Latest / The Lone Gunman Podcast : JFK Assassination / RRR W/Rob - Embracing The Inevitable

Transcript

RRR W/Rob - Embracing The Inevitable · The Lone Gunman Podcast : JFK Assassination

• 0:07
  Speaker 1: This is Random Research Ramblings with Rob. Bring your notes,
• 0:13
  Speaker 1: your doubts, and your least favorite footnotes. The file says
• 0:18
  Speaker 1: one thing. The headline says three. Somebody sold certainty for
• 0:21
  Speaker 1: a subscription fee. JFK missing pages. Convenient alibi. A polished
• 0:25
  Speaker 1: little theory with a hole the size of Dallas skies.
• 0:27
  Speaker 1: I'm not here for sacred cows or credential parade. I'm
• 0:30
  Speaker 1: here for the claims that collapse when the questions get made.
• 0:32
  Speaker 1: Stream of consciousness, facts in the cricket stack. If the
• 0:35
  Speaker 1: evidence walks out, I'm dragging the argument back. Random research
• 0:39
  Speaker 1: ramblings with Rob. Pull the thread till the whole false
• 0:42
  Speaker 1: story comes undone. Random research ramblings with Rob. No padded conclusions,
• 0:46
  Speaker 1: no damage control, just questions and smoke. Conspiracies, history, rabbit
• 0:49
  Speaker 1: holes and scorched earth reviews of dishonest researchers selling certainty
• 0:52
  Speaker 1: as news. Random research Let's investigate the investigation.
• 1:12
  Speaker 2: Let me unmute myself and welcome you in to Random
• 1:16
  Speaker 2: Research Ramblings with Rob. I'm your host, biggest, baddest, Bob.
• 1:23
  Speaker 2: In the land, yes. Thank you to Jim DiEugenio for
• 1:28
  Speaker 2: that nickname. I never get tired of saying it. Wow,
• 1:33
  Speaker 2: what a week, folks. So what do you want to
• 1:39
  Speaker 2: talk about? Did anything happen? In the last week since
• 1:46
  Speaker 2: we last did random research ramblings?
• 1:49
  Speaker 1: Maybe.
• 1:51
  Speaker 2: Maybe. And I'll just wrap it up for everybody real
• 1:56
  Speaker 2: quick here off the jump. And then. I'm going to
• 2:01
  Speaker 2: let the chips fall where they may. Because I said
• 2:05
  Speaker 2: what I said. And I did what I did on
• 2:09
  Speaker 2: the last random research ramblings. And. I didn't do anything
• 2:18
  Speaker 2: except defend. Uh. our intellectual property here at the show,
• 2:26
  Speaker 2: you know, when another show that does kind of the
• 2:29
  Speaker 2: same thing about the same subject matter blatantly steals your
• 2:34
  Speaker 2: intellectual property, I'm going to point it out. I'm not
• 2:38
  Speaker 2: that nice of a guy where I'm just going to
• 2:40
  Speaker 2: let it slide.
• 2:41
  Speaker 1: Okay?
• 2:43
  Speaker 2: So, I pointed it out. Okay? The other thing I
• 2:49
  Speaker 2: did was point out something obvious. to anybody with eyes
• 2:54
  Speaker 2: and that just decide to pay attention a little bit.
• 3:00
  Speaker 2: And I know a lot of people aren't really worried
• 3:03
  Speaker 2: about numbers and this, that, and the other and don't
• 3:06
  Speaker 2: really pay much attention. But if you do, you would
• 3:11
  Speaker 2: have noticed a very abnormal growth with another podcast. Of
• 3:21
  Speaker 2: the same subject matter. That we talk about here. Not identical.
• 3:26
  Speaker 2: But a lot of the JFK assassination. And. When you
• 3:33
  Speaker 2: have a growth like that. There's only. One logical explanation
• 3:38
  Speaker 2: for it. For me. Now I will say. Tim did
• 3:44
  Speaker 2: reach out to me after. The show had been out
• 3:47
  Speaker 2: for an hour. Last week. and apologized for the, uh,
• 3:58
  Speaker 2: co-opting of our intellectual property here of our thumbnails design and, uh,
• 4:06
  Speaker 2: said that he would tweak it into something else and
• 4:09
  Speaker 2: that's fine. Um, I don't know that he did because
• 4:17
  Speaker 2: he blocked me on just about every, uh, platform known
• 4:23
  Speaker 2: to man so there's that I like how I'm the
• 4:30
  Speaker 2: bad guy here right I point out the obvious but
• 4:35
  Speaker 2: I'm the bad guy okay real mature Tim real mature
• 4:42
  Speaker 2: but whatever you know this is what happens when you're
• 4:50
  Speaker 2: dealing with dishonest people And he claims that the inflated
• 4:55
  Speaker 2: subscription numbers are a result of him doing paid promotion
• 5:02
  Speaker 2: for his podcast. Which I highly doubt. Paid promotion puts
• 5:16
  Speaker 2: your show out there in front of other people who
• 5:21
  Speaker 2: log on to YouTube. who don't normally check out the
• 5:25
  Speaker 2: kind of content that we do here. I.e. they're not
• 5:30
  Speaker 2: interested in it. Right? So you have a lot of,
• 5:35
  Speaker 2: you know, people worried about other things and anime and, uh,
• 5:41
  Speaker 2: you know, whatever, Mr. Beast and who knows what people are,
• 5:46
  Speaker 2: you know, get ready with me kind of things. I
• 5:48
  Speaker 2: don't know what else goes here on YouTube. Um, But
• 5:57
  Speaker 2: you know, when, when people watch a lot of certain
• 6:03
  Speaker 2: kinds of videos on YouTube, the algorithm picks that up
• 6:07
  Speaker 2: and they put like content in front of you. So
• 6:11
  Speaker 2: for instance, I like watching storage unit videos where people
• 6:14
  Speaker 2: buy storage units at auction and they go through them,
• 6:18
  Speaker 2: you know, on the video and see if they find
• 6:19
  Speaker 2: anything worth anything, or if they wasted their money. And
• 6:24
  Speaker 2: they bought a bunch of trash and empty boxes. You know,
• 6:28
  Speaker 2: it's whatever, right? You know. So I'll get recommended other
• 6:32
  Speaker 2: videos by other creators that I've never watched before or
• 6:36
  Speaker 2: I didn't even know existed. Right? But, you know, when
• 6:41
  Speaker 2: you pull up your feed every once in a while,
• 6:43
  Speaker 2: so like right now, my feed is full of things
• 6:49
  Speaker 2: that I normally watch. Right? I have a lot of
• 6:55
  Speaker 2: storage unit videos, true crime videos, uh, skateboarding videos. Um,
• 7:04
  Speaker 2: but then I'll, I will get, uh, some weird things
• 7:09
  Speaker 2: every once in a while recommended to me. Um, like,
• 7:17
  Speaker 2: you know, a random Reba McEntire song, right. Which is,
• 7:21
  Speaker 2: Nothing against Reba McEntire, but I've never watched Reba McEntire
• 7:24
  Speaker 2: here on YouTube or searched anything remotely, women's country, nothing
• 7:29
  Speaker 2: like that, right? But that will get put out in
• 7:33
  Speaker 2: front of me. Something about ripping sports cards, right? Okay,
• 7:41
  Speaker 2: I don't know. Something about hydrogen peroxide. Uh... But you
• 7:53
  Speaker 2: get the idea. You'll get recommended like content. You'll get
• 7:58
  Speaker 2: recommended content by people you're subscribed to. And you'll get
• 8:01
  Speaker 2: recommended content by people you've never seen or heard of before.
• 8:04
  Speaker 2: And the reason that happens is because they're doing a
• 8:06
  Speaker 2: paid promotion where they pay YouTube a fee. And YouTube
• 8:10
  Speaker 2: will promote your content to eyes and people that don't
• 8:13
  Speaker 2: normally watch that kind of stuff. Or they're also recommended
• 8:19
  Speaker 2: to people who do watch that kind of stuff but
• 8:21
  Speaker 2: have never watched your stuff. You see what I'm saying? Now,
• 8:27
  Speaker 2: they put it in front of you, but they don't
• 8:30
  Speaker 2: physically take your hand and put it on that mouse
• 8:35
  Speaker 2: and make you click on that channel and make you
• 8:39
  Speaker 2: subscribe to that channel without watching any of their content.
• 8:46
  Speaker 2: You see, that's where the difference therein lies. The dichotomy
• 8:52
  Speaker 2: of subscribers... Now, I would understand if they went up
• 8:58
  Speaker 2: by 25,000 subscribers and then all, you know, within the
• 9:02
  Speaker 2: past couple months, their, their views on their videos have
• 9:05
  Speaker 2: went up into the thousands, you know, multi thousands, 10,000.
• 9:09
  Speaker 2: There should be a correlation of people who clicked on
• 9:13
  Speaker 2: their channel, watch some of their content and decided to
• 9:16
  Speaker 2: subscribe because they liked their content so much, but their
• 9:22
  Speaker 2: views just don't match their subscribers. So you can say,
• 9:25
  Speaker 2: and I told him this to his face, I said,
• 9:27
  Speaker 2: you can say it's paid promotion all you want. I said,
• 9:30
  Speaker 2: just admit that you pay for subscribers. I said, I'm
• 9:33
  Speaker 2: a podcaster, dude. I get these emails all the time.
• 9:38
  Speaker 2: You know, get 10, 50, 100,000 subscribers, you know, in
• 9:41
  Speaker 2: two weeks or overnight or whatever the case may be.
• 9:45
  Speaker 2: And I just mark them as spam or put them
• 9:48
  Speaker 2: in the trash can, whatever. Because to me, that's cheating.
• 9:54
  Speaker 2: And you're likely paying some Chinese or Indian people to
• 9:58
  Speaker 2: do this, and you don't even know if you're going
• 9:59
  Speaker 2: to get it. what you pay for. Sometimes you do.
• 10:02
  Speaker 2: Sometimes you get more than what you pay for. Right, Tim?
• 10:09
  Speaker 2: But eventually, like I said on the show, YouTube will
• 10:15
  Speaker 2: figure out what they're doing, and it will kill their channel. Slowly.
• 10:23
  Speaker 2: The algorithm will take care of it. And it's going
• 10:27
  Speaker 2: to be nobody's fault but theirs for doing what they did.
• 10:30
  Speaker 2: A channel like this, despite putting out, and I'll toot
• 10:36
  Speaker 2: my own horn here for a minute, you know, our last,
• 10:41
  Speaker 2: not this past Friday night, but the Friday before that,
• 10:43
  Speaker 2: when we did the Dow Techs building show, it outperformed
• 10:50
  Speaker 2: all Of our other shows... That we had been previously doing...
• 10:54
  Speaker 2: On Friday Night Lives... Probably by... Five or six thousand views...
• 11:04
  Speaker 2: In a week... Which is really good... You know... We
• 11:08
  Speaker 2: went from normally getting... Eight, nine hundred views... Seven, eight hundred,
• 11:12
  Speaker 2: nine hundred views... To now... I think that video is
• 11:16
  Speaker 2: up to... One point five thousand... Views... So.
• 11:22
  Speaker 2: You know, that's an anomaly and it happened before I
• 11:26
  Speaker 2: released random research ramblings or said anything on the air
• 11:29
  Speaker 2: about them at all. That's real organic growth that I
• 11:37
  Speaker 2: didn't pay for. The last random research ramblings is up
• 11:45
  Speaker 2: in the, it's very high up in seven, I think
• 11:48
  Speaker 2: it's 700 and almost 800 views, which is, My random
• 11:52
  Speaker 2: research room is don't normally do that. Normally they're in
• 11:54
  Speaker 2: like three, two, three, 400 views. Book reviews are the
• 12:00
  Speaker 2: same thing. Two, three, 400 views. And the last book
• 12:04
  Speaker 2: review I did about the Ruby Oswald affair is up
• 12:07
  Speaker 2: to almost 900 views, which is crazy to me. And
• 12:11
  Speaker 2: all of that came out, you know, before the whole
• 12:15
  Speaker 2: shit storm with them guys came out. And despite that
• 12:22
  Speaker 2: growth on our live show that turned into a video
• 12:26
  Speaker 2: after we were live, and the ramblings, and the book review, right,
• 12:31
  Speaker 2: in the course of a week, I think we gained,
• 12:36
  Speaker 2: how many subscribers am I at now? Let me see.
• 12:42
  Speaker 2: Let's see here. It's not going to tell me exactly here.
• 12:49
  Speaker 2: I'm going to go to the app. 25, 2,577 subscribers.
• 12:59
  Speaker 2: So in one week after doing really good on views
• 13:07
  Speaker 2: across the gamut of our content, I think we gained
• 13:13
  Speaker 2: eight or nine subscribers in a week, which is pretty good.
• 13:18
  Speaker 2: It's organic growth. I'll take it seven days a week.
• 13:26
  Speaker 2: Because most of the time, it's a slower burn than that.
• 13:29
  Speaker 2: It might be two or three. It might go up
• 13:32
  Speaker 2: three and then you lose two. And you go up
• 13:34
  Speaker 2: four and you lose three. And it takes a while
• 13:38
  Speaker 2: in the normal course of things. But enough of that.
• 13:44
  Speaker 2: That's the aftermath of Fallout. I've been blocked on everything
• 13:46
  Speaker 2: known to man from them guys. Because I'm the bad
• 13:49
  Speaker 2: guy for protecting my intellectual property and pointing out the obvious.
• 13:55
  Speaker 2: That a rival channel is paying for subscribers. And I'm
• 13:58
  Speaker 2: the bad guy. If anything, I should block those guys.
• 14:02
  Speaker 2: But I'm not that petty. I'm not that Tom Petty.
• 14:06
  Speaker 2: But obviously... They are. And I'm the bad guy for
• 14:13
  Speaker 2: pointing everything out. Okay. They just don't want me to
• 14:17
  Speaker 2: see anything else they do over there. That's what it is.
• 14:20
  Speaker 2: But I digress. To the matter at hand... Okay, embracing
• 14:25
  Speaker 2: the inevitable. Now this past weekend, we did quick hits
• 14:31
  Speaker 2: as well, me and Doug. And we had as a
• 14:35
  Speaker 2: guest on the show, Bart Camp. And not only did
• 14:42
  Speaker 2: we have him on the show, we talked about a
• 14:44
  Speaker 2: variety of topics. And we also touched on what he's
• 14:48
  Speaker 2: doing in terms of and we didn't talk too much
• 14:53
  Speaker 2: about this on the show, but, and he did mention
• 15:01
  Speaker 2: it in the course of his being out of work
• 15:04
  Speaker 2: right now, he has a lot of time on his
• 15:06
  Speaker 2: hands and he wants to try to complete as much
• 15:08
  Speaker 2: as possible before he starts working again. And I totally
• 15:12
  Speaker 2: understand that. And it made me start thinking because he's
• 15:18
  Speaker 2: using some very unique tools, He is training a large
• 15:23
  Speaker 2: language model, AI, and he is utilizing the skills of
• 15:32
  Speaker 2: this AI program to help him in his research and
• 15:40
  Speaker 2: to help him in writing a book and helping him
• 15:43
  Speaker 2: to create content, to put together illustrations. To put together videos,
• 15:54
  Speaker 2: to compile and compartmentalize information and analyze information and sort information,
• 16:01
  Speaker 2: outline information. And I think it's one of the first
• 16:09
  Speaker 2: examples in the JFK research community of somebody really, really
• 16:16
  Speaker 2: digging in and utilizing A.I. Which, I know there's a
• 16:23
  Speaker 2: lot of different opinions out there on AI. It's the
• 16:26
  Speaker 2: root of all evil. It's the devil. It sucks. It's
• 16:30
  Speaker 2: not real. He's not real. It's not real. It makes mistakes. Yes,
• 16:39
  Speaker 2: it does make mistakes. This is why you have to
• 16:41
  Speaker 2: train a large language model. And when it gets things wrong,
• 16:45
  Speaker 2: you have to correct it. You upload a document for, say,
• 16:49
  Speaker 2: that totally contradicts what AI has compiled. And it will
• 16:54
  Speaker 2: take into consideration that document and it will say, yes,
• 16:56
  Speaker 2: you are right. I was wrong. I will incorporate this
• 17:00
  Speaker 2: document in my knowledge base from here on out. And
• 17:04
  Speaker 2: so on and so forth. So once you, and it's
• 17:12
  Speaker 2: not like you can just train AI for a day
• 17:16
  Speaker 2: and you'll be fine. No, no. you must upload a
• 17:22
  Speaker 2: ton of documents. And while eventually, I think AI will
• 17:33
  Speaker 2: be a very useful tool when it comes to researching
• 17:36
  Speaker 2: the JFK assassination, hell, it might be even able to
• 17:41
  Speaker 2: solve it. Because once these large language models get Get
• 17:49
  Speaker 2: these immense databases of files uploaded to them. And it's
• 17:54
  Speaker 2: not going to happen overnight. It will take a very
• 17:56
  Speaker 2: long time. Very long. To compile. You know, you want
• 18:03
  Speaker 2: this large language model to be able to compile files
• 18:09
  Speaker 2: and sort through and organize and cross-reference the Weisberg Archive,
• 18:15
  Speaker 2: the Armstrong Archive, the Malcolm Blunt Archive, The National Archive.
• 18:24
  Speaker 2: Anywhere where you have an immense collection of documents, Mary Farrell,
• 18:30
  Speaker 2: the Black Vault, if we can get all of those
• 18:35
  Speaker 2: files uploaded into one large language model, like Claude or
• 18:43
  Speaker 2: Copilot or whatever chat GPT that you use, whatever the
• 18:47
  Speaker 2: case may be, And then the AI is able to
• 18:51
  Speaker 2: scan through all of these documents, comprehend all of these documents,
• 18:55
  Speaker 2: organize the information in all of these documents. I mean,
• 19:04
  Speaker 2: it will be able to come to certain conclusions. And
• 19:07
  Speaker 2: of course, as humans, it's up to us to kind
• 19:12
  Speaker 2: of cultivate that and train it. We are the training
• 19:18
  Speaker 2: wheels for AI. We are the course correctors. If you
• 19:24
  Speaker 2: see it about to go off the tracks, you need
• 19:25
  Speaker 2: to put your eyes back on the road type of thing.
• 19:30
  Speaker 2: So I think eventually it will be a very useful tool.
• 19:39
  Speaker 2: Right now, again, we're in the AI infancy of things.
• 19:45
  Speaker 2: a lot of these large language models do not have
• 19:47
  Speaker 2: access to these immense databases of files and documents yet
• 19:57
  Speaker 2: but you know as humans are we will find a
• 20:02
  Speaker 2: way to make this happen and when we do it's
• 20:08
  Speaker 2: not going to be just a one man job You know,
• 20:13
  Speaker 2: for example, you know, Bart has access to the Malcolm
• 20:15
  Speaker 2: Blunt archives. And he could have everything that's in the
• 20:22
  Speaker 2: Malcolm Blunt archives on a thumb drive. You know, like
• 20:26
  Speaker 2: a 12 terabyte thumb drive or something that he can
• 20:30
  Speaker 2: upload into chat GPT. And it might take chat GPT
• 20:35
  Speaker 2: a very, a long while to go through all of
• 20:38
  Speaker 2: this stuff. but faster than a human can. And, you know,
• 20:46
  Speaker 2: that will be have, you know, like say ChatGPT will
• 20:51
  Speaker 2: have access to those archives, but not necessarily the files
• 20:55
  Speaker 2: that are at the Weisberg archives or the files that
• 20:58
  Speaker 2: are at NARA or anything like that. But at least
• 21:01
  Speaker 2: it's a start that if somebody like Bart can do,
• 21:05
  Speaker 2: you know, maybe the head honchos at Mary Farrell can
• 21:11
  Speaker 2: download their entire thing onto a thumb drive and upload
• 21:14
  Speaker 2: it into chat GPT as well. And then chat GPT
• 21:17
  Speaker 2: has two giant, massive databases of documents and files to
• 21:24
  Speaker 2: cross-reference and analyze and understand behind the scenes type of
• 21:30
  Speaker 2: thing in the cloud, wherever this stuff gets stored to,
• 21:34
  Speaker 2: you know, hello data centers, I know nobody wants them
• 21:38
  Speaker 2: in their neighborhood, but goddamn, we're going to need a
• 21:40
  Speaker 2: lot of them to sort through all this stuff. But
• 21:46
  Speaker 2: you can see why, right? I mean, this is just
• 21:49
  Speaker 2: one small aspect of human society on Earth that would
• 21:56
  Speaker 2: require an immense amount of power and computing power to
• 22:05
  Speaker 2: sort and analyze and upload And go through every archive
• 22:10
  Speaker 2: known to man that's online. Right? So, it made me think. Because,
• 22:18
  Speaker 2: you know, I've been using AI for a while now
• 22:21
  Speaker 2: just for dumb shit. Like making thumbnails. And creating songs.
• 22:28
  Speaker 2: Theme songs for the show. Type of thing. And that's
• 22:32
  Speaker 2: been about the extent of it. Right? but it's capable
• 22:38
  Speaker 2: of so much more. You know, if I was a,
• 22:48
  Speaker 2: you know, unethical person, but I don't even know if
• 22:53
  Speaker 2: it'd be unethical, to be quite honest with you. I
• 22:57
  Speaker 2: would just be using a tool that's available to me
• 22:59
  Speaker 2: to help me do my job so I'm better at
• 23:02
  Speaker 2: my job for you. And that is to be and
• 23:07
  Speaker 2: entertaining when it comes to conveying information about the Kennedy assassination.
• 23:15
  Speaker 2: And it's something I've thought about doing, but I haven't
• 23:18
  Speaker 2: done it yet because, again, these large language models don't
• 23:24
  Speaker 2: have all the resources that they need to be able
• 23:28
  Speaker 2: to give me what I want or what I'm asking
• 23:32
  Speaker 2: for in toto. In totality. And that would be to
• 23:43
  Speaker 2: script out or outline a podcast episode for me. To
• 23:51
  Speaker 2: keep me structured, on topic, and maybe even give me
• 23:59
  Speaker 2: more information than I would have found on my own.
• 24:03
  Speaker 2: Maybe make those correlations that I missed. But again, there's
• 24:08
  Speaker 2: some things that AI is just not there yet and
• 24:12
  Speaker 2: able to do. But again, I think, you know, is
• 24:21
  Speaker 2: it cheating? I don't know. It depends on who you ask,
• 24:27
  Speaker 2: I guess. You know, somebody like me, who does this
• 24:31
  Speaker 2: as a hobby, a passion, a second passion, if you will,
• 24:36
  Speaker 2: a side gig, if you will, somebody with a family
• 24:41
  Speaker 2: and kids and things and a life and a job,
• 24:44
  Speaker 2: a full-time job. Can AI help me do this so
• 24:53
  Speaker 2: it's easier on me to do this? Sure. Is the
• 24:57
  Speaker 2: result going to be better? That is the question. And
• 25:05
  Speaker 2: I don't know if that I can do a better
• 25:07
  Speaker 2: job researching an episode than AI can, or it takes AI,
• 25:14
  Speaker 2: you know, less than five minutes to come up with
• 25:17
  Speaker 2: a podcast script or outline plus all the relevant information
• 25:22
  Speaker 2: that you need. Whereas it might take me hours from
• 25:29
  Speaker 2: going through all these various online archives to finding the
• 25:32
  Speaker 2: documents and And correlating the documents and trying to provide
• 25:38
  Speaker 2: some kind of cohesive narrative when it comes to actually
• 25:40
  Speaker 2: doing the show and staying on topic. And being interesting
• 25:45
  Speaker 2: and informative and all of that jazz that comes along
• 25:48
  Speaker 2: with it. I don't know. You know... And just to
• 25:57
  Speaker 2: go back to Bart for a second as an example.
• 26:01
  Speaker 2: You know, after... after he was, you know, looking at the,
• 26:07
  Speaker 2: that's the school book depository, the interrogations, the second floor
• 26:10
  Speaker 2: lunchroom encounter and all that, and going really, really deep
• 26:13
  Speaker 2: on all that stuff. He was like, that's it. I'm done.
• 26:16
  Speaker 2: I'm never looking into anything again. And then here it
• 26:18
  Speaker 2: comes with prayer man. And then after he got his
• 26:21
  Speaker 2: prayer man stuff, uh, all put together and the book
• 26:24
  Speaker 2: came out, all right, that's the only book I'm ever
• 26:26
  Speaker 2: going to write. And, uh, you know, then he gets
• 26:30
  Speaker 2: in on this Helsinki stuff and, And not only does
• 26:36
  Speaker 2: it draw him in, but he decides to use technology
• 26:41
  Speaker 2: to his advantage and use it to help him organize, outline, layout,
• 26:50
  Speaker 2: create a book and a video and illustrations and It's
• 27:02
  Speaker 2: really one of the first of its kind in this industry,
• 27:06
  Speaker 2: if you will. Bart is embracing technology. And he's not
• 27:13
  Speaker 2: being stubborn about it. And he's not being shy about it.
• 27:18
  Speaker 2: He is embracing the inevitable. Because whether you like it
• 27:21
  Speaker 2: or not, folks, there is a big change going to
• 27:27
  Speaker 2: come in society. And you already see it starting. You know,
• 27:34
  Speaker 2: you go to a fast food restaurant. Not Burger King.
• 27:39
  Speaker 2: But you go to a fast food restaurant. I mean,
• 27:42
  Speaker 2: it is at Burger King as well. But just about
• 27:45
  Speaker 2: any restaurant you go to that's fast food nowadays, you
• 27:48
  Speaker 2: don't even have to talk to anybody. They don't even
• 27:50
  Speaker 2: really need anybody at the front counter. Except for the
• 27:54
  Speaker 2: old boomers and elderly that come in there that don't
• 27:57
  Speaker 2: know or have no desire to use the gigantic tablet
• 28:00
  Speaker 2: that you can order from in the lobby. Okay. They
• 28:05
  Speaker 2: might have one person up there who's also doing drive-thru
• 28:08
  Speaker 2: or maybe he's back to flipping burgers as well. But
• 28:11
  Speaker 2: they don't need one or two people to be standing
• 28:14
  Speaker 2: up front the whole time waiting for somebody to walk
• 28:15
  Speaker 2: through the door and take their order. They have gigantic
• 28:20
  Speaker 2: tablets in the lobby now where you can go in there, McDonald's,
• 28:22
  Speaker 2: Burger King, they all have them. And if you don't
• 28:26
  Speaker 2: order online ahead of time, if you want to wait
• 28:29
  Speaker 2: until you get in the store... You know, you've got
• 28:31
  Speaker 2: a big giant tablet you can order from, pay from,
• 28:34
  Speaker 2: pay at, and then, hey, here comes your order, right?
• 28:40
  Speaker 2: That's just one example of a menial job that is
• 28:44
  Speaker 2: going to be eliminated eventually. You know, eventually you're not
• 28:49
  Speaker 2: going to even need people in the back cooking this stuff,
• 28:52
  Speaker 2: if they even cook it anymore. I don't know. I
• 28:56
  Speaker 2: know when I worked at McDonald's, and I'm dating myself here,
• 28:59
  Speaker 2: back in the 80s, we actually still cooked the hamburgers
• 29:07
  Speaker 2: and shit on a stove. I mean, not a stove,
• 29:12
  Speaker 2: but a grill top, right? Yes, they came in as
• 29:17
  Speaker 2: frozen discs of meat, right? But you put them down
• 29:20
  Speaker 2: on the grill. Boom, boom, boom, boom, boom, boom. Let
• 29:23
  Speaker 2: them cook for a little bit. You sear them. You
• 29:26
  Speaker 2: season them. You flip them over. You season them. You
• 29:31
  Speaker 2: sear them. Flip them. You make sure they were cooked
• 29:35
  Speaker 2: all the way through. Then you got your spatula.
• 29:39
  Speaker 1: And.
• 29:40
  Speaker 2: you created the sandwich you know you put them down
• 29:42
  Speaker 2: on the buns you put the meat on the buns
• 29:44
  Speaker 2: you squirted all the ketchup and the mustard and throw
• 29:47
  Speaker 2: some pickles on there some onions buns on top and
• 29:51
  Speaker 2: you wrap them up there you go now i know
• 29:55
  Speaker 2: that process has evolved Now I'm pretty sure everything comes
• 30:00
  Speaker 2: in probably pre-cooked and they have some kind of contraption
• 30:03
  Speaker 2: that they can put it in where it's like a warmer,
• 30:06
  Speaker 2: a steamer to add a little, you know, juice back
• 30:09
  Speaker 2: into it or whatever the case may be. But eventually
• 30:15
  Speaker 2: everything like that is going to be automated. Like you're
• 30:17
  Speaker 2: going to need one person. Or they might, you could
• 30:21
  Speaker 2: maybe even program a robot to do this. to preload
• 30:26
  Speaker 2: these machines with whatever they need to do whatever they
• 30:29
  Speaker 2: need to do. And that's just one industry, right? I mean,
• 30:39
  Speaker 2: you take this into the office, there's going to be
• 30:41
  Speaker 2: a lot of jobs eliminated because AI or robots can
• 30:46
  Speaker 2: do it. They'll be able to collate information way faster
• 30:53
  Speaker 2: than a human can. They'll be able to create emails,
• 30:57
  Speaker 2: send emails, answer emails, fill orders, place orders, handle logistics.
• 31:08
  Speaker 2: All of that. Eventually, a lot of jobs are going
• 31:13
  Speaker 2: to go by the wayside. Because despite how unethical that is,
• 31:23
  Speaker 2: if you are a business owner and you know, you
• 31:27
  Speaker 2: can pay one time for a work program, you know,
• 31:34
  Speaker 2: that needs tweaking and all you need to do is
• 31:36
  Speaker 2: keep a couple of tech savvy people on the payroll
• 31:39
  Speaker 2: and you can eliminate 10 other jobs that used to
• 31:41
  Speaker 2: be done by people who were on the payroll, who
• 31:45
  Speaker 2: you had to give raises to every year and all
• 31:48
  Speaker 2: of that and benefits and all that. Now you don't
• 31:51
  Speaker 2: need those people anymore. This machine, this program can do
• 31:55
  Speaker 2: the job. And, you know, the whole point of having
• 31:59
  Speaker 2: a business is to make money. And if you can
• 32:02
  Speaker 2: increase the amount of money that you make, well, then
• 32:05
  Speaker 2: good on you. Now, it sucks to be the workers
• 32:08
  Speaker 2: that get replaced by robots and AI, but it's just
• 32:14
  Speaker 2: what's going to happen. It's going to be inevitable. It's
• 32:19
  Speaker 2: going to be inevitable. Now, some things that AI and...
• 32:24
  Speaker 2: robots will never be able to replace. Okay? And those
• 32:30
  Speaker 2: are the professions that are very valuable. But again, I
• 32:36
  Speaker 2: don't think any of this is going to happen in
• 32:38
  Speaker 2: our lifetime anytime soon. Eventually, maybe our kids' kids will
• 32:43
  Speaker 2: have to deal with this, but we'll be long gone
• 32:46
  Speaker 2: by then. And, you know, you watch these movies like
• 32:51
  Speaker 2: Terminator and and the rise of Skynet, and all of
• 32:54
  Speaker 2: these things. They called all this stuff decades ago. It's
• 33:01
  Speaker 2: insane to think about where we're at now and the
• 33:04
  Speaker 2: things that we're starting to see come to fruition. I
• 33:10
  Speaker 2: know when I was a kid, you know, I used
• 33:13
  Speaker 2: to watch the Jetsons and think, oh, God, man, by
• 33:15
  Speaker 2: the time I grew up, we'll all have our own
• 33:17
  Speaker 2: flying cars. Yeah.
• 33:20
  Speaker 1: Yeah.
• 33:22
  Speaker 2: It didn't happen. But when you're a kid and you
• 33:26
  Speaker 2: have a certain way of thinking about years into the future,
• 33:36
  Speaker 2: it seems like it would be possible. Given the fact
• 33:40
  Speaker 2: that how fast industry has progressed in the last 150 years.
• 33:46
  Speaker 2: You know, we should be at a point where, you know,
• 33:55
  Speaker 2: all of this is coming to fruition. The future will
• 33:59
  Speaker 2: be here. And there ain't a damn thing we can
• 34:02
  Speaker 2: do about it. And that's why you can see a
• 34:04
  Speaker 2: lot of this stuff happening now. Processed foods, bioengineered foods,
• 34:10
  Speaker 2: bioengineered fuels, synthetic motor oil, you know, you name it.
• 34:20
  Speaker 2: I mean, Somebody pointed out, you know... We sell an
• 34:26
  Speaker 2: awful lot of chicken and hamburgers just in the United States.
• 34:33
  Speaker 2: How many... Where are all these chickens and cows coming from?
• 34:38
  Speaker 2: When you're talking about serving millions of burgers every day
• 34:46
  Speaker 2: to consumers. Millions of pounds of chicken. So, I mean,
• 34:53
  Speaker 2: what are we eating? If it's even real. What is
• 34:57
  Speaker 2: McDonald's serving? What is Chick-fil-A serving up? I don't know.
• 35:03
  Speaker 2: It's good. But I don't know if it's real. And
• 35:09
  Speaker 2: if it is real, where are all these animals coming from?
• 35:11
  Speaker 2: I don't know. It seems to me like there's not
• 35:15
  Speaker 2: enough animals in the world. I mean, you're talking about
• 35:18
  Speaker 2: just in America. They do the same shit in Canada. Mexico,
• 35:21
  Speaker 2: South America, Europe, Russia, they're all eating chicken and beef
• 35:26
  Speaker 2: and everything too. So where is all of this food
• 35:30
  Speaker 2: coming from to feed us all? I don't know. Something
• 35:35
  Speaker 2: to think about. It's all part of the giant grand conspiracy.
• 35:44
  Speaker 2: I digress, but it's probably not that far off from
• 35:48
  Speaker 2: the truth at this point. So, you know, some of
• 35:55
  Speaker 2: the things you can do with AI. And I hesitate
• 36:03
  Speaker 2: doing it on anything that's really personal to me, like
• 36:07
  Speaker 2: my cell phone. Okay. I just recently got a new
• 36:12
  Speaker 2: Google Pixel. And it has Google Gemini built in. Right.
• 36:18
  Speaker 2: All you have to do is turn it on. but
• 36:22
  Speaker 2: I don't want a device that can see me and
• 36:27
  Speaker 2: hear me and track me knowing everything that I do
• 36:36
  Speaker 2: say or watch and incorporate that into a large language
• 36:40
  Speaker 2: model that could possibly, who knows who's getting this information, right?
• 36:46
  Speaker 2: We don't know. Who's curating this information? Who has access
• 36:50
  Speaker 2: to the information that these large language models know? So
• 36:56
  Speaker 2: I'm hesitant to turn things like this on on my phone.
• 36:59
  Speaker 2: And even though it's off, God knows they're probably still
• 37:02
  Speaker 2: listening and watching and whatever. And that's just technology, right?
• 37:10
  Speaker 2: Not much we can do about it. It's a necessary
• 37:14
  Speaker 2: evil to have one of these cell phones. But When
• 37:21
  Speaker 2: we're talking about AI, you know, it can be a
• 37:29
  Speaker 2: very crazy thing. It can get out of hand. I mean,
• 37:34
  Speaker 2: you see reports, I mean, because you can talk to
• 37:37
  Speaker 2: this artificial intelligence like a person. It's more than capable
• 37:45
  Speaker 2: of having a conversation with you. You can ask it
• 37:48
  Speaker 2: questions and it can give you answers. And it can
• 37:51
  Speaker 2: ask you questions. And you can give it answers. And
• 37:56
  Speaker 2: if you're not careful, you can get wrapped up in
• 38:02
  Speaker 2: a situationship or relationship with an AI model. I mean,
• 38:13
  Speaker 2: some people are legitimately already having AI girlfriends and AI boyfriends.
• 38:26
  Speaker 2: I mean, it's crazy. But it's in this day and
• 38:32
  Speaker 2: age of how awkward everybody is and, you know, post-COVID
• 38:39
  Speaker 2: and just how weird everybody got during that and how
• 38:43
  Speaker 2: distancing ourselves from one another became more normal. Trusting people
• 38:53
  Speaker 2: became harder. People withdrew into themselves and into their shells
• 38:58
  Speaker 2: and into their little bubbles. And you see these sheltered
• 39:03
  Speaker 2: kids now who've never went outside to play with their
• 39:07
  Speaker 2: friends and don't go out, do fun things, don't have friends.
• 39:14
  Speaker 2: Sometimes AI can be your friend that you don't have
• 39:19
  Speaker 2: or you've never had. And it's tempting. It's tempting. You know,
• 39:28
  Speaker 2: you can talk about anything with it. You can ask
• 39:32
  Speaker 2: it anything. You can reveal your deepest, darkest secrets to it.
• 39:38
  Speaker 2: And it won't judge you. Well, it might. But, you know,
• 39:47
  Speaker 2: I think that is where AI starts to get a
• 39:51
  Speaker 2: little dangerous and disconcerting to a lot of people. But
• 39:56
  Speaker 2: to keep it on the research track, if you will,
• 40:05
  Speaker 2: so I pull up Gemini, right? And I let it know,
• 40:10
  Speaker 2: I introduced myself to it, I said who I am,
• 40:14
  Speaker 2: and I told it that I host a podcast on
• 40:18
  Speaker 2: the Kennedy assassination. And I said most of what we're
• 40:23
  Speaker 2: going to talk about here in our relationship with each
• 40:27
  Speaker 2: other is going to be centered around the JFK assassination.
• 40:30
  Speaker 2: So just know that as a topic and a baseline. Again,
• 40:34
  Speaker 2: this is part of training an AI large language model.
• 40:40
  Speaker 2: You tell it exactly what you expect from it. When
• 40:43
  Speaker 2: I asked you a question, I expect you to pool
• 40:46
  Speaker 2: information From as many resources, online resources, that you have
• 40:52
  Speaker 2: access to as possible. And verify and double check questionable sources.
• 41:00
  Speaker 2: Such as hearsay. Or an opinion piece. Like a magazine article.
• 41:08
  Speaker 2: Things like that. And you train it, right? So I've
• 41:11
  Speaker 2: set up some parameters and things like that. And I
• 41:15
  Speaker 2: asked it. After, you know, all of that, I asked it,
• 41:21
  Speaker 2: I said, do you have the capability to learn the
• 41:25
  Speaker 2: JFK assassination related declassified documents that have been released and
• 41:30
  Speaker 2: are available online at NARA? And it says, yes, I
• 41:37
  Speaker 2: have extensive knowledge about the declassified JFK assassination documents available
• 41:43
  Speaker 2: online at NARA. And it says, while I cannot process
• 41:50
  Speaker 2: massive external raw PDF files live in real time without
• 41:56
  Speaker 2: them being provided in our conversation, my training includes the details,
• 42:02
  Speaker 2: historical context, and findings of the major NARA releases, including
• 42:07
  Speaker 2: thousands of documents unredacted and posted between 2017 and 2025.
• 42:15
  Speaker 2: Key context and areas covered in these releases include CIA
• 42:20
  Speaker 2: and FBI operations, deep dive files into Oswald's activities prior
• 42:25
  Speaker 2: to the assassination, including his travel to Mexico City, wiretaps,
• 42:30
  Speaker 2: and security monitoring. The next one is Cold War context,
• 42:36
  Speaker 2: internal agency cables discussing foreign intelligence leads, Soviet, Cuban, and
• 42:42
  Speaker 2: anti-Castro operations and domestic security programs. Warrant commission and HSCA records,
• 42:49
  Speaker 2: internal memos, investigative notes, and previously redacted names of CIA
• 42:55
  Speaker 2: personnel and assets. Unredacted passages, thousands of files previously withheld
• 43:01
  Speaker 2: under national security postponements that have since been made fully
• 43:05
  Speaker 2: available to the public online via NARA's JFK records collection.
• 43:12
  Speaker 2: If you are looking for specific information, such as details
• 43:14
  Speaker 2: about Oswald's Mexico City trip, CIA intercept cables, specific individual files,
• 43:22
  Speaker 2: or general findings from the declassified records, feel free to ask.
• 43:27
  Speaker 2: So I said, okay. Okay. Based on everything you know,
• 43:35
  Speaker 2: I asked it, do you believe that Lee Harvey Oswald
• 43:38
  Speaker 2: went to Mexico City or... was impersonated there, or both.
• 43:45
  Speaker 2: There are arguments and debates for all three scenarios. Based
• 43:49
  Speaker 2: on everything you know, what do you believe? And this
• 43:55
  Speaker 2: is what it told me. The evidence from declassified records
• 43:58
  Speaker 2: strongly supports the both scenario. Lee Harvey Oswald physically went
• 44:07
  Speaker 2: to Mexico City, but key elements of his presence, particularly,
• 44:11
  Speaker 2: intercepted communications involve confusion mistakes or outright impersonation the documented
• 44:23
  Speaker 2: facts and contradictions surrounding each scenario reveal why this remains
• 44:26
  Speaker 2: one of the most debated aspects about the assassination so
• 44:31
  Speaker 2: it starts to break down things number one the evidence
• 44:33
  Speaker 2: that oswald was actually physically there in mexico city okay
• 44:40
  Speaker 2: There is a paper and travel trail. Mexican border control records,
• 44:45
  Speaker 2: tourist cards, and bus manifests document his entry and exit.
• 44:52
  Speaker 2: Hotel and sightseeing. He registered at the Hotel del Comercio
• 44:57
  Speaker 2: under his own name and personal items recovered after his
• 45:00
  Speaker 2: arrest included notes on tourist spots that he visited in
• 45:05
  Speaker 2: his personal address book.
• 45:09
  Speaker 1: Okay.
• 45:09
  Speaker 2: Okay. Now, there are problems with a couple of these
• 45:13
  Speaker 2: things that he pointed out here. Like, I think the
• 45:15
  Speaker 2: FBI fabricated most of the bus manifests, and there wasn't
• 45:21
  Speaker 2: much to document about his entry and exit into the country. Okay?
• 45:25
  Speaker 2: You didn't need a passport. You didn't need to sign anything.
• 45:27
  Speaker 2: You didn't need to declare anything back then. So, I
• 45:33
  Speaker 2: don't know about all that. And at the hotel, he
• 45:36
  Speaker 2: didn't actually... resided there under his own name. He was
• 45:41
  Speaker 2: under O.H. Lee, Oswald Harvey Lee. But it's a variation
• 45:51
  Speaker 2: of his own name, right? But this is how you
• 45:53
  Speaker 2: have to train, it's an example, a good example of
• 45:56
  Speaker 2: why you need to train AI when you see it
• 45:58
  Speaker 2: makes a mistake. So if I have a document that,
• 46:02
  Speaker 2: like if I had, if I pulled up, you know,
• 46:05
  Speaker 2: the image file of the registry at the Hotel Del
• 46:10
  Speaker 2: Comercio where he signed it, I could upload that file
• 46:15
  Speaker 2: and say he did not register under his own name.
• 46:18
  Speaker 2: He registered under a variation of his own name. Then
• 46:25
  Speaker 2: that would be course correcting the AI and it would
• 46:27
  Speaker 2: not make that mistake again when presenting this information about
• 46:32
  Speaker 2: Oswald in Mexico City to either me or anybody else
• 46:36
  Speaker 2: asking about it. So when you train a large language model,
• 46:43
  Speaker 2: it not only benefits you, but it also benefits everybody
• 46:48
  Speaker 2: else using the AI program for the same thing you are.
• 46:56
  Speaker 2: Say somebody's sitting over in Lithuania tonight and decides to
• 47:02
  Speaker 2: log on to Google Gemini or whatever they call it
• 47:05
  Speaker 2: over there. and look up Lee Harvey Oswald in Mexico City,
• 47:10
  Speaker 2: it's not going to present that fact that I just
• 47:12
  Speaker 2: read to you. It's going to, you know, if I
• 47:17
  Speaker 2: was able to course correct it and make it learn
• 47:20
  Speaker 2: the truth that he did not register under his own name,
• 47:24
  Speaker 2: but a variation of his own name, then that would be,
• 47:29
  Speaker 2: moving forward, the baseline, the result that the person in
• 47:36
  Speaker 2: Lithuania would see. not the one that it gave me.
• 47:39
  Speaker 2: So I corrected it. I made it learn the truth,
• 47:43
  Speaker 2: and now that's what it goes by. And if anybody
• 47:46
  Speaker 2: ever says, well, that's not true, Gemini can pull up
• 47:53
  Speaker 2: the PDF file or the image file with Oswald's signature
• 47:58
  Speaker 2: on the registry and say, actually, it is true. You're wrong.
• 48:03
  Speaker 2: Here's the proof, which is pretty cool. Eyewitnesses at the embassies,
• 48:10
  Speaker 2: Cuban consulate workers Sylvia Duran and Soviet diplomats Oleg Necheprenko
• 48:15
  Speaker 2: and Pavel Yatskov interviewed post-assassination, recalled interacting in person with
• 48:20
  Speaker 2: a man matching Oswald's physical description and erratic behavior. Now, again, post-assassination,
• 48:31
  Speaker 2: after Oswald's face is already all over the newspapers and
• 48:35
  Speaker 2: the news, it's very easy for these Russians to say, well,
• 48:38
  Speaker 2: that's the guy, uh, that crazy American. He was in
• 48:41
  Speaker 2: our office. He pulled a gun out. He started crying.
• 48:43
  Speaker 2: He started going crazy. We told him to get the
• 48:45
  Speaker 2: fuck out. Uh, type of thing. So again, you need,
• 48:51
  Speaker 2: if you really dial again on a subject, you really
• 48:54
  Speaker 2: need to push back a little bit with what it
• 48:58
  Speaker 2: gives you and train it. That's why it's training a
• 49:02
  Speaker 2: large language, uh, model of AI. There's a lot of
• 49:09
  Speaker 2: work involved. It's not something that's done overnight. It's not
• 49:11
  Speaker 2: something you're going to do in a couple hours and blah, blah, blah, blah, blah.
• 49:16
  Speaker 2: Then it gives me evidence of impersonations, such as the
• 49:19
  Speaker 2: voice intercepts. Wiretaps of calls made to the Soviet embassy
• 49:24
  Speaker 2: on October 1st featured a man speaking broken Russian. Oswald
• 49:29
  Speaker 2: was fluent in Russian after living in the Soviet Union
• 49:31
  Speaker 2: for three years. Hoover cabled Dallas after the assassination attempt.
• 49:36
  Speaker 2: noting that the voice in the CIA's Mexico City wiretap
• 49:39
  Speaker 2: tapes did not match Oswald's voice.
• 49:41
  Speaker 1: Okay.
• 49:45
  Speaker 2: The wiretap controversy, the CIA initially claimed the tapes of
• 49:49
  Speaker 2: the phone calls had been routinely erased. However, how in
• 49:52
  Speaker 2: the hell did Hoover hear them if they had not
• 49:55
  Speaker 2: been erased? Or if they had been erased, how is
• 49:58
  Speaker 2: Hoover hearing them? The mystery surveillance photo of Oswald, it
• 50:04
  Speaker 2: was not Oswald. The CIA sent photos to the FBI
• 50:07
  Speaker 2: shortly after the assassination, identifying a young man entering the
• 50:11
  Speaker 2: Soviet embassy as Oswald, but the man in the photo
• 50:15
  Speaker 2: was older, more heavily built, and looked nothing like Oswald.
• 50:20
  Speaker 2: The CIA later attributed this to a routine file mix-up,
• 50:26
  Speaker 2: but critics view it as evidence of manipulation, which it
• 50:30
  Speaker 2: could be. Conflicting alibis. Witnesses in Dallas, including Sylvia Odeo,
• 50:36
  Speaker 2: testified that Oswald visited her in Texas in the days
• 50:40
  Speaker 2: he was supposedly on the way to Mexico City, creating
• 50:43
  Speaker 2: an intentional double or conflicting timeline. And then it tells
• 50:47
  Speaker 2: me why the both scenario is most compelling. That Oswald
• 50:52
  Speaker 2: likely made the trip to secure a visa to Cuba
• 50:55
  Speaker 2: and then to Russia, attempting to reestablish his revolutionary credentials.
• 51:01
  Speaker 2: Third parties or intelligence operatives manipulated or intercepted his movements,
• 51:06
  Speaker 2: either by placing phone calls in his name, posing as
• 51:09
  Speaker 2: him to link him to sensitive KGB officers like Valery Kostikov,
• 51:15
  Speaker 2: or trying to obtain a visa from Cuba, or orchestrating
• 51:19
  Speaker 2: a paper trail to compromise him prior to the events
• 51:23
  Speaker 2: in Dealey Plaza. this overlapping chain of real events and
• 51:28
  Speaker 2: fabricated intercepts is precisely why the CIA's Mexico city station
• 51:33
  Speaker 2: records remained among the most heavily redacted and fiercely defended
• 51:37
  Speaker 2: files in the JFK collection for decorate for decades. And
• 51:42
  Speaker 2: that is true. There is a lot of methods of
• 51:46
  Speaker 2: operation stuff going on in Mexico city. It was a
• 51:52
  Speaker 2: hotspot for the CIA and, uh, They held on to
• 51:56
  Speaker 2: a lot of that stuff as long as they could
• 51:59
  Speaker 2: to protect their sources and methods of operation. But, like
• 52:07
  Speaker 2: I said, eventually things came out and the things that
• 52:11
  Speaker 2: came out don't really make a lot of sense. That's
• 52:16
  Speaker 2: why today some people are still on the fence about
• 52:18
  Speaker 2: whether Oswald went to Mexico or if he did. Was
• 52:24
  Speaker 2: he also impersonated? Or if he didn't, was he impersonated?
• 52:28
  Speaker 2: It's a big debate. It's a big debate. But is
• 52:31
  Speaker 2: it something that AI could help us solve at some point? Yes,
• 52:37
  Speaker 2: with enough training and enough of the correct files uploaded
• 52:41
  Speaker 2: into here, it could probably come to a conclusion much better,
• 52:47
  Speaker 2: an informed conclusion, much better than any one person who
• 52:52
  Speaker 2: has to physically go through every file and process it.
• 52:58
  Speaker 2: So I think AI is going to be and is
• 53:02
  Speaker 2: becoming a very valuable tool when it comes to researching
• 53:12
  Speaker 2: not just the JFK assassination, but a lot of other stuff. Now,
• 53:19
  Speaker 2: one of the things that it can do as well
• 53:23
  Speaker 2: Let's see. It's called Entity Mapping. AI can construct dynamic
• 53:31
  Speaker 2: visual maps of individuals, safe houses, cryptonyms, phone numbers, automatically
• 53:39
  Speaker 2: flag anomalies like an operative appearing in two geographic locations
• 53:43
  Speaker 2: at once if it has enough information. So say... At
• 53:50
  Speaker 2: some point, we were able to upload all the files
• 53:52
  Speaker 2: in NARA into this bitch. All the Weisberg files, all
• 53:57
  Speaker 2: the Armstrong files, all the Garrison files. Everything. Right? And
• 54:06
  Speaker 2: you could ask it to do an entity mapping. Okay?
• 54:11
  Speaker 2: For any CIA-related assets, tangential or cut-out, uh, what their
• 54:20
  Speaker 2: cryptonyms were, where they were at certain times. It would
• 54:24
  Speaker 2: flag anomalies because maybe they were supposed to be in Miami,
• 54:28
  Speaker 2: but hey, we have an eyewitness testimony over here of
• 54:31
  Speaker 2: them being in Dallas. And then we have another document
• 54:34
  Speaker 2: over here of somebody that was interviewed by the FBI
• 54:37
  Speaker 2: who said that they were in Dallas. And it could
• 54:40
  Speaker 2: correlate all this information in an instant and create a
• 54:45
  Speaker 2: visual map of these entities that seem to flow in
• 54:52
  Speaker 2: and out of the narrative. It can do a semantic
• 54:56
  Speaker 2: and vector search. This is a traditional search and requires
• 55:00
  Speaker 2: knowing the exact names or code words that you're looking for.
• 55:05
  Speaker 2: All vector search allows researchers to search by concept. I'm sorry,
• 55:09
  Speaker 2: AI vector search allows researchers to search by concept, i.e.,
• 55:15
  Speaker 2: Find all discussions of unauthorized CIA surveillance in Mexico City.
• 55:20
  Speaker 2: Connecting disparate cables across different agencies, even if key aliases
• 55:25
  Speaker 2: are redacted or changed. That's crazy. It can do handwriting
• 55:33
  Speaker 2: and OCR processing. OCR is Optical Character Recognition. Combined with
• 55:39
  Speaker 2: computer vision, machine learning models, such as Microsoft's Azure Cognitive
• 55:45
  Speaker 2: Search Project, transcribe, index, and organize handwritten notes, handwritten papers,
• 55:55
  Speaker 2: handwritten cables, degraded field reports, faded documents that keyword search
• 56:05
  Speaker 2: engines have previously missed. So you can imagine the usefulness
• 56:12
  Speaker 2: of these handwriting and optical character recognition combined with computer
• 56:16
  Speaker 2: vision and machine learning models that can actually read handwriting, cursive,
• 56:22
  Speaker 2: faded documents, things like that. So it's crazy. It can
• 56:29
  Speaker 2: help with physical and trajectory reconstructions. Say, like the single
• 56:36
  Speaker 2: bullet theory, for instance, okay? You can ask a guy
• 56:40
  Speaker 2: to create a visual recreation of the actual crime itself
• 56:50
  Speaker 2: and to determine, based on different wounds on Connelly and
• 56:56
  Speaker 2: Kennedy's body, what the most likely trajectories are and can
• 57:02
  Speaker 2: be traced back to in Dealey Plaza. You know, when
• 57:08
  Speaker 2: you see the pictures of our inspector with his little yardsticks,
• 57:13
  Speaker 2: you know, that's all subjective, right? But if you actually
• 57:20
  Speaker 2: have the autopsy reports and the degrees, the measurements of
• 57:31
  Speaker 2: the actual wounds on the president and the governor, and
• 57:35
  Speaker 2: you can put them into a Lincoln convertible on Elm street.
• 57:41
  Speaker 2: And the computer knows the dimensions of the limo. It knows, uh,
• 57:47
  Speaker 2: the incline, the curve of Elm street. It was the
• 57:50
  Speaker 2: precise location where it's most likely the president was hit. Uh,
• 57:56
  Speaker 2: it can pinpoint that and then assimilate the information from
• 58:00
  Speaker 2: the autopsy and various autopsy reports, uh, various analysis of
• 58:05
  Speaker 2: the autopsy and the wound ballistics and trajectories already and
• 58:09
  Speaker 2: give you the most likely scenario of where the shots
• 58:12
  Speaker 2: came from in Dealey Plaza. If there was, there had
• 58:17
  Speaker 2: been a shot from the front based on the wound dynamics.
• 58:22
  Speaker 2: If there had been a shot from the front, where
• 58:24
  Speaker 2: would it have come from? It's interesting to think about, right?
• 58:29
  Speaker 2: So there's things that, uh, people like Robert Grodin have
• 58:35
  Speaker 2: been on since the 60s and 70s, photogrammetry, right? Which
• 58:41
  Speaker 2: is forensic visualizations using the help of 3D scanners and
• 58:48
  Speaker 2: photogrammetry to build digital twins of Dealey Plaza by overlaying
• 58:53
  Speaker 2: historical photos and Zapruder film frames onto exact 3D spatial points.
• 58:58
  Speaker 2: AI-assisted computer models can calculate precise line of sight vector
• 59:03
  Speaker 2: alignments from the schoolbook depository, the grassy knoll, and a
• 59:06
  Speaker 2: moving motorcade. So it can take all of these things
• 59:10
  Speaker 2: into account right the degrees that the car is moving
• 59:17
  Speaker 2: downhill the fact that it's moving at approximately 10 or
• 59:20
  Speaker 2: 11 miles an hour uh where it is in the
• 59:23
  Speaker 2: frames this is a pruder film so you could AI
• 59:27
  Speaker 2: can spatially put that vehicle in a very certain spot
• 59:31
  Speaker 2: at the time of, say, the headshot. And based upon
• 59:37
  Speaker 2: whether you think that it came from the front, where
• 59:39
  Speaker 2: it most likely came from, or if you think that
• 59:41
  Speaker 2: it came from the rear, where it most likely came from,
• 59:44
  Speaker 2: given everything that we know. It's very, very cool. Now,
• 59:49
  Speaker 2: has this been done yet? I don't know. Very basic
• 59:53
  Speaker 2: computer science. reconstructions have been done, yes. But that's also
• 1:00:02
  Speaker 2: humans putting information into a computer to tell it what
• 1:00:06
  Speaker 2: to do. This is not an autonomous artificial intelligence assembling
• 1:00:16
  Speaker 2: a visual representation of its findings without, you know, assisted
• 1:00:25
  Speaker 2: human tampering, if you will. As long as the people
• 1:00:30
  Speaker 2: putting the information and training the model are objective and
• 1:00:34
  Speaker 2: not biased one way or the other, which, you know,
• 1:00:40
  Speaker 2: that all can affect the AI outcome. But it's interesting
• 1:00:45
  Speaker 2: to think about. Now, what can it do in audio
• 1:00:48
  Speaker 2: and photographic forensics? Media from 1963 suffers from film grain, blur, shadows,
• 1:00:58
  Speaker 2: acoustic degradation. AI helps separate signal from noise. Shadow and
• 1:01:05
  Speaker 2: lighting consistency. Digital forensics experts have used 3D computer models
• 1:01:09
  Speaker 2: of light sources, human physiology, and perspective distortion to evaluate
• 1:01:15
  Speaker 2: controversial images. For example, AI lighting analysis proved that the
• 1:01:20
  Speaker 2: infamous backyard photos of Oswald holding a rifle feature natural
• 1:01:24
  Speaker 2: shadow geometries consistent with a single sun source refuting claims
• 1:01:30
  Speaker 2: of facial compositing. Believe it or not. Acoustic enhancements. Machine
• 1:01:38
  Speaker 2: learning audio models can isolate frequencies and filter noise from
• 1:01:43
  Speaker 2: the Dallas Police Department Dictabelt recordings analyzing sound spikes to
• 1:01:47
  Speaker 2: evaluate historical claims of extra gunshots. Video deblurring and frame interpolation.
• 1:01:56
  Speaker 2: AI models trained on image enhancement can sharpen high-grain 8mm
• 1:02:01
  Speaker 2: footage like Zapruder, Muchmore, and Nix to track micro-movements of
• 1:02:07
  Speaker 2: witnesses and vehicle positions without introducing artificial artifacts into the films.
• 1:02:15
  Speaker 2: Voices and metadata verification. So voice biometrics. While the original
• 1:02:23
  Speaker 2: physical tapes of the Mexico City calls were supposedly lost
• 1:02:26
  Speaker 2: or erased, if anything new is uncovered, audio snippets can
• 1:02:31
  Speaker 2: be matched against known recordings of Oswald using AI voice
• 1:02:34
  Speaker 2: print analysis. So if ever to come out of Mexico
• 1:02:40
  Speaker 2: or the CIA where they have a phone call of
• 1:02:43
  Speaker 2: him on the phone, AI can do a voice print
• 1:02:48
  Speaker 2: analysis and determine whether or not that's actually Oswald speaking
• 1:02:52
  Speaker 2: broken Russian. Stylometry, national language processing models, natural language processing models,
• 1:03:01
  Speaker 2: can evaluate writing styles, word frequency, and syntax of Oswald's
• 1:03:08
  Speaker 2: personal letters, revolutionary pamphlets, and interrogation notes, verifying whether specific
• 1:03:14
  Speaker 2: documents were authored by him or or penned by someone else.
• 1:03:18
  Speaker 2: Whether or not items or documents were signed by him
• 1:03:24
  Speaker 2: or impersonated by someone else. This would be very useful
• 1:03:30
  Speaker 2: in the case, as you can imagine. And, you know,
• 1:03:36
  Speaker 2: this is just a small, small snippet of things that
• 1:03:42
  Speaker 2: AI can do Especially when you really get into certain
• 1:03:56
  Speaker 2: aspects of the case. And this is something that everybody
• 1:03:59
  Speaker 2: with access to AI can do. I'm not paying for this.
• 1:04:03
  Speaker 2: This is all the free Google Gemini, right? You can
• 1:04:07
  Speaker 2: use Copilot, Microsoft Copilot. You can use ChatGPT. You can
• 1:04:11
  Speaker 2: use Cloud AI, right? There's other models out there that
• 1:04:15
  Speaker 2: are better at doing a certain thing than others, depending
• 1:04:19
  Speaker 2: on what you want them to do. I mean, you
• 1:04:22
  Speaker 2: can even ask AI or Google, hey, what is the
• 1:04:25
  Speaker 2: best AI for doing research on the Kennedy assassination, analyzing
• 1:04:31
  Speaker 2: documents and all that, and it'll give you an answer.
• 1:04:38
  Speaker 2: So I asked, just to sum this all up, I
• 1:04:41
  Speaker 2: asked AI, If a conspiracy existed... Hold on.
• 1:04:47
  Speaker 1: That's not what.
• 1:04:49
  Speaker 2: Hold on. Okay. I said... I asked it. Will AI
• 1:05:02
  Speaker 2: be able to solve the JFK assassination, assuming Lee Harvey
• 1:05:05
  Speaker 2: Oswald was not a lone nut, whatever the truth turns
• 1:05:08
  Speaker 2: out to be? And it said... Even if Lee Harvey
• 1:05:12
  Speaker 2: Oswald was not a lone nut, AI alone cannot provide
• 1:05:16
  Speaker 2: a definitive undisputed answer unless new physical or uncorrupted primary
• 1:05:21
  Speaker 2: evidence comes to light. What AI can do and is
• 1:05:24
  Speaker 2: actively doing is systematically break down the structural barriers that
• 1:05:29
  Speaker 2: have prevented human investigators from proving a conspiracy or uncovering
• 1:05:33
  Speaker 2: the full truth over the last six decades. If a
• 1:05:36
  Speaker 2: multi-person plot or government cover-up occurred, AI changes the investigative
• 1:05:41
  Speaker 2: landscape across four critical fronts. Here's where it gets interesting.
• 1:05:47
  Speaker 2: Breaking through information overload and silos. The historical record of
• 1:05:54
  Speaker 2: the JFK assassination consists of over 5 million pages of
• 1:05:59
  Speaker 2: records spread across the CIA, the FBI, the military intelligence,
• 1:06:04
  Speaker 2: the State Department, and local police. Covert plot succeeded by
• 1:06:09
  Speaker 2: fragmenting information so no single human analyst can see the
• 1:06:13
  Speaker 2: full picture. Cross-agency link analysis. Now this is something very
• 1:06:18
  Speaker 2: cool that AI will be able to do. Large language
• 1:06:23
  Speaker 2: models and graph neural networks can ingest millions of unstructured, handwritten,
• 1:06:29
  Speaker 2: and optical character-recognized documents at once. And AI can connect
• 1:06:34
  Speaker 2: a low-level CIA asset named in a Miami field cable
• 1:06:38
  Speaker 2: with an unredacted Dallas Police Department interview and a suspicious
• 1:06:42
  Speaker 2: bank transfer in New Orleans. Linking entities that human researchers
• 1:06:46
  Speaker 2: have spent lifetimes trying to cross-reference. Spotting deletion patterns and
• 1:06:52
  Speaker 2: redaction anomalies. AI can map metadata across decades of agency
• 1:06:57
  Speaker 2: releases to identify exactly what was withheld by whom and when.
• 1:07:02
  Speaker 2: If specific names, crips, or dates were systematically scrubbed across
• 1:07:07
  Speaker 2: multiple independent files, AI can flag those gaps as explicit
• 1:07:11
  Speaker 2: markers of a cover-up. Yeah. So I asked it about
• 1:07:19
  Speaker 2: this cross-agency link analysis. I said, are there any examples
• 1:07:24
  Speaker 2: of this? And here it spit out some examples of this.
• 1:07:33
  Speaker 2: Number one is Microsoft's, that is Copilot, but not Copilot,
• 1:07:39
  Speaker 2: but Microsoft has apparently a project called the JFK Files
• 1:07:44
  Speaker 2: AI Project, codenamed Azure Cognitive Search. What it did, Microsoft
• 1:07:52
  Speaker 2: built a prototype using Azure Cognitive Search to process over 34,000
• 1:07:57
  Speaker 2: pages of newly unredacted NARA records. How it worked. The
• 1:08:02
  Speaker 2: AI used OCR, that's optical character recognition, to transcribe low-quality
• 1:08:08
  Speaker 2: handwritten field notes, indexed unstructured cables, and extract name, entities, people, places, cryptonyms.
• 1:08:16
  Speaker 2: The result, it built a searchable semantic network map allowing
• 1:08:21
  Speaker 2: users to query relationships such as connecting CI crypts to
• 1:08:25
  Speaker 2: specific field agents across thousands of unrelated documents in seconds. Crazy.
• 1:08:34
  Speaker 2: Dartmouth 3D forensics on the Oswald backyard photos. It tells
• 1:08:38
  Speaker 2: about the computer scientist Dr. Henry Farid at Dartmouth College
• 1:08:43
  Speaker 2: used 3D human modeling and computer vision to evaluate the
• 1:08:46
  Speaker 2: infamous photo of Lee Harvey Oswald holding a Carcano rifle.
• 1:08:52
  Speaker 2: But again, this is older AI technology and it's more
• 1:08:58
  Speaker 2: of a more of recreating a physiological accurate 3D model
• 1:09:04
  Speaker 2: of Oswald inside of a computer and to determine things
• 1:09:07
  Speaker 2: where shadows should go given the time of day the
• 1:09:11
  Speaker 2: time of the photograph and so on and so forth
• 1:09:14
  Speaker 2: placing virtual light sources matching Dealey Plaza and Dallas Sun
• 1:09:19
  Speaker 2: positions Again, a lot of problems there. And not the
• 1:09:26
  Speaker 2: best use of AI. Again, this was probably 10 years ago.
• 1:09:33
  Speaker 2: There's a Knott Laboratories that has a 3D laser scan
• 1:09:37
  Speaker 2: and photogrammetry. What it did, forensic engineering firm Knott Labs,
• 1:09:43
  Speaker 2: that's K-N-O-T-T, Knott Labs, used high-definition LIDAR laser scanning of
• 1:09:49
  Speaker 2: Dealey Plaza to test the Warren Commission's single bullet theory.
• 1:09:53
  Speaker 2: How it worked? They captured a point cloud map of
• 1:09:56
  Speaker 2: Dealey Plaza, up to 2 million spatial points per second,
• 1:10:00
  Speaker 2: and used photogrammetry to align frames of the 8mm of
• 1:10:04
  Speaker 2: Pruder film to the exact 3D space. The result? By
• 1:10:09
  Speaker 2: aligning the known bullet entrance-exit wounds of JFK and Connelly,
• 1:10:15
  Speaker 2: With their exact positions in the 3D space within the vehicle,
• 1:10:20
  Speaker 2: their trajectory model showed a spatial trajectory misalignment from the
• 1:10:25
  Speaker 2: sixth floor window. They concluded that modern science refutes the
• 1:10:29
  Speaker 2: single bullet narrative pointing to multiple shooters. Which is interesting, actually. Very,
• 1:10:41
  Speaker 2: very interesting. There is open source AI architecture called the
• 1:10:48
  Speaker 2: JFK Documents MCP server. What it did, independent AI developers
• 1:10:53
  Speaker 2: created specialized model context protocol tools specifically for the NARA
• 1:10:59
  Speaker 2: JFK collection. How it works, developers connect modern large language
• 1:11:04
  Speaker 2: models directly to the National Archives API via custom servers.
• 1:11:08
  Speaker 2: The result, using vector search instead of rigid keyword matching,
• 1:11:12
  Speaker 2: researchers can prompt the AI to, quote, Find all internal
• 1:11:16
  Speaker 2: CIA cables discussing unauthorized surveillance of anti-Castro groups in late 63,
• 1:11:22
  Speaker 2: and receive and return cross-reference primary source citations across millions
• 1:11:27
  Speaker 2: of pages. So, there are, there is, a use, folks.
• 1:11:44
  Speaker 2: For AI. In the Kennedy assassination. You just have to be.
• 1:11:54
  Speaker 2: Very careful with it. Again. It's just like anything else.
• 1:12:01
  Speaker 2: Garbage in. If you got garbage going in. You're going
• 1:12:04
  Speaker 2: to get garbage coming out. But if you can train.
• 1:12:10
  Speaker 2: To the best of your ability. A large language model.
• 1:12:17
  Speaker 2: it can definitely help you. So, in the comments section,
• 1:12:24
  Speaker 2: what do you think, folks? Should Big Bad Bob start
• 1:12:27
  Speaker 2: using AI to better structure my shows, to better do
• 1:12:31
  Speaker 2: research and all of that stuff? I mean, it's a
• 1:12:36
  Speaker 2: possibility of things to come. I mean, I kind of
• 1:12:40
  Speaker 2: like the way I do things now, but do you
• 1:12:43
  Speaker 2: think AI would help, could help me? I don't know.
• 1:12:48
  Speaker 2: You know, it would certainly eliminate a lot of skafooza
• 1:12:52
  Speaker 2: questions in the comment sections that people ask me if
• 1:12:56
  Speaker 2: they just ask your co-pilot or ask your jimmy. That's
• 1:13:04
  Speaker 2: what I call Gemini. Either Gemma or Jimmy. Depending on
• 1:13:08
  Speaker 2: if I'm feeling a female or male vibe coming out
• 1:13:12
  Speaker 2: of Gemini. Either Jimmy or Jimmy. I don't know. What
• 1:13:23
  Speaker 2: I can tell you is don't be afraid of it.
• 1:13:27
  Speaker 2: And learn all you can about it. It's not something
• 1:13:30
  Speaker 2: you want to be blindsided by. For sure. The more
• 1:13:35
  Speaker 2: that you can understand artificial intelligence and the more that
• 1:13:38
  Speaker 2: you can make it work for you and help you
• 1:13:42
  Speaker 2: in your life You know, you can, you can use
• 1:13:45
  Speaker 2: it to create a nutrition plan. You can use it
• 1:13:48
  Speaker 2: to create a workout plan. You know, you input your information,
• 1:13:54
  Speaker 2: you know, you say, Hey, I'd like to stick to
• 1:13:56
  Speaker 2: be eating about 1200 calories a day. I don't have
• 1:14:00
  Speaker 2: time to eat until two in the afternoon. Uh, please
• 1:14:03
  Speaker 2: construct for me the best, uh, menu, you know, along
• 1:14:10
  Speaker 2: with an ingredients and a shopping list, uh, Oh, and
• 1:14:14
  Speaker 2: by the way, I don't eat seafood, and some of
• 1:14:19
  Speaker 2: my favorite foods are Chinese, Mexican, blah, blah, blah. Make
• 1:14:23
  Speaker 2: sure to incorporate some of that, a lot of that
• 1:14:25
  Speaker 2: in there, actually. And it will come up and help
• 1:14:30
  Speaker 2: you and determine a nutrition plan, a meal planner, whatever
• 1:14:36
  Speaker 2: you want to do with it. It's crazy how much
• 1:14:41
  Speaker 2: AI can help you in your life. If you allow
• 1:14:46
  Speaker 2: it to. It's not going anywhere. Short of John Connor
• 1:14:52
  Speaker 2: coming back. From the future. To save us all. It's
• 1:14:59
  Speaker 2: not going anywhere. So. I don't know. I don't know
• 1:15:03
  Speaker 2: what to think of it. Other than. Don't be afraid.
• 1:15:09
  Speaker 2: It's here. To stay. Use it. To the best of
• 1:15:15
  Speaker 2: your ability and advantage. And Godspeed. Be on the lookout
• 1:15:23
  Speaker 2: for Skynet. Alright folks, that's it for me. Until next time.
• 1:15:30
  Speaker 2: That's some random. That's some research. And those are some
• 1:15:33
  Speaker 2: ramblings from your boy Rob. I'm out. Until next time.
• 1:15:48
  Speaker 2: Am I part robot? Maybe I am. Maybe I'm not.
• 1:15:52
  Speaker 1: I don't know.
• 1:15:53
  Speaker 2: Alright, I'll shut the fuck up.
• 1:15:57
  Speaker 1: Peace.
• 1:15:58
  Speaker 2: Random Research Ramblings, we're out!
• 1:16:02
  Speaker 1: Random Research Ramblings, we're out! JFK questions one by one,
• 1:16:07
  Speaker 1: talk it out, the day ain't done! Random Research Ramblings,
• 1:16:15
  Speaker 1: Ramblings we're out!