Health Ranger - Mike Adams - AI Capabilities Advancing RAPIDLY Even on the Same Hardware
Aired: 2026-04-04
Duration: 11:10


=== AI Reviewing Entire Code Base (05:04) ===

[00:00:03]  So here's an AI update for you.
[00:00:05]  You know, I'm an AI developer and I've built several very popular platforms and I'm building more things that are going to be really fascinating.
[00:00:14]  And one of the techniques that I use now, which shows you the capabilities of artificial intelligence is that I ask AI.
[00:00:24]  See what?
[00:00:24]  Once I'm part way through a project, I ask the AI engine to tell me what features or improvements would make this project more successful given the project goals.
[00:00:37]  So, in other words, I ask it to give me suggestions, and sometimes I ask it to ask me questions.
[00:00:45]  Like, you know, here I want to add this feature.
[00:00:48]  You ask me what you think is relevant about this feature.
[00:00:52]  Just give me a series of questions, and I'll give you the answers and give you the details.
[00:00:56]  Other times, I tell it to come up with new ideas and new features and to review the entire code base and find anything that's missing or anything that might make.
[00:01:07]  Processing more efficient or more redundant, more resilient to errors or disconnects or things like that.
[00:01:14]  And I found that today, the AI systems that exist, like Claude Code, for example, are incredibly capable now with this process.
[00:01:25]  They're very good at coming up with suggestions.
[00:01:28]  In fact, it was just yesterday I was working, let me back up, I've rewritten my document processing engine.
[00:01:37]  For classifying and cleaning documents to be used as indexed reference documents for our bright answers.ai engine, as well as the bright learn.ai book engine.
[00:01:49]  And this also powers the research engines for the articles at naturalnews.com, as you probably have guessed, because they're so well researched now.
[00:01:57]  You're like, how do they do all that research?
[00:02:00]  Yeah, AI agents with this incredible, massive document base.
[00:02:04]  That's how it's actually done.
[00:02:06]  But anyway, I.
[00:02:07]  I decided to rewrite the whole thing, the entire way that document processing is handled, because I was running into some inefficiencies with the old design.
[00:02:17]  Anyway, so I rewrote the whole thing in about four hours with the help of AI.
[00:02:27]  That was a project that used to take six months, let's say a year and a half ago, or yeah, even heck, even a year ago would have taken months.
[00:02:36]  But I got that done in four hours because I was able to reuse a lot of the existing code, but I had to restructure the whole workflow, et cetera.
[00:02:44]  Anyway, that took about four hours.
[00:02:45]  And then after the four hours was done, I then asked the, I was using Claude code for this one.
[00:02:52]  And so I asked Claude, I said, I want you to review the entire code base now, and here are the goals, and I want you to give me suggestions.
[00:03:00]  on what's missing, you know, what have I left out of this?
[00:03:03]  Now, before, let's say six months ago, it could not have reviewed the entire code base because the code base was too large.
[00:03:10]  It didn't have a large enough context window.
[00:03:12]  It couldn't keep it all in its memory at the same time.
[00:03:15]  You know, it couldn't really process the whole code base.
[00:03:18]  Now it can.
[00:03:19]  So it goes through the whole code base, which for this project, I don't know, it's a couple hundred K of Python code.
[00:03:26]  So it's not massive, but it's not tiny either, right?
[00:03:31]  So it goes through the whole code base and it comes back with 14 suggestions.
[00:03:36]  14 suggestions separated by high priority, medium priority, and low priority.
[00:03:43]  And I went through all 14.
[00:03:45]  I just, you know, reading through, like, oh, yeah, that sounds good.
[00:03:48]  I forgot about that.
[00:03:49]  Oh, yeah, the retries over here.
[00:03:51]  Yeah.
[00:03:52]  Oh, there's a file renaming collision problem over here, blah, blah, blah.
[00:03:58]  And I looked at all 14 and I said, Yes, all 14 do it, you know.
[00:04:05]  So the engine says, okay, sir, you know, we'll do all 14, and it spawns 14 agents and then it updates all the code, it adds all 14 things.
[00:04:16]  And then there's another step that I always do after it makes a bunch of changes like that.
[00:04:19]  I always go back and I tell it, review all the changes you just made one more time to see if you introduced any errors or if you left something out.
[00:04:32]  And occasionally it will find problems that way.
[00:04:37]  Not always.
[00:04:38]  Sometimes it just comes back and says, everything was perfect, no problem.
[00:04:41]  But every once in a while it's like, oh yeah, I found there's two bugs and here they are, Bloom.
[00:04:45]  I'm going to fix these.
[00:04:47]  So, anyway, I had to go back and check all 14 features that it added.
[00:04:51]  Everything was good.
[00:04:53]  And so then I ran the program and smooth as silk, you know, smooth as silk.
[00:04:59]  It's just like churning away.
[00:05:02]  And again, this would have taken months for a human programmer to do not that long ago.

=== NVIDIA Hardware Memory Limits (04:57) ===

[00:05:08]  Months.
[00:05:09]  And I know this because I hired human programmers to do this, you know, a couple of years ago when I started the whole AI project and the data pipeline processing.
[00:05:19]  I was paying human programmers to do this.
[00:05:22]  Now, I just use AI and myself, and it all happens in a few hours or a few minutes in some cases.
[00:05:33]  So, anyway, this is where AI is today.
[00:05:38]  And AI advancements have not slowed down.
[00:05:42]  They have not slowed down.
[00:05:44]  There have been advancements that make the existing AI inference hardware, the hardware base that's installed around the world, including all of our hardware in my mini data center, I call it.
[00:05:55]  48 workstations.
[00:05:57]  Our hardware will be able to do more and more with each passing month as more innovation takes place.
[00:06:03]  For example, Google famously released a paper called TurboQuant, and TurboQuant allows about a six times compression of the KV cache, which is one of the, it's like the context cache that's needed for the model to process all your context.
[00:06:22]  You know, when you paste in a hundred K of text and you ask a question like, here, you know, Here's an entire book, or here's a PDF file, and I want you to find an answer that's somewhere in this book, right?
[00:06:37]  So you paste in the whole book.
[00:06:39]  That's a lot of context.
[00:06:40]  And it turns out that that context takes up loads of memory in the GPU.
[00:06:47]  It's the memory hog, actually.
[00:06:49]  I mean, obviously, the model itself takes up memory, but the KV cache takes up even more.
[00:06:54]  Well, it can, depending on your context size.
[00:06:57]  And remember that DeepSeq version four, which is supposed to be coming out soon, Has a 1 million token context window.
[00:07:06]  A million tokens, that's going to burn a lot of memory in your GPU, obviously.
[00:07:12]  So, anyway, Google comes out with TurboQuant and says we can reduce that by a factor of six without losing any fidelity in the model.
[00:07:21]  And through some clever math, it actually works.
[00:07:24]  And people are now demonstrating that.
[00:07:26]  So, all of a sudden, not only does the same hardware handle six times as much context, you know, theoretically.
[00:07:35]  But also then looking through that KV cache is much, much faster for the GPU, you know, for, for the inference process.
[00:07:43]  It's faster because it's less stuff to sort through.
[00:07:46]  So it's faster and smaller.
[00:07:48]  It's still the same base model, and it's the same base hardware, but it does more.
[00:07:54]  And also, in a similar fashion, we've seen some new OCR models that are very, very small, like 1 billion parameter OCR models that are incredibly good, just ridiculously good, better than models that were 20 times larger at OCR.
[00:08:11]  TTS models for text to speech are also, some of them are getting very good and very fast, et cetera.
[00:08:18]  So, a lot of improvements are coming on the same hardware.
[00:08:22]  And that's why demand for high bandwidth memory is actually suddenly, you know, falling because people are realizing that, hey, we might not actually need as much high bandwidth memory.
[00:08:33]  We can do more with smaller GPUs.
[00:08:36]  And this sucks for NVIDIA, obviously, because NVIDIA wants to sell you the largest, most expensive card possible.
[00:08:44]  And by the way, I had a bad thing with NVIDIA recently.
[00:08:48]  I had a $9,000 GPU that was faulty, it kept.
[00:08:56]  Dying in inference.
[00:08:59]  And I went through a warranty replacement process with NVIDIA.
[00:09:02]  And at first, they were helpful and they're like, you know, send us the serial number and this and that, and give a photo and proof of purchase and all this stuff.
[00:09:10]  And they were like, you know, they were acting like we're going to replace it.
[00:09:14]  And then at the end, they said, no, we're not going to replace it.
[00:09:17]  Even though it has a three year warranty, they said, you got to talk to the retailer that sold you this.
[00:09:22]  I'm like, are you kidding me?
[00:09:23]  This is a manufacturer's warranty replacement.
[00:09:26]  So I'm still fighting with NVIDIA over that.
[00:09:29]  You know, you spend $9,000 on a card, you expect it to work.
[00:09:32]  And if it doesn't work, you expect them to replace it.
[00:09:35]  But so far, they have refused.
[00:09:36]  So NVIDIA sucks for that reason.
[00:09:40]  But I mean, I spend a lot of money with NVIDIA, you know, hundreds of thousands of dollars a year.
[00:09:46]  And you would think that they would treat me like a customer, you know, instead of a piece of trash, but whatever.
[00:09:55]  So just be careful with NVIDIA.
[00:09:58]  A lot of times, their cards don't work.
[00:10:01]  And you may have to get a warranty replacement.

=== Free BrightLearn AI Engines (01:05) ===

[00:10:05]  Anyway, just letting you know about all of that AI is getting a lot more technical and capable.
[00:10:13]  You can use my AI engines, they're all free.
[00:10:16]  You can use my AI engine at brightanswers.ai, for example.
[00:10:22]  That's our deep research engine.
[00:10:23]  Or you can use our book creation engine at brightlearn.ai.
[00:10:27]  And there you can also, of course, download books and the audiobooks that we have available now.
[00:10:33]  They're all completely free.
[00:10:35]  Downloadable.
[00:10:36]  That's at brightlearn.ai.
[00:10:39]  And you can follow more of my podcasts at brightvideos.com and my articles at naturalnews.com.
[00:10:46]  So, a lot of AI advancements coming, and I have more announcements coming up this year as well.
[00:10:51]  Major things are happening, major improvements.
[00:10:53]  So, just be ready for that.
[00:10:54]  Thank you for listening.
[00:10:55]  Take care.
[00:10:59]  Start your day right with our organic, hand roasted whole bean coffee.
[00:11:03]  Low acid, smooth, and bold.
[00:11:04]  Lab tested, and ethically sourced.
[00:11:07]  Taste the difference only at HealthRangerStore.com.