it's technically not rag, but it's rag-like.
And it's massive.
You know, the number of documents is massive.
So now that we have been able to get 50,000 science papers into the system, what's coming next, within, I would say, within one week, is that we're going to make those available to the BrightLearn.ai book creation engine.
And then BrightLearn is going to get a major upgrade to where it starts having this improved citation format where the references for each subchapter are listed with their science paper citations and everything.
And more citations that are in line in the sentences and in the paragraphs or at the end of paragraphs.
So inline citations and science paper citations in there and more books coming.
So the Brightlearn.ai engine is going to get a huge upgrade.
And once that kicks in, then those new books that are generated will have much more research available.
And we may go back and then regenerate some of the previous books.
Well, in fact, I'm sure we will, especially the ones that are more popular, to regenerate them now that we have new research.
And when I say regenerate, I should really be clear.
I mean to rewrite them.
So to start the writing over again, bringing in new research information and so on.
But I think I want to wait until I get 100,000 science papers and maybe at least 50,000 books into the indexing system.
And then I'll probably start rewriting books.
I don't want to just do it with a little bit better information.
I want to have a lot better information.
So that will be coming over the next few weeks.
And I'm watching the ingestion engine literally right here right now on my screen.
And it's like every science paper takes, oh my goodness.
This is putting 28,000 more papers in right now.
This little routine.
every science paper takes what is it one with the two with the that's about a second It's about a second per science paper.
So you can do the math on that and figure out how long it takes.
And then when it's ingesting science papers, it can't also ingest books at the same time.
And books take a lot longer per book because books are longer than science papers typically.
So a book might take 10 seconds to fully ingest.
And that's after months and months of normalization and cleaning of all the books and everything.
So this is just the last step, which is indexing.
Anyway, it all takes time, but it's all coming online.
And the bottom line for you is that now at brightanswers.ai, you've got the world's best AI engine by far that has the only curated indexed knowledge base that exists on any AI engine.
There's nothing that even compares to this, nothing.
And it's very difficult to achieve, as I just described.
It takes years to put this together.
So if you want deep research on almost any topic, including survival and preparedness and emergency first aid and food and food preservation, use brightanswers.ai.
It's going to give you by far the best answers.
Oh, wow.
Speaking of answers, it's funny.
Oh, wow.
I was running a long query to find out how many unique science papers had been ingested into the engine.
And this does not count the 28,000 that I just mentioned that are in progress.
But before the 28,000, it was 46,093.
So add 28,000 to that.
Or, I don't know, roughly.
We're talking about 75,000 science papers roughly by the time you hear this.
So that's a lot.
That's a lot of science papers.
Some of them are readable, too.
Like a lot of them are not.
It's just filled with numbers and citations and charts and things that didn't really translate that well.
Okay, in any case, again, the engine is brightanswers.ai and the book creation engine is at brightlearn.ai.
So take advantage of all of it and enjoy.
Stock up on the long-term storable Ranger Bucket Set.
536 servings of clean organic superfoods for your survival pantry.