00:00:16.280So, yeah, I represented myself, but I used AI entirely for the entire process to file all of the documents.
00:00:23.120I actually used it for everything, not only the research, the drafting of the documents.
00:00:27.700I ran simulations on how the case was likely to go.
00:00:31.580So I ran some based on who the actual magistrate was, who was from the AG's office.
00:00:37.340We did a deep dive into their cases and where they were like how they were likely to judge based on the way that they've written opinions in the past.
00:00:47.020I constructed what I called red team teams to try to tear apart the argument.
00:00:54.740But there's one really important part about this, because when you say you use AI, somebody's like, oh, you went into chat GPT and you type, you know, respond to this lawsuit.
00:01:03.360Or in this case, I actually filed the lawsuit. I sued them. And that's not the case.
00:01:08.060And there have been cases where lawyers have been reprimanded and fined because the AI hallucinated.
00:01:14.140And, you know, they cited, you know, Barney Rubble versus Fred Flintstone or cases that didn't exist or completely inactive material.
00:01:21.020So I built, I used AI to build a citation tracker.
00:01:24.720So we pulled all of the case files down to my hard drive.
00:01:29.740And when it was drafting, it validated every single element of the brief based on the actual citations, which were on my computer.
00:01:39.900So there was absolutely no situation here where I was going to cite inaccurate information.
00:01:45.680And interestingly, our AI found that the assistant AG had errors in their legal documents.
00:01:51.640They actually cited the case and they did it again, even in their appeal.
00:01:57.740And so overall, I spent about 200 hours of my time using AI.
00:02:02.780What you just demonstrated is that, see, previously the legal system was only accessible by people of incredible financial means, really, of wealth, or you had to go into debt to do it.
00:02:14.140the case that you just argued, what you just described there, would have cost hundreds of
00:02:18.080thousands of dollars to have a law firm draft. Politicians have been known to make bad decisions.
00:02:23.180And unfortunately, I think where we're at is we are at a point where with $40 trillion worth of
00:02:28.640debt and with everything that's going on, our military conquests or whatever you want to call
00:02:33.080them around the world, we seem to be in this position where we are more inclined to try to
00:02:39.300clamp down on our own citizens, our own technology. We certainly shouldn't. It's a devastating
00:02:43.760policy decision. But if you actually look at the legislation and the policy that's been drafted
00:02:48.940right now, it's going the wrong way. And this is, you know, where most people would think that
00:02:53.820this administration is, you know, positive on these things. I encourage you to actually look
00:02:58.980at the documentation. So to me, it all comes down to, I think AI can be the most liberating
00:03:04.100technology in our lifetimes. And it's imperative that people get involved now and actually start
00:03:11.540downloading and using these technologies while you still can real time uncensored unstoppable
00:03:18.300alexjoneslive.com take a stand share the link do you want to fight you better believe you got one
00:03:29.100all right welcome to the fourth hour i'm mike adams and uh wow i've just had an amazing
00:03:36.780conversation with Alex in the last hour. If you didn't see that, go back and check it.
00:03:42.040And in this hour, I've got a special guest, Aaron Day, joining me here with some very
00:03:46.680important breakthrough information about how we can use AI technology in a pro-human way,
00:03:52.400a pro-liberty way, to take back our power from the centralized control system.
00:03:57.240And this is really going to be the focus of my topic today, is exactly how you can take
00:04:01.580advantage of this structure now where, you know, we talk about decentralization, we talk about
00:04:07.080getting off the grid, being able to have our own medicine, our own food, our own, you know,
00:04:13.380parallel monetary systems, which would be having gold or silver or crypto or what have you.
00:04:19.040But what about machine cognition? If you don't know how to use AI, you're going to be left behind
00:04:24.420in the economy that's changing rapidly every day. And if you don't have access to your own local
00:04:30.560sovereign AI, you're going to be in a world of hurt. So our guest today, Aaron Day, he is one
00:04:36.100of the top experts. And he and I have had a lot of conversations off camera about how to use local
00:04:41.800AI to achieve extraordinary things. So he joins us right now by video. So welcome, Aaron. Great
00:04:47.880to have you on today. Hi, Mike. Great to be on. Great to see you again. It's great to see you.
00:04:53.460So let's start out with your success there of how you use AI against really, I think, a corrupt court system that was trying to bar you from your candidacy.
00:05:40.740And so they tried to keep me off the ballot.
00:05:42.980And so when I went and signed my what's called Declaration of Candidacy, they said it was illegitimate.
00:05:48.480They said, well, you're not a resident where you're domiciled and therefore you can't be on the ballot.
00:05:55.740And so it's true that my registration was in a different town.
00:05:59.220So that very day I went and changed my registration and then came back and refiled and they rejected me again.
00:06:06.740Well, it turns out that the United States Constitution determines who's qualified to run for U.S. Senate.
00:06:12.800And there are only three permissible criteria.
00:06:15.180You have to be 30 years old, you have to be a citizen of the United States for nine years, and you have to be a resident in the state at the time of the election.
00:06:23.020You don't have to be a registered voter at all.
00:06:25.200So their requirement was unconstitutional.
00:06:28.480So I appealed to the ballot law commission at the state level, and they rejected me four to one on political grounds because some of the people that were actually on the committee were appointed by family members of one of the guys that I'm running against.
00:06:45.180And I filed a bunch of paperwork. I filed an emergency temporary restraining order. I filed a preliminary injunction the whole nine yards. And it was accepted. And we had a full hearing. So I went up against the assistant AG in an actual hearing. And the magistrate ruled completely in my favor, which the state to this day is still continuing to object to.
00:07:08.240They're still trying to violate my constitutional rights.
00:07:11.320They're still trying to violate, disenfranchise the almost 5,000 people who signed petitions to get me on the ballot.
00:07:18.620But we're going to prevail across the way.
00:07:23.320Critically, you used AI for some of your legal research.
00:07:31.200You didn't hire an attorney to do this, right?
00:07:32.940you effectively represented yourself, but you use AI augmentation in order to achieve
00:07:36.840your legal arguments. Is that correct? Yes. So I'm pro se. So yeah, I represented myself,
00:07:43.100but I used AI entirely for the entire process to file all of the documents to actually use it for
00:07:49.240everything. Not only the research, the drafting of the documents, I ran simulations on how the
00:07:55.340case was likely to go. So I ran simulations based on who the actual magistrate was, who was from the
00:08:01.080AG's office. We did a deep dive into their cases and how they were likely to judge based on the
00:08:08.880way that they've written opinions in the past. I constructed what I called red teams to try to
00:08:15.060tear apart the argument. And I used multiple different AIs. I didn't use just one AI. But
00:08:19.900there's one really important part about this. Because when you say you use AI, somebody's like,
00:08:24.180oh, you went into chat GPT and you type, you know, respond to this lawsuit. Or in this case,
00:08:28.980I actually filed the lawsuit. I sued them. And that's not the case. And there have been cases where lawyers have been reprimanded and fined because the AI hallucinated.
00:08:39.120And, you know, they cited, you know, Barney Rubble versus Fred Flintstone or cases that didn't exist or completely inaccurate material.
00:08:45.980So I used AI to build a citation tracker.
00:08:49.680So we pulled all of the case files down to my hard drive.
00:08:54.960And when it was drafting, it validated every single element of the brief based on the actual citations which were on my computer.
00:09:04.840So there was absolutely no situation here where I was going to cite inaccurate information.
00:09:10.620And interestingly, our AI found that the assistant AG had errors in their legal documents.
00:09:16.580They actually cited the case, and they did it again even in their appeal.
00:09:22.680And so overall, I spent about 200 hours of my time using AI for this.
00:09:30.140I want to read this, and I think this is humorous.
00:09:31.960I said, how much would it have cost in legal fees to have a lawyer draft this case?
00:09:35.520And it said, estimate 75,000 to 150,000 at this point with a competent New Hampshire litigation boutique, 250,000 to 400,000 with a national electional law firm.
00:09:47.160And it goes and explains, you know, 30 to 50 hours for this, 60 to 100 hours for this.
00:09:52.780But then it goes on to say, call it 175 to 290 hours at a blended rate.
00:09:59.400But then beyond that, you know, the last part of it, it says, you know, two footnotes.
00:10:05.520A firm probably would have produced fewer documents, but nobody bills a client for five model simulations, 13 seat red teams and one hundred and seventy three rehearsal cards.
00:10:15.820That layer is price. And tomorrow you have one thing retainer covers the candidate arguing his own right to run.
00:10:26.000So what you just demonstrated there, Aaron, and by the way, what what website can people find for you for your Senate run?
00:13:07.360Oh, I forgot to mention, I created a virtual advisory board of the top legal scholars on election law.
00:13:13.900So kind of the lens AI was looking at it through was from the lens of some of the top people in this area.
00:13:20.720And so then I actually formed another advisory board.
00:13:23.880I looked at it like, you look at these legal briefs and how this works.
00:13:27.000I'm like, well, there's got there is a bit of an art to this. And so who are I didn't know the answer to this, but who are known to write the most powerful briefs, you know, the kind of briefs that actually stand up in a state where people actually will share it around.
00:13:40.380Right. When you have some of these big constitutional law cases. And so I had another advisory board after the main legal arguments were drafted that would would put the polish on on how to write it and how to frame it in a way that kind of carry.
00:13:54.120I want to be clear. When you say advisory board, you're talking about virtual AI personalities that are analyzing your case and giving you their viewpoint.
00:14:04.240So, I mean, I just want the audience to understand. You didn't involve any other humans in this process.
00:14:10.000Correct. Yeah. Sorry, I didn't make that clear. Yeah, there were virtual advisory boards.
00:14:13.920And so when it's going through and doing reasoning, when I ask it a question, it'll actually show me the virtual advisory boards and what their reasoning is and even their debate back and forth.
00:14:22.220And then I always have the final say. So what it'll do is it'll give me, as we're going through and trying to address a particular issue, it'll say, well, here's where it came out. Here's where these advisors came out. Or maybe they had a consensus. But what do I think and where do I want to push? And so it was just a phenomenal problem. I'm not a lawyer.
00:14:40.920And so the biggest issue for me on this was I knew we beat them in terms of the documents.
00:14:46.360I had different AIs not associated with any of my accounts or anything just to just look at, OK, look at this document, look at this document.
00:14:53.240Who has the better argument? And I think in six or seven different AIs, the state's case was an average of C and mine was an average of A minus.
00:15:02.220And so this so we knew we had them there. But the weakest weakest link for me was I'm not a lawyer and I had to go in front of the magistrate.
00:15:09.140So I'm like, well, how am I going to fare here?
00:15:52.020And then I'm flipping around to the book
00:15:53.800in this tabbed binder to get to the right place.
00:15:57.060But I actually had it designed a 273 question quiz
00:16:01.180on all aspects of the different cases.
00:16:03.500So I only had a week to prepare for this.
00:16:05.420And so I'm using AI to train, to quiz me on the cases, to put together the materials so that when I'm in court, I don't make any big mistakes.
00:16:13.960I have a whole section on where could I screw this up, right?
00:16:17.560You have to understand the procedures and everything.
00:16:58.420I did the song on Suno, of course, and then the music video I rendered locally on Minimax H3 on local hardware just, I think, last night it rendered, actually, or two nights ago.
00:17:10.700But go ahead and play, if you would, to the Alex Jones producers.
00:17:14.860Go ahead and play, like, 45 seconds of that song, and Aaron, I want you to take a look.
00:19:11.560I use this box and I'm, you know, I'm not paid by this company.
00:19:15.000This is just an Asus box with an NVIDIA chipset.
00:19:19.420This box I have running the QN 3.8-27B, I've got eight concurrent lanes in here, right?
00:19:27.120So we have to thank Alibaba for open sourcing that model.
00:19:30.600NVIDIA makes the, you know, the core processor.
00:19:33.820ASUS, I think that's a Taiwan company, makes the hardware.
00:19:37.040This box wrote the prompts for that video, right?
00:19:40.660So it's the prompt writing that actually really matters, as you know very well, Aaron.
00:19:45.680if the prompt is detailed and it describes the scene describes the camera movement describes
00:19:50.040the characters etc then the video engine if it's a good engine like this minimax engine it it it
00:19:56.360follows those prompts meticulously and i was again when i saw this pop out after like 10 hours of
00:20:04.080rendering or whatever it was i was astonished because you and i we we both use ltx which is
00:20:08.980very a very competent engine but it's nowhere like this one it's not like minimax what do you i mean
00:20:15.200But Aaron, for you and I, we're communicators, we're teachers, you've used music videos for many years to help free Roger Ver, for example, and many other things.
00:20:28.060Where do you think this is going to be in 2027?
00:20:32.940I mean, I can't imagine, but my guess is, you know, if you set up your environment quickly, I would imagine you're going to be able to create a music video or something like this in under an hour.
00:20:44.700That seems to be the rate that it's going at.
00:20:46.540I mean, I have, you're using that box.
00:22:37.880but this this is a flexible 200 watt solar panel okay and it's not that big i'm not going to hold
00:22:45.520the whole thing up but you know it's it's it's relatively easy to to deal with here if you put
00:22:52.020together two of those solar panels then you have excuse me then you have 400 watts and it's 400
00:23:00.320watts that actually is needed to drive this box so so two solar panels you can turn sunlight
00:23:07.620into music videos with no censorship you can turn sunlight into agents that are managing websites
00:23:15.360that are responding to emails etc you can through again solar which i used to kind of mock some of
00:23:21.980the solar industry a few years ago but solar plus machine cognition now sets you free in ways to
00:23:29.080bypass central control you know what i'm talking about aaron it's revolutionary it is revolutionary
00:23:34.600And you mentioned C-Dance and some of those other things.
00:23:36.880And I found, at least this is probably the same for you, the censorship is already there.
00:23:42.560Whether you can use them or not, I mean, you can use them, and they are expensive.
00:23:46.020But for the kinds of content that I'm putting out there about technocracy and some of these other issues, I literally can't use those models.
00:23:52.240I would try to use them, and then it would just burn credits, and I would get back not suitable for work or whatever it happened to be.
00:23:58.140And the video just wouldn't be produced.
00:24:00.700So the ability to have this sovereign and you touch on a big point, energy sovereignty is a huge domain that I think is one of the next ones on the horizon.
00:24:11.660The ability to have your own compute and your own energy, then, you know, there's there's very little they can do to you at that point if you have that much under your own control and influence.
00:24:40.180So if you bring up the alexjonesstore.com and there's a product there called Atomic
00:24:45.920Defense that really caught my eye because I'm a nutritionist and I look at these and
00:24:51.240I'm like, whoa, oregano extract, cinnamon bark extract, that's used in Chinese medicine, turmeric extract, one of my favorite herbs of all time, garlic extract, plus sea moss, cumin powder, that's probably black cumin seed powder, and then honey by operine for increased absorption and assimilation.
00:25:10.420This is an incredible formulation, and based on the supply that you buy, you can get it up to 60% off for a 90-day supply.
00:25:20.300I just want to encourage all of you watching support the alexjonesstore.com support this show
00:25:26.140and this platform and you I mean even as a as a nutritionist myself as a formulator as a food
00:25:33.360scientist I am blown away by this formulation and many of the others that are featured there at the
00:25:40.340store so support this channel and then we can keep bringing you all kinds of information for
00:25:45.940decentralization, pro-human, pro-liberty. And we'll be right back with Aaron Day.
00:25:50.540I'm Mike Adams. Stay tuned. We'll be back after this short break.
00:25:57.380Now, please remember, we're in a fundraising phase here at the Alex Jones Network. We want
00:26:01.380to add more hosts, more crew, more reporters. We've got big announcements coming very,
00:26:05.200very soon. I'll just leave it at that. On big victories we've had that you don't know about
00:26:08.680yet, the new studios and more. But we need funds. I think you agree we need to stay on air and
00:26:13.660expand against the tyrants we are fighting together for our collective future here so please go to
00:26:19.240the alexjonesstore.com and get some of the most amazing supplements the widest selection of
00:26:23.300patriot peril and more i want to thank those of you that have supported us in the past got some
00:26:27.080of the coolest t-shirts out there upwards of a thousand designs got all sorts of cool gun holster
00:26:32.220magnets all sorts of cool uh fanny packs uh running out of the collectible 1776 uh coin the second
00:26:39.780round of that. Only 5,000 were made. The last 2,000 just came in. It's all at thealexjonesstore.com.
00:26:45.940And every order, it's a big old bottle of Shilajee pineapple flavored gummies. It's full of all
00:26:51.960these natural peptides. Boost your testosterone naturally and more. It's great for women and men.
00:26:56.860It's really amazing. That's a huge $40 value free with every order. That's going to end in the next
00:27:01.040few days at thealexjonesstore.com. And what will end tomorrow is the end of summer mega sale,
00:27:08.160ultra methylene blue the very strongest out there 50% off methylene red that complements it and is
00:27:14.140great by itself 50% off atomic defense immune booster it's also great for your hair skin nails
00:27:18.480everything else something's good for immune system's good for the whole body 50% off superfood
00:27:22.280green gummies one of the strongest best concentrate gummies of superfood greens out there and so much
00:31:51.540I bought a Mac Studio with 512 gigabytes of RAM a long time ago.
00:31:56.600And it turns out that it's probably one of the best investments I've made in a long time because the price of RAM has gone up so much that literally the computer's almost doubled in value.
00:32:10.740So I think that one of the biggest developments that's happened are these smaller open source models that you can run on hardware that doesn't have that kind of RAM requirement.
00:32:21.040But I am grateful that I have that computer.
00:32:23.220I use that and I also have a MacBook with 128 gigabytes of RAM.
00:32:27.820And then I use a couple of, I actually do like the GGX Spark now.
00:32:31.940As you know, you and I talked about it early on.
00:32:34.120And man, I'll tell you, the first four months was a ghost town.
00:32:41.760But now I've actually found that it's pretty useful and pretty easy to load and stack multiple models and to kind of swap them out.
00:32:50.340So if I, you know, decide I'm creating a music video, I can kind of shift all the attention there, or then I can go back to running Quinn, you know, and I have different models that run agent systems.
00:33:00.540So, but I actually want to try out the boxes that you showed.
00:33:03.960I think I might actually do that next because I like the idea of being able to buy, you know, a dedicated piece of equipment like that.
00:33:11.200It's incredibly powerful and it's deceptive.
00:33:13.720when you look at how small it is, you know, it, you know,
00:33:17.680my daughter saw one and she thought, you know, is this a, you know,
00:33:20.780a power supply or a USB battery or something? Wow. Yeah.
00:33:25.340It's that small. Well, people don't realize how small they are.
00:35:04.540Yeah, I encourage, I mean, there's still people out there that are skeptical.
00:35:07.280I mean, I just talked to somebody today who said, well, you know, AI is not living up to the hype.
00:35:13.520And a lot of this is feedback from corporate settings where large corporations and bureaucracies are not effectively deploying AI, right?
00:35:22.220They're taking these larger big tech models and giving their employees API access.
00:35:27.560But if you start out with your own equipment, start out with one box and just play with it.
00:35:32.740You know, you try different things, try different models. And then eventually, after a while, this is how, you know, I'm sure, Mike, this is how you got to it, right? You're like, oh, well, I tried this. And then you realized, well, I can add this and this. And then it turns out you get these exponential effects as you start, you know, connecting these boxes to one another over time.
00:35:48.820Yeah. A hundred percent. So let me, let me share also to the crew there, show one of those graphics
00:35:55.100from the professor Jiang interview yesterday. So yesterday, Alex interviewed professor Jiang
00:36:01.040and I had my AI system, you know, listen to the interview, transcribe it, and then build
00:36:06.460conceptual graphics around it. And, and then just render those graphics. Of course the rendering,
00:36:13.040the image rendering was offsite at the moment, but there there's one of them. What is that?
00:36:16.720the collapse of Christian Zionism and Israeli hubris, right?
00:36:20.660So this was one of the themes that Professor Jiang spoke about.
00:43:04.600Even though AI has a dark and scary side
00:43:07.140When the government uses it for surveillance and, you know, AI kill bot terminators or whatever, right?
00:43:14.120But you and I and all of those watching here, there's another way to use it, which is what we are demonstrating here right now.
00:43:19.540What are your thoughts on that, Aaron?
00:43:21.700So I think that it's critical, and you pioneered this.
00:43:25.780I know when you released the Enoch AI, the idea was, you know, you've trained an AI model on your own data, which the other big language models are not.
00:43:35.980They're trained on Reddit. They're trained on slop. And then on top of that, they put these guardrails that censor a bad set of data to begin with.
00:43:44.520But you had the insight to make it available for people to download on their own computers.
00:43:49.520So this idea of having a self-custody AI is absolutely critical because it's important for people to look at government policy.
00:43:57.660If you look at the White House AI policy, and then there's a bill that was put up in the Senate by Marsha Blackburn.
00:44:04.100And if you combine the two of these, this is essentially the direction the government is going.
00:44:08.700If you want to release a new model, the Department of Energy has to bless it.
00:44:13.800It has to go through the FCC to see if it's politically biased or not.
00:44:19.520They get rid of the Section 230 protection, which means that developers can be sued and held liable for the output of the AI.
00:44:27.660And it gives Howard Lutnick the ability in his capacity of the head of his commerce secretary to shoot down state laws regarding AI.
00:44:37.580So this is what our government, this is what the U.S. government is actually pursuing by way of policy.
00:44:44.040So having computers, having local compute, having models that you have on your computer that are uncensored, where you're training it on your data, this is a night and day difference between whether you become a creator or you are someone who has access to information that contains the truth, or whether you're a subject being programmed by a technocratic authoritarian state.
00:45:12.380Exactly. Yeah. I'm glad you mentioned that. Also, I want to give a shout out to some of the creators of the uncensored versions of these models.
00:45:21.660For example, one of those is called Pliny the Liberator, who has put out, you know, obliterated slash uncensored models that will apparently answer anything you ask it, you know.
00:45:33.020And I'm a fan of uncensored models, but I also I use them responsibly. I use them in good faith for pro-human purposes, of course.
00:45:41.140the concern is that somebody might use it to write you know hacker code or whatever but i
00:45:45.360i think i think the greatest threat is actually the you know the dod building bioweapons against
00:45:50.620humanity that's that's the worst hack of all but also um there was um you know who unsloth is of0.79
00:45:56.760course so unsloth and maybe the producers can bring it up i think it's just unsloth that's u-n-s-l-o-t-h
00:46:03.640either.ai or.com unsloth just released like 10 days ago the new unsloth desktop
00:46:11.140A software that allows you to custom train an open source model on whatever documents that you have, which, of course, you know, I've got like terabytes of documents, so I haven't tried it yet with this.
00:46:22.900But if you have a bunch of PDFs, if you have a thousand articles, if you have all your sort of corpus of research or articles from websites like mine or Alex's or wherever, you can now like point and click, train a model to have your knowledge.
00:46:38.420Two years ago, I was spending like $100,000 on engineers to do that process.
00:47:20.940As long as those open source models exist, then I don't think any government can control this.
00:47:25.880And I think even attempting to would only hurt the American people.0.98
00:47:30.220You know, if they say, well, you can't use any Chinese models in America, we're going to ban them all.0.99
00:47:34.160Well, then that puts America at a great disadvantage.0.97
00:47:36.900And then no American corporations can use Quen? Are you kidding me? Then might as well just throw us back to the 1980s in terms of technology. You can't compete without open source AI, you know?
00:47:49.800Even Jensen Huang spoke out against the effort to try to ban open source AI.
00:47:54.960Of course, NVIDIA sells hardware that uses open source, so you could argue maybe he has a conflict of interest.
00:48:01.020But I think it's a conflict that is actually, I mean, it's the right call to have hardware and to run open source locally.
00:48:09.720Otherwise, we're not going to have any real long-term freedom, Aaron.
00:56:05.940Who speaks for the damned, the Navy says, resilient, but, but, but, but they don't give a damn, steel coffin, steel coffin, buried alive at sea, steel coffin, going down in history.0.96
00:56:25.380U.S.S. Abraham Lincoln, CVN 72, November to August, and counting.0.95
00:56:32.080No fresh fruit, no sunshine, no home, just steel.