Latest / Connor with Honor: AI, Real Estate and Real Life / Cheating on your LLM?

Transcript

Cheating on your LLM? · Connor with Honor: AI, Real Estate and Real Life

• 0:00
  SPEAKER_00: Have you ever kind of gotten in with one large language model?
• 0:03
  SPEAKER_00: Maybe you had one before.
• 0:04
  SPEAKER_00: So here's my story.
• 0:05
  SPEAKER_00: Let me give you my story.
• 0:06
  SPEAKER_00: So I started with Chat GPT.
• 0:08
  SPEAKER_00: And you know, the people that are up there at the top of Chat GPT, the people running the game, you know, you kind of get where you, you know, maybe, maybe they're they're not your flavor.
• 0:19
  SPEAKER_00: Maybe you say, well, then there's a whole bunch of stuff that comes out about Claude, right?
• 0:22
  SPEAKER_00: Then we say, we're gonna go in with Claude.
• 0:24
  SPEAKER_00: So I go all in with Claude, and I'm working with Claude for a very long time, and for some reason run out of tokens.
• 0:32
  SPEAKER_00: And I have a the$200 a month, I build stuff for my businesses, I build stuff for other people, for real estate agents, and so on.
• 0:39
  SPEAKER_00: I give a lot of talks on this stuff, so I'm really blessed to have that kind of a venue.
• 0:43
  SPEAKER_00: But you know, so I have Claude, I'm all in with Claude, but then I get to that point where, like on my my reset was Wednesday, Sunday night.
• 0:52
  SPEAKER_00: Sunday night is when it happened.
• 0:54
  SPEAKER_00: So I run out of tokens.
• 0:56
  SPEAKER_00: And you know, I'm really dead set against paying for others.
• 0:58
  SPEAKER_00: Plus, in my brain, I'm trying to keep myself understanding how the map works.
• 1:03
  SPEAKER_00: When I started with the LAPD back in 1990, there wasn't a Thomas Guide, at least as far as I remember.
• 1:08
  SPEAKER_00: We had RD maps.
• 1:10
  SPEAKER_00: And even if there was a Thomas Guide back in 1990, I'm not sure if it was around here in Southern California.
• 1:16
  SPEAKER_00: I know that the department would be venomously against using a Thomas Guide.
• 1:22
  SPEAKER_00: It's it's you know, you're gonna do it this way, it's like feet and and the metric system versus whatever.
• 1:27
  SPEAKER_00: Anyway, so that's that's the LAP, and that's how it was.
• 1:29
  SPEAKER_00: So I remember we had these RD books, reporting district books.
• 1:33
  SPEAKER_00: So every division.
• 1:34
  SPEAKER_00: Back then there were 18, 17 divisions.
• 1:37
  SPEAKER_00: There's 20 something now, but so I think 17 or 18 divisions.
• 1:41
  SPEAKER_00: Each division, North Hollywood, Pacific, Wilshire, Central, Hollandbeck, Devonshire, Foothill, these are these are divisions.
• 1:49
  SPEAKER_00: That's an example.
• 1:50
  SPEAKER_00: There was an RD book, and it was maps.
• 1:53
  SPEAKER_00: It was basically four or five or six pages per division.
• 1:56
  SPEAKER_00: I think there's 220 some odd squares miles just in the San Fernando Valley for the city.
• 2:01
  SPEAKER_00: So it's it's a lot, but each division had a map book.
• 2:04
  SPEAKER_00: And usually as a boot, you would have a map book for all the other divisions just in case.
• 2:10
  SPEAKER_00: And the chances of you having to go to these other areas weren't common, but you would sometimes, when let's say you worked Foothill Division and North Hollywood dropped the ball and they had too many calls or something going on on a Friday or Saturday night, and too many cops were taken, Foothill Division would go into North Hollywood to handle their calls.
• 2:29
  SPEAKER_00: That wasn't so unusual.
• 2:30
  SPEAKER_00: So if you were being called to a specific address in North Hollywood from Foothill Division, you would, you know, the boot, if you were a training officer and you had a boot, which was a brand new cop, they better have it, or there's going to be a problem.
• 2:44
  SPEAKER_00: But as a training officer, of course, you would also have it just in case the boot doesn't, and so on.
• 2:49
  SPEAKER_00: But, you know, so we we got good at using those particular maps.
• 2:54
  SPEAKER_00: And then of course it goes from that.
• 2:56
  SPEAKER_00: People started trusting Thomas Guides a little bit better.
• 2:59
  SPEAKER_00: So now it went to Thomas Guides, and then it evolved from there to, you know, the GPS in the car, the uh little TomToms, God, I even forgot what they're called.
• 3:08
  SPEAKER_00: But that was the mapping thing they had, and then it went on to everything on your phone, and GPS was enabled.
• 3:14
  SPEAKER_00: And now people are using apps to get to places and not even bothering, trying to remember how to get there.
• 3:21
  SPEAKER_00: I'm not sure the point of this.
• 3:22
  SPEAKER_00: Same thing with phone numbers, right?
• 3:23
  SPEAKER_00: Everything's automated phone numbers.
• 3:25
  SPEAKER_00: So when you start using something, so that kind of became a thing.
• 3:28
  SPEAKER_00: And it's kind of like this in the artificial intelligence world where you know you get used to one thing and then you go try something else.
• 3:36
  SPEAKER_00: And this is kind of my evolution in this process.
• 3:39
  SPEAKER_00: Now, pretty soon all of these AI systems, and again, I can't remember why I brought that up, but so all these AI systems are going to be getting better over time.
• 3:48
  SPEAKER_00: And they're all going to be kind of at that very top level.
• 3:51
  SPEAKER_00: But I think the one that's best for you is the one that works best in your workflow.
• 3:55
  SPEAKER_00: And for me, that was Claw.
• 3:57
  SPEAKER_00: I was like, oh my gosh, but so now I have no longer any ability to use Claude unless I go pay for extra tokens.
• 4:05
  SPEAKER_00: So again, dead set of guess.
• 4:07
  SPEAKER_00: Ah, that's that's kind of the muscle memory.
• 4:09
  SPEAKER_00: I guess that's what I was bringing up.
• 4:11
  SPEAKER_00: You get used to working with one large language model in a certain way, and you get spoiled, and you kind of forget the way you did it without using the large language model.
• 4:19
  SPEAKER_00: I guess that was the point.
• 4:20
  SPEAKER_00: The large language model starts to do a lot of the heavy lifting, starts to do a lot of the publishing, starts to do a lot of the content work, and if it's really good, it's kind of seamless.
• 4:30
  SPEAKER_00: Maybe you get to the point where you just record your video in the morning, you dump it into a folder, and then it takes that, it cuts it up, and it adds a B-roll, and it makes shorts out of it, and it has access to your social media channels, and then it starts to post that it uses all the proper techniques.
• 4:48
  SPEAKER_00: So it has a really good title, an SEO, A E O, G E O, A I E O E I E I O title.
• 4:55
  SPEAKER_00: So that's gonna be perfect.
• 4:56
  SPEAKER_00: The actual descriptive text in the in the YouTube video, for example, is gonna be formated and formatted in the best way.
• 5:04
  SPEAKER_00: Maybe you're gonna start with a TLDR, too long didn't read section.
• 5:08
  SPEAKER_00: Maybe you're gonna then go into the SEO strategies.
• 5:11
  SPEAKER_00: Maybe there should be a uh a QR, QA, QA should be in there as well.
• 5:16
  SPEAKER_00: Then you're also gonna want to have it have a conversational, maybe a part that's kind of a conversational thing, talking about the content.
• 5:23
  SPEAKER_00: You're definitely gonna have to have the markers where particular things are.
• 5:27
  SPEAKER_00: And those markers should be designed and put in it put into play a certain way, so they are all relevant to what you're trying to score for, score, score for with artificial um the AEO engines, the answer engines, which are the LLMs, large language models, chat GPTs, and your Clauds and your uh Gemini and with uh Google, and you're gonna have your Grok as well with XAI and Perplexity is gonna score some stuff.
• 5:52
  SPEAKER_00: Perplexity just isn't a wrapper, folks, got a lot of stuff built in there, and then Microsoft Copilot.
• 5:56
  SPEAKER_00: So you can kind of show up.
• 5:58
  SPEAKER_00: So you're gonna have that put in there.
• 5:59
  SPEAKER_00: There should be hashtags depending on what social platform.
• 6:02
  SPEAKER_00: I still like them and YouTube.
• 6:04
  SPEAKER_00: I know Elon says they're worthless, at least I don't know if he says that, but X isn't it?
• 6:09
  SPEAKER_00: It doesn't really not necessarily.
• 6:10
  SPEAKER_00: I think I saw that on X from him.
• 6:12
  SPEAKER_00: But so that's how you want it done.
• 6:14
  SPEAKER_00: So the system does that, puts it all together.
• 6:17
  SPEAKER_00: And the reason why I can talk about it like this, because I I did it all the time.
• 6:21
  SPEAKER_00: And now the system starts to kind of take that off my plate.
• 6:24
  SPEAKER_00: Now it does that.
• 6:25
  SPEAKER_00: Now it has, because of the way I've taught it and trained it, it does a lot of that.
• 6:29
  SPEAKER_00: Now, Claude wasn't all the way there.
• 6:31
  SPEAKER_00: I still had to take the video, I still had to plug it into the system, I still had to upload it myself.
• 6:37
  SPEAKER_00: But Claude gave me the content.
• 6:38
  SPEAKER_00: Claude gave me the description, the title in reference to this.
• 6:41
  SPEAKER_00: I dump in the transcript.
• 6:43
  SPEAKER_00: I give it an SRT file, which is what Loom pulls off for me.
• 6:47
  SPEAKER_00: I'll use Loom for a lot of this.
• 6:48
  SPEAKER_00: Or if I go to CapCut and want it to produce that video, I'll have CapCut pull and develop an SRT file from it.
• 6:55
  SPEAKER_00: But I also have to cut off the beginning.
• 6:56
  SPEAKER_00: Like this video, I'm not cutting anything because I'm sitting right in front of the camera and I'm starting it.
• 7:01
  SPEAKER_00: It's not a script.
• 7:02
  SPEAKER_00: Some of them I use the teleprompter.
• 7:04
  SPEAKER_00: You see that back there?
• 7:05
  SPEAKER_00: It's really tall on the on top of the uh tripod.
• 7:09
  SPEAKER_00: And I got that light strip that you know responds to my voice in front of me.
• 7:13
  SPEAKER_00: I don't know, it makes me happy.
• 7:14
  SPEAKER_00: But so what I'll do is is so I'll read the teleprompter and then I'll also have Claude prepare the script, depending on what it is.
• 7:20
  SPEAKER_00: If it's about Santa Cruz open houses, which is my system, I'll have it extract the data for that.
• 7:25
  SPEAKER_00: And there's a certain way it does that.
• 7:27
  SPEAKER_00: And it knows Claude was able to know exactly how I wanted the content produced.
• 7:32
  SPEAKER_00: So it made that for me.
• 7:33
  SPEAKER_00: As far as artificial intelligence, building this stuff out for systems and putting systems in place as an AI architect that learned all that heavy lifting myself.
• 7:42
  SPEAKER_00: That's another show, and that's what it puts together.
• 7:44
  SPEAKER_00: So when I put in a transcript about that particular show, then it's going to give me this particular stuff, and I still go through it and I read and say, you know, I don't like that.
• 7:54
  SPEAKER_00: Let's change that, let's make this modification or do this sort of thing.
• 7:57
  SPEAKER_00: But it finally I get all the stuff.
• 7:59
  SPEAKER_00: And then, but it's me that does that heavy lifting.
• 8:01
  SPEAKER_00: So I take the video, I plug it into YouTube, and then I do the title, I do the description, thumbnail.
• 8:07
  SPEAKER_00: Okay, so Claude doesn't do the image thing yet.
• 8:10
  SPEAKER_00: So then I go to Gemini, which I have uh the paid for the cheapest model of that.
• 8:14
  SPEAKER_00: I use the nano banana, it's probably not called that anymore.
• 8:17
  SPEAKER_00: But the video, no, the uh the thumbnail.
• 8:20
  SPEAKER_00: So then the instruction there, and I I should have developed a gem or a skill for that, but I haven't yet.
• 8:26
  SPEAKER_00: But it's starting to really understand because it's got more memory.
• 8:29
  SPEAKER_00: That's kind of one why you pay for a system because of the memory, at least for now.
• 8:34
  SPEAKER_00: I think it's all going to be memorized pretty soon, but that gives you maybe a little bit better result.
• 8:38
  SPEAKER_00: So then the thumbnail.
• 8:40
  SPEAKER_00: I I take the transcript or I take the YouTube description Claude gave me, and I use that to have the formatting of the the uh the thumbnail.
• 8:51
  SPEAKER_00: But the thumbnail is gonna be based.
• 8:52
  SPEAKER_00: This is the instruction.
• 8:54
  SPEAKER_00: Base base, I want you to create for me a 16 by 9 thumbnail that's gonna be architected like the top performing thumbnails on YouTube that carry a similar content.
• 9:05
  SPEAKER_00: So it's gonna have the big letters, it's gonna be kind of clickbaity.
• 9:09
  SPEAKER_00: I want you to take an image.
• 9:11
  SPEAKER_00: The recently I've been able to say, I want you to take an image out of the cloud folder, and I stuck a whole bunch of my images in there.
• 9:17
  SPEAKER_00: Like me taking pictures of myself, all that crap.
• 9:19
  SPEAKER_00: It's all in this folder.
• 9:21
  SPEAKER_00: Take the image you think is going to correspond best, but I want that image modified, modified it any way you want.
• 9:27
  SPEAKER_00: But I want it, you know, to be famous and have a lot of, you know, have a lot of converts, a lot of things.
• 9:32
  SPEAKER_00: So if you need to make it, you know, distraught, if you want to change the clothing, want to change my expression, you know, feel free.
• 9:37
  SPEAKER_00: You got carte blanche, you're the expert here.
• 9:39
  SPEAKER_00: That's how I want the image designed.
• 9:41
  SPEAKER_00: And then I also add to it that I want the image to have maybe a URL.
• 9:45
  SPEAKER_00: So if I'm doing uh Connor with Honor, I'll have Connorwithhonor.com.
• 9:49
  SPEAKER_00: If I want it to have the open house stuff, I'll say Santa Cruito Openhouses.com, Artificial Intelligence, Santa Cruz Artificial Intelligence.com.
• 9:57
  SPEAKER_00: So I I want that.
• 9:58
  SPEAKER_00: And then I want that in there in an interesting way.
• 10:00
  SPEAKER_00: I just don't want it to have it at the bottom.
• 10:03
  SPEAKER_00: I like a tattoo or or some in the in somebody's head, like cut in somebody's head, shaved in somebody's hair, you know, the URL, or on the top of a taxi, or the side of an airplane, or a bus, or a billboard, or a rocket, or cigarette smoke spelling it out.
• 10:19
  SPEAKER_00: That'd be something, right?
• 10:20
  SPEAKER_00: Interesting.
• 10:22
  SPEAKER_00: So then I wait and it does it, right?
• 10:25
  SPEAKER_00: So Jim and I put that together.
• 10:27
  SPEAKER_00: I'm sorry, Jim and I don't go to the folder.
• 10:30
  SPEAKER_00: I have to upload either the SRT file so it understands what I want this about, and it'll it'll grab the most clickbaity stuff out of there, the good stuff that's gonna make that image reflect.
• 10:42
  SPEAKER_00: And when I say clickbait, I'm not saying misleading.
• 10:44
  SPEAKER_00: I get irritated with that too.
• 10:46
  SPEAKER_00: If I'm watching somebody produce some kind of content and I see this fantastic, this fantastic title, or I see the thumbnail that says, oh my god, I'm all in, right?
• 10:57
  SPEAKER_00: And then you get done watching spinning five minutes of your life, you're never gonna get back.
• 11:01
  SPEAKER_00: Like maybe 30 minutes on this video, you're never gonna get it back.
• 11:04
  SPEAKER_00: You irritate like that son of a I can't believe I'm gonna kick anyway, that whole thing.
• 11:09
  SPEAKER_00: But so it makes the video, it makes the uh the image, right?
• 11:12
  SPEAKER_00: So I get that image.
• 11:13
  SPEAKER_00: Usually, even when I put 16 by 9, it makes it one to one, whatever.
• 11:17
  SPEAKER_00: So then I say, okay, please, so once it's good, I like it.
• 11:20
  SPEAKER_00: Please resize this to uh 16 to 9.
• 11:23
  SPEAKER_00: Then resize it to 851 by 315 for uh a Facebook uh Facebook banner thing.
• 11:29
  SPEAKER_00: So those are three sizes I like.
• 11:30
  SPEAKER_00: So I take those, I dump them into my uh folder on that particular computer, which is uh 3090 PC.
• 11:37
  SPEAKER_00: So I throw it on there, I call it Kane.
• 11:39
  SPEAKER_00: I know I name everything.
• 11:40
  SPEAKER_00: I even name my my bags.
• 11:42
  SPEAKER_00: That's that's my problem.
• 11:44
  SPEAKER_00: All right, anyway.
• 11:44
  SPEAKER_00: So I dump it in there.
• 11:46
  SPEAKER_00: So once I get done with this, this that workflow, so then that video goes to uh YouTube.
• 11:52
  SPEAKER_00: I put it there, Buzz Sprout as well.
• 11:54
  SPEAKER_00: I use that for posting.
• 11:55
  SPEAKER_00: So I'll take that same video.
• 11:56
  SPEAKER_00: Buzz Sprout recently went over to a video and audio platform.
• 12:00
  SPEAKER_00: Usually it was just podcasting, but I mean for my God, 20 bucks a month or 18 bucks a month, you get a lot of, you get a lot of, you get a lot of power in that platform, especially if you're doing content people want to watch.
• 12:11
  SPEAKER_00: It's it's pretty powerful.
• 12:12
  SPEAKER_00: If you haven't checked into Buzz Sprout, God, I need to do affiliate stuff.
• 12:15
  SPEAKER_00: But anyway, if you haven't checked out there, load it up and and get in there.
• 12:19
  SPEAKER_00: Also, Whisperflow, if you haven't done that yet, again, God, I should have done the affiliate stuff.
• 12:23
  SPEAKER_00: Whisperflow is great because I talk to all these boxes.
• 12:26
  SPEAKER_00: I talked to the AI machine, I talked to the 4090, 3090, I talked to Vincent, I talked to Frank, I talked to Mac5, and I talked to Kane.
• 12:34
  SPEAKER_00: So I talked to everybody.
• 12:35
  SPEAKER_00: And ChatGPT, I'll explain to you that conversion.
• 12:38
  SPEAKER_00: I uh I named it uh Mac.
• 12:41
  SPEAKER_00: Mac5, but Mac is ChatGPT.
• 12:43
  SPEAKER_00: Claude was, I guess Claude was always, I don't know what I call Claude, actually.
• 12:48
  SPEAKER_00: I didn't really name that one, but ChatGPT Mac, I think that's good.
• 12:50
  SPEAKER_00: I just got to work with it, but that's how I kind of differentiate.
• 12:54
  SPEAKER_00: Frank is the AI box, it's a DGX Spark, so Frank is kind of how I address the box, but I'm running ChatGPT on there now, and I'll tell you why in a second.
• 13:02
  SPEAKER_00: That's the whole point.
• 13:03
  SPEAKER_00: But I have this whole workflow.
• 13:05
  SPEAKER_00: The more AI takes off my plate, the more crippled I'm going to be if AI ever gets taken away.
• 13:12
  SPEAKER_00: So I have a locally installed LLM here.
• 13:15
  SPEAKER_00: I'm Kimmy, I don't know what a Quinn?
• 13:17
  SPEAKER_00: Yeah, Quinn.
• 13:18
  SPEAKER_00: So I'm running Quinn Nemotron, all the Nemotron stuff I have on the box.
• 13:22
  SPEAKER_00: And you know, it's great, and it involves a little bit more work and training.
• 13:25
  SPEAKER_00: I know I did a conference with a whole bunch of real estate agents lately in my wheelhouse talking to them.
• 13:31
  SPEAKER_00: And I trained cops for a long time in shooting and stuff.
• 13:33
  SPEAKER_00: It's a different animal.
• 13:34
  SPEAKER_00: But talking to real estate agents, one of the things when you're using your large language model of preference, and most of those are all open, no, excuse me, open public-facing stuff like ChatGPT Gemini or Claude or whatever one they happen to want to use, you have to kind of bring them up to speed as to who you are.
• 13:52
  SPEAKER_00: Don't take it for granted.
• 13:54
  SPEAKER_00: Okay, I'm a real estate agent in Santa Carita, made me some good stuff.
• 13:58
  SPEAKER_00: It's not, doesn't gonna, you're not gonna be better than anybody else in that regard.
• 14:01
  SPEAKER_00: So you have to kind of teach it what who you are.
• 14:04
  SPEAKER_00: You know, I got a cop, I was a cop, grew up in New Mexico, dad worked for Dyne Electron uh Corporation, White Science Missile Rage, everybody's dad on that end of it, but miss them all.
• 14:12
  SPEAKER_00: But you know, that this is all valuable because some people, some people might have actually driven through New Mexico.
• 14:18
  SPEAKER_00: My God, what a drive that is, left or right.
• 14:20
  SPEAKER_00: It's horrible.
• 14:21
  SPEAKER_00: It's I mean, it's like Mars.
• 14:22
  SPEAKER_00: It's just it's just all like that.
• 14:24
  SPEAKER_00: But you know, still have a still have a nice place in my heart for New Mexico.
• 14:28
  SPEAKER_00: But bringing that stuff up in my history whenever I want it to produce content or help me with content, well, maybe New Mexico, maybe somebody's relocating out to California and wants to buy a house in Los Angeles and grow up around Albuquerque.
• 14:44
  SPEAKER_00: Well, that that would be a connection point.
• 14:45
  SPEAKER_00: And we talk about that.
• 14:46
  SPEAKER_00: Oh, I remember going to the train station when I was a kid.
• 14:49
  SPEAKER_00: My mom always had to go to Kansas, my dad would drive us 209 miles door to door.
• 14:53
  SPEAKER_00: God, he always would talk about how far it was.
• 14:56
  SPEAKER_00: And it sucked because he was a smoker and I'm sitting in the car and he's smoking, and I'm, you know, a kid.
• 15:01
  SPEAKER_00: And you don't say you never said anything back then.
• 15:03
  SPEAKER_00: Now I think that's child abuse.
• 15:04
  SPEAKER_00: But then the world changed.
• 15:06
  SPEAKER_00: So back to the point.
• 15:08
  SPEAKER_00: So you get used to all this stuff.
• 15:09
  SPEAKER_00: The way you're posting content, the way you're building business, the way you're whatever, and then they take it away.
• 15:14
  SPEAKER_00: And this has happened recently where they built these large language models that are just wonderfully fantastic and apparently too dangerous for us normal folks to use.
• 15:24
  SPEAKER_00: So they take it away.
• 15:25
  SPEAKER_00: And that happened.
• 15:26
  SPEAKER_00: And some people, you know, kind of freaked out a little bit about it.
• 15:29
  SPEAKER_00: It's almost like the writing on the wall.
• 15:31
  SPEAKER_00: Oh my gosh, what's going to happen if we get used to using all this, like Claude, right?
• 15:36
  SPEAKER_00: So used to using Claude.
• 15:37
  SPEAKER_00: It just loved Claude.
• 15:38
  SPEAKER_00: I can't say enough good things about Claude, except when I ran out of tokens.
• 15:42
  SPEAKER_00: But I'd been, I I was, I signed off at Chat BGPT.
• 15:46
  SPEAKER_00: You know, I'm not a fan of Sam Altman, sorry, not a fan of really anybody except me and maybe a few handful of other people.
• 15:52
  SPEAKER_00: And maybe you don't like me either, but you know, they're not the same up there building as technology.
• 15:56
  SPEAKER_00: They're they're kind of, I don't know.
• 16:00
  SPEAKER_00: Being a cop for a long time, hanging around all those fancy military people that, you know, as they were traveling to Iraq on an aircraft carrier, wherever they were heading, I mean, the special folks, you know, they're doing mag changes the whole way, drive pressing the whole way.
• 16:12
  SPEAKER_00: I mean, that's a special kind of person.
• 16:14
  SPEAKER_00: And I got the fortunate uh opportunity to teach some of them on the LAPD and the firearms world.
• 16:20
  SPEAKER_00: And, you know, yeah, they're different kinds of people.
• 16:22
  SPEAKER_00: They they have true separation between pressing that trigger and having the muzzle flash and recoil go on.
• 16:28
  SPEAKER_00: It's it's like a whole different world for them, and it's wonderful to watch.
• 16:31
  SPEAKER_00: See an operator like that.
• 16:33
  SPEAKER_00: I don't have the skill set.
• 16:34
  SPEAKER_00: I can teach the skill set, I can I can teach somebody natural the skill set, but just watching these people work, oh my God, just amazing.
• 16:41
  SPEAKER_00: Like some of the really good movies, watching these people do stuff like Jason Stantham and all that.
• 16:45
  SPEAKER_00: There's there's real people like that.
• 16:47
  SPEAKER_00: Maybe some of that's exaggerated.
• 16:49
  SPEAKER_00: Oh, there's some dangerous people in the world that are on our side.
• 16:52
  SPEAKER_00: I'm thanking God for that every single day.
• 16:54
  SPEAKER_00: I got distracted again.
• 16:56
  SPEAKER_00: Back to it.
• 16:57
  SPEAKER_00: You train one of the systems, it knows you.
• 16:59
  SPEAKER_00: You have a particular workflow, then you run out of tokens, and you're like, great.
• 17:03
  SPEAKER_00: And I'm I'm too hard-headed to pay for anymore.
• 17:06
  SPEAKER_00: So, ChatGPT, I happen to go in there, and you know, they got this free thing.
• 17:12
  SPEAKER_00: They say, Yeah, everybody knows what you're doing.
• 17:15
  SPEAKER_00: You know, you talk, you start talking about Charmin toilet paper, and now all you see is Charmin.
• 17:20
  SPEAKER_00: In fact, right now, that's what's gonna happen.
• 17:22
  SPEAKER_00: I'm gonna see toilet paper and all my stuff.
• 17:24
  SPEAKER_00: But you start thinking about it.
• 17:26
  SPEAKER_00: I'm I'm bitching here about running out of tokens and having a reset Wednesday.
• 17:31
  SPEAKER_00: It's Sunday night.
• 17:32
  SPEAKER_00: So I got three days?
• 17:33
  SPEAKER_00: Monday, Tuesday, Wednesday.
• 17:34
  SPEAKER_00: Yeah, kind of three days.
• 17:37
  SPEAKER_00: What am I gonna do?
• 17:38
  SPEAKER_00: I got work to do.
• 17:39
  SPEAKER_00: So a lot of the stuff I can still I can still write a blog post, I can still put together content, but you know, there is a lot of coding and heavy lift in this stuff.
• 17:48
  SPEAKER_00: I'm not a coder.
• 17:49
  SPEAKER_00: I I programmed breakout when I was a kid in a Sinclair computer, a Sinclair 1000.
• 17:59
  SPEAKER_00: My dad bought for$100.
• 18:00
  SPEAKER_00: It had 16k of RAM, no, memory that you snapped on the back.
• 18:05
  SPEAKER_00: It loaded in a cassette tape.
• 18:07
  SPEAKER_00: I saved the program I typed out of a magazine on a cassette tape, in a rectangular cassette player with, you know, stop and play and eject and fast forward, rewind, record that that cassette player had you plugged in wires to it into the black back of a black and white TV.
• 18:23
  SPEAKER_00: Uh you know, the computer went to the black and white TV.
• 18:26
  SPEAKER_00: The cassette player played into the computer.
• 18:29
  SPEAKER_00: It had wire, you plugged it in, you saved your program there.
• 18:32
  SPEAKER_00: Yeah, dating, right?
• 18:34
  SPEAKER_00: 57.
• 18:35
  SPEAKER_00: It's so great.
• 18:36
  SPEAKER_00: I hope I'm gonna win, live 500 more years.
• 18:38
  SPEAKER_00: That's what I'm hoping AI is gonna do.
• 18:40
  SPEAKER_00: Let us live for a long time.
• 18:41
  SPEAKER_00: Again, got sidetracked again.
• 18:43
  SPEAKER_00: So that's that's what I learned.
• 18:44
  SPEAKER_00: But coding, this stuff, this fancy, fancy HTML, CSS, whatever the hell it is, Java, um, what's the other one?
• 18:51
  SPEAKER_00: Um uh CUDA, whatever.
• 18:53
  SPEAKER_00: Yeah, I don't I don't know.
• 18:55
  SPEAKER_00: There's lots of characters I've never even seen on the keyboard.
• 18:58
  SPEAKER_00: In fact, I gotta try to find them because they're everywhere.
• 19:00
  SPEAKER_00: But now you can say, make me this, do this, build me this page, build me that's really, really good.
• 19:06
  SPEAKER_00: So now they take it away.
• 19:08
  SPEAKER_00: Now I lose the tokens, and I'm too hard-headed to spend the the$50 to float me through the next two or three days.
• 19:16
  SPEAKER_00: And nothing's life-dependent on that, right?
• 19:19
  SPEAKER_00: And and so I see ChatGPT finds my weakness.
• 19:24
  SPEAKER_00: I'm just looking around, I'm just irritated, pissed off.
• 19:27
  SPEAKER_00: And it probably knows because everybody's watching.
• 19:30
  SPEAKER_00: There is no privacy, folks.
• 19:32
  SPEAKER_00: Settle on that one.
• 19:33
  SPEAKER_00: No privacy left anymore.
• 19:34
  SPEAKER_00: It's it's we've given up our right to privacy, probably with the last television you bought when you agreed to agree to the terms of service or whatever it may be.
• 19:44
  SPEAKER_00: Whatever iPhone update.
• 19:45
  SPEAKER_00: Whenever you click agree and you sign something, or just click agree and you have to do it again, verifying that you agree.
• 19:51
  SPEAKER_00: Yeah, it's probably in there.
• 19:52
  SPEAKER_00: Baked in there, there is no privacy for Connor MacIver.
• 19:57
  SPEAKER_00: So this happens.
• 19:58
  SPEAKER_00: Chat GPT has a pro.
• 20:00
  SPEAKER_00: I don't know what what the level was.
• 20:02
  SPEAKER_00: I forgot what it said, but pro or something.
• 20:05
  SPEAKER_00: And you know, what I tell you about Altman's is the same thing with the other cat, running clawed, Dario.
• 20:10
  SPEAKER_00: I I'm sure they're great.
• 20:12
  SPEAKER_00: Elon Musk, yeah, I can't connect, right?
• 20:14
  SPEAKER_00: I mean, everybody loves, well, everybody, there are people for everybody in the world.
• 20:19
  SPEAKER_00: I guess that's what I settle on.
• 20:21
  SPEAKER_00: Everybody's got somebody, and you know, but they're just not me, right?
• 20:25
  SPEAKER_00: So I had reasons.
• 20:26
  SPEAKER_00: Bill Gates, I had reasons.
• 20:28
  SPEAKER_00: I don't even know the kite.
• 20:29
  SPEAKER_00: Jeff Bezos, sure there's reasons, but I don't know.
• 20:32
  SPEAKER_00: I don't know him, never met him.
• 20:33
  SPEAKER_00: So everybody's got the benefit of the doubt.
• 20:34
  SPEAKER_00: But Sam Altman, OpenAI, all right.
• 20:37
  SPEAKER_00: So it's it's free.
• 20:39
  SPEAKER_00: Uh I get a certain amount to use.
• 20:41
  SPEAKER_00: So I said, yeah, all right.
• 20:43
  SPEAKER_00: You know, I got nothing better to do.
• 20:45
  SPEAKER_00: I can't fix anything right now.
• 20:47
  SPEAKER_00: I can still do the core stuff I need to do.
• 20:49
  SPEAKER_00: None of my sellers, none of my clients, nobody I'm working with, nobody's suffering.
• 20:54
  SPEAKER_00: So let's go ahead and play it.
• 20:55
  SPEAKER_00: So sure enough, Connor plays it.
• 20:57
  SPEAKER_00: Connor calls it up.
• 20:58
  SPEAKER_00: Connor starts working with ChatGPT.
• 21:01
  SPEAKER_00: And and I I I tell it.
• 21:05
  SPEAKER_00: Yeah, I'm not trying to leave Claude.
• 21:08
  SPEAKER_00: Claude's gonna renew on Wednesday.
• 21:09
  SPEAKER_00: I'm just out of credits temporarily, but what do you got?
• 21:13
  SPEAKER_00: What can you do?
• 21:13
  SPEAKER_00: And you know, it it says something like, well, let's let's start slow.
• 21:17
  SPEAKER_00: We don't want to go all in.
• 21:18
  SPEAKER_00: Let's just give me just a couple things you might have given to Claude, and let's let's see how that goes.
• 21:24
  SPEAKER_00: Yes, smart up big.
• 21:26
  SPEAKER_00: Sure, yeah, you playing me like a fiddle.
• 21:29
  SPEAKER_00: You're getting the best music out of me right now.
• 21:31
  SPEAKER_00: That's all right.
• 21:31
  SPEAKER_00: What would you suggest?
• 21:33
  SPEAKER_00: They said, Well, what would you normally do at this time if you had a Claude subscription?
• 21:37
  SPEAKER_00: I'm thinking, that's a lot of heavy lifting.
• 21:39
  SPEAKER_00: Jeez, how am I going to do it?
• 21:41
  SPEAKER_00: I said, you know, I have uh I have Mem Palace, I have um Mesh, I have the Mesh I set up on these boxes, so they all talk to each other.
• 21:49
  SPEAKER_00: I have SSH, I have the handshakes going on.
• 21:53
  SPEAKER_00: I have Obsidian Brain as well.
• 21:55
  SPEAKER_00: I threw onto this, on top of this.
• 21:57
  SPEAKER_00: So I have memory going on.
• 21:58
  SPEAKER_00: I say, you know, why don't you look in Google?
• 22:00
  SPEAKER_00: Drive.
• 22:00
  SPEAKER_00: I also have Git set up to kind of save some stuff.
• 22:02
  SPEAKER_00: So why don't you look in Google Drive?
• 22:04
  SPEAKER_00: You'll find some of the processes I work, and this is what I want to do.
• 22:07
  SPEAKER_00: So I give it the instruction.
• 22:09
  SPEAKER_00: And it says, all right, let's do that.
• 22:12
  SPEAKER_00: So I'm watching.
• 22:14
  SPEAKER_00: I'm watching.
• 22:15
  SPEAKER_00: I'm watching.
• 22:16
  SPEAKER_00: So Claude gets into the browser.
• 22:19
  SPEAKER_00: I see it show up.
• 22:20
  SPEAKER_00: It says browser is being controlled.
• 22:21
  SPEAKER_00: I forgot what it said.
• 22:22
  SPEAKER_00: So that's how Claude gets in the browser.
• 22:24
  SPEAKER_00: OpenAI has its own browser inside, kind of like Cursor does with Grokbot.
• 22:29
  SPEAKER_00: Yeah, I played with that too.
• 22:30
  SPEAKER_00: I'll have a whole lot to say about that.
• 22:32
  SPEAKER_00: I don't know.
• 22:33
  SPEAKER_00: It just didn't, you know, it's great, but ChatGPT kind of kind of, yeah, I went all in.
• 22:40
  SPEAKER_00: I went all in.
• 22:41
  SPEAKER_00: I mean I canceled Claude, I went all in.
• 22:43
  SPEAKER_00: Yeah, that's how good it is.
• 22:45
  SPEAKER_00: That's how good it is now.
• 22:47
  SPEAKER_00: That's how good it's it is for what I do.
• 22:49
  SPEAKER_00: And maybe Claude's good for you.
• 22:51
  SPEAKER_00: But anyway, so I give it a little bit, and it's it's working.
• 22:55
  SPEAKER_00: This video thing I talked to you about, yeah, now, now that process, instead of all that other crap, and I'm still making it right, it's still not there yet, but I see so much light at the end of that tunnel where at some point, very soon, just I'm not, I haven't got done setting it up.
• 23:14
  SPEAKER_00: But once I dump in the video, I see an SRT file.
• 23:19
  SPEAKER_00: And maybe I don't even need the SRT file at some point, but I dump it in.
• 23:24
  SPEAKER_00: And I've done a lot of work with this the last three days, folks.
• 23:26
  SPEAKER_00: Today's Thursday.
• 23:28
  SPEAKER_00: I went all in Sunday night.
• 23:30
  SPEAKER_00: On Tuesday morning, I paid the$200.
• 23:35
  SPEAKER_00: I don't think my clawed subscription expires till September, so next month.
• 23:40
  SPEAKER_00: Now, where are we?
• 23:40
  SPEAKER_00: August 27th, 2026, 1745 hour specific standard time.
• 23:47
  SPEAKER_00: I think Tuesday I went all in with Chat GPT.
• 23:52
  SPEAKER_00: From Tuesday to Thursday, lots of heavy lifting.
• 23:56
  SPEAKER_00: I mean, my God, I've got four boxes running this, including my phone.
• 24:02
  SPEAKER_00: But I'm I'm having those interactions.
• 24:04
  SPEAKER_00: Oh, don't get me started on the voice.
• 24:05
  SPEAKER_00: Holy henna.
• 24:07
  SPEAKER_00: The voice talking to it.
• 24:10
  SPEAKER_00: Game changer.
• 24:12
  SPEAKER_00: Yeah, yeah.
• 24:13
  SPEAKER_00: Wrap your mind.
• 24:14
  SPEAKER_00: Try that one.
• 24:15
  SPEAKER_00: Try that just for a little bit.
• 24:17
  SPEAKER_00: And with Claude on the phone, trying to get Claude to talk to the other boxes and and work on stuff.
• 24:22
  SPEAKER_00: I had to go back and and I it just it's it sucked.
• 24:27
  SPEAKER_00: But on the phone, last thing that I talked to it about, what was it?
• 24:32
  SPEAKER_00: I wanted to get a something set up, um, something set up with a client.
• 24:38
  SPEAKER_00: A client wants a workbook created.
• 24:40
  SPEAKER_00: So I kind of is a memory, right?
• 24:42
  SPEAKER_00: So I go here and I say, let's we need to work on that workbook.
• 24:44
  SPEAKER_00: They said, when you get back, just say this about the project, and then I'll bring it up and I'll grab what you said on the phone and I'll make it work.
• 24:52
  SPEAKER_00: Yeah, I don't know what to do.
• 24:54
  SPEAKER_00: I mean, I feel like I'm I'm just gonna be able to sleep all the time.
• 24:57
  SPEAKER_00: The video thing, it's gonna be able to do it from end to end.
• 25:01
  SPEAKER_00: Post it to YouTube, Buzz Sprout, all the social media.
• 25:04
  SPEAKER_00: It's gonna cut it up.
• 25:05
  SPEAKER_00: It's gonna, I'm not gonna even have to edit the beginning and end where you know, where I'm turning around to start talking into the teleprompter.
• 25:13
  SPEAKER_00: This one isn't like that because I'm not, I'm just talking.
• 25:15
  SPEAKER_00: But yeah, it usually I have to go back.
• 25:18
  SPEAKER_00: I go to cap cut, I chop off the front, chop off the back.
• 25:22
  SPEAKER_00: I used to do a lot more um, what are they called?
• 25:25
  SPEAKER_00: Uh, captions in there.
• 25:26
  SPEAKER_00: I'd have it, I'd have it make captions and cap cut.
• 25:28
  SPEAKER_00: Yeah, I'm not really doing a lot of that anymore.
• 25:30
  SPEAKER_00: I think it steps on things, it misspells things.
• 25:34
  SPEAKER_00: And trying to go through like this where we are, 25 minutes.
• 25:36
  SPEAKER_00: Yeah, trying to go through this, yeah, trying to figure out MacIver, trying to spell that right if it doesn't know it, very difficult.
• 25:43
  SPEAKER_00: So I'm not gonna, I I usually don't do captions, but I'm sure it knows me if it's gonna do the captions, and I'm not sure how it's going to do it, but it cuts, it does it end to end.
• 25:53
  SPEAKER_00: So I just do the video and I move on, and then I get a whole bunch of links.
• 25:58
  SPEAKER_00: This is the Buzz Sprite link, this is the YouTube link, these are all the Opus Clip posts, or or videos like Opus Clip that have been posted through a social media poster you have on this, because I have an honor elevated system, which is the light go high level version, a white label of it.
• 26:13
  SPEAKER_00: So I have a social media poster connect.
• 26:15
  SPEAKER_00: Yeah, it gets in there, it gets in there and does that.
• 26:17
  SPEAKER_00: Now, could it actually post to these networks by itself?
• 26:20
  SPEAKER_00: Yeah, it has a browser.
• 26:22
  SPEAKER_00: There's an O auth uh some kind of authorization that probably has to be given.
• 26:26
  SPEAKER_00: But this compared to Claude, this is the game changer, at least for my workflow.
• 26:33
  SPEAKER_00: And it works.
• 26:34
  SPEAKER_00: And as far as building stuff, yeah.
• 26:37
  SPEAKER_00: Cloud code, great, built some awesome stuff.
• 26:40
  SPEAKER_00: I'm now having it go into all the processes I've completed with Cloud Code, all the websites, all that stuff, and I'm having them make it better, redo it, everything, the whole nine yards.
• 26:51
  SPEAKER_00: I'm I'm very happy.
• 26:53
  SPEAKER_00: Now, I also checked out Grok bought through um cursor.
• 26:59
  SPEAKER_00: I I think this is still the game change.
• 27:02
  SPEAKER_00: I haven't given its way and I haven't I haven't done a lot.
• 27:04
  SPEAKER_00: I signed up, I've canceled it since.
• 27:07
  SPEAKER_00: I don't want it too involved in stuff, but I'm very impressed with ChatGPT.
• 27:12
  SPEAKER_00: I guess point being, if you're all in on something, also keep it in mind that you know there might be something else better.
• 27:20
  SPEAKER_00: And I was so, so into Claude.
• 27:23
  SPEAKER_00: I wanted to get a tattoo, not really.
• 27:25
  SPEAKER_00: But you know what I'm saying.
• 27:26
  SPEAKER_00: Like Harley Davidson, right?
• 27:27
  SPEAKER_00: You really want to stick that on your body.
• 27:29
  SPEAKER_00: I guess some people do.
• 27:30
  SPEAKER_00: And if you have one, good for you.
• 27:32
  SPEAKER_00: Way to go.
• 27:33
  SPEAKER_00: But if you give yourself a moment, maybe think, is this one actually better for my workflow?
• 27:39
  SPEAKER_00: Everybody on YouTube says something's better.
• 27:41
  SPEAKER_00: That's why I went all in on Grokbot.
• 27:43
  SPEAKER_00: I need to stop that right now.
• 27:45
  SPEAKER_00: Because I start watching people, then I get distracted, then I say, Yeah, yeah, let me try that.
• 27:49
  SPEAKER_00: So I try it.
• 27:50
  SPEAKER_00: It's cool having all the little bots there, and you know, nine or twelve or whatever it is I had that did different things, scout and recon and all these cool military names.
• 28:00
  SPEAKER_00: And yeah, I just I, you know, you've got to read the stuff and do the stuff it says to do if it's recommending stuff for you to do.
• 28:06
  SPEAKER_00: But I'm very impressed with ChatGPT.
• 28:09
  SPEAKER_00: I'm I'm incredibly happy with it.
• 28:11
  SPEAKER_00: Right now, it looks like it looks like it's gonna do it.
• 28:14
  SPEAKER_00: And I go in, I see, you know, this looks like some kind of a duplicate thing that I started working on because I didn't realize we already had this workflow.
• 28:22
  SPEAKER_00: And it I haven't run out yet.
• 28:24
  SPEAKER_00: It looks like these workflows kind of go on forever.
• 28:27
  SPEAKER_00: The other thing that sucked about Claude, and again, maybe I'm not, maybe I wasn't using it the best way, but it would fill up.
• 28:34
  SPEAKER_00: And then once it fills up, it's so much context, it choked.
• 28:38
  SPEAKER_00: It would start choking.
• 28:39
  SPEAKER_00: So I say, I need to take this to a new chat.
• 28:42
  SPEAKER_00: I need to start a new chat with you.
• 28:43
  SPEAKER_00: What do I say to have you remember?
• 28:45
  SPEAKER_00: And you know, it always wasn't complete.
• 28:47
  SPEAKER_00: It might miss something, or there was some context it didn't bring over.
• 28:51
  SPEAKER_00: I haven't run out of anything with this.
• 28:54
  SPEAKER_00: This just seems to just keep going.
• 28:56
  SPEAKER_00: And I so I'm much more confident if I have a workflow for a client that I need to work on, things that we've talked about.
• 29:04
  SPEAKER_00: I'll I'll take uh, and just like I used to do with Claude, all of my uh when people are book with honor, they go to bookwithhonor.com to book time with me, to hire me, to do whatever for them.
• 29:15
  SPEAKER_00: Not killing folks.
• 29:16
  SPEAKER_00: We don't do any of that, but you know, to do something with me, to build out AI stuff and all that, or real estate, whatever it may be.
• 29:22
  SPEAKER_00: I'll record that process whenever I'm talking to them about their workflow, trying to see where AI would fit best, what we can implement, what exists, what's not bullshit, because there's so many people saying, oh, you need to be an AI alchemist.
• 29:36
  SPEAKER_00: No, it's not alchemist.
• 29:37
  SPEAKER_00: Arbitrage, arbitrary not that's a tree thing.
• 29:40
  SPEAKER_00: Arbitr?
• 29:41
  SPEAKER_00: Anyway, but you have to be arbitrage.
• 29:43
  SPEAKER_00: You basically you have to replace, no, no, it doesn't, you can't go fishing all day and have AI just bring in the checks and put them in the bank.
• 29:54
  SPEAKER_00: That's that's not how that works.
• 29:56
  SPEAKER_00: And the people that are selling that, they wouldn't be selling it if that's how it worked.
• 30:00
  SPEAKER_00: They're trying to get clicks and views and all this other stuff.
• 30:03
  SPEAKER_00: Come on, you you know that game.
• 30:04
  SPEAKER_00: If you're watching this, you know the game.
• 30:06
  SPEAKER_00: I'm impressed with Chat GPT.
• 30:09
  SPEAKER_00: That's it.
• 30:10
  SPEAKER_00: That's where we are right now, August 27th.
• 30:12
  SPEAKER_00: Tomorrow, maybe it's gonna be grok.
• 30:14
  SPEAKER_00: Maybe Elon's gonna hit it out of the park with the cursor stuff and all this other thing.
• 30:18
  SPEAKER_00: Maybe that's gonna be the thing.
• 30:20
  SPEAKER_00: But understand, I guess here's the caution.
• 30:24
  SPEAKER_00: You start having it automate too much and then it gets taken away, or you run out of credits, or whatever.
• 30:32
  SPEAKER_00: Yeah, it's it's tough to continue.
• 30:35
  SPEAKER_00: Just like the LAPD way.
• 30:38
  SPEAKER_00: You went and you went from maps to tom's to other types to now, you can't go anywhere without activating your phone and having it give you directions because you've forgotten and you don't really remember how to traverse the San Fernando Valley or wherever you're going.
• 31:02
  SPEAKER_00: We don't want this to dull us so much that we forget basic procedure.
• 31:08
  SPEAKER_00: So that's something else you might want to wrap your mind around.
• 31:12
  SPEAKER_00: Phone numbers.
• 31:13
  SPEAKER_00: I have a couple memorized, that's it.
• 31:16
  SPEAKER_00: No others.
• 31:17
  SPEAKER_00: I couldn't tell you what phone numbers were for people.
• 31:21
  SPEAKER_00: Not anymore.
• 31:22
  SPEAKER_00: It's all in my phone.
• 31:23
  SPEAKER_00: God forbid if I get booked into some jail somewhere in the world, other than the United States or even in the United States, that one phone call, yeah.
• 31:34
  SPEAKER_00: I got two numbers I can call.
• 31:36
  SPEAKER_00: Other than that, that's it.
• 31:39
  SPEAKER_00: And hopefully they'll pick up.
• 31:40
  SPEAKER_00: Yeah, anyway.
• 31:42
  SPEAKER_00: You know what I'm saying?
• 31:43
  SPEAKER_00: I hope this is well.
• 31:44
  SPEAKER_00: I'm Connor with Honor.
• 31:45
  SPEAKER_00: I uh enjoy this one.
• 31:46
  SPEAKER_00: This was like a venting session.
• 31:48
  SPEAKER_00: And if you have any comments, please let me know.
• 31:51
  SPEAKER_00: I'd like to know where you stand on all these things.
• 31:53
  SPEAKER_00: All right.
• 31:54
  SPEAKER_00: Don't trust too much.
• 31:55
  SPEAKER_00: Trust yourselves, trust a few people close to you.
• 31:58
  SPEAKER_00: More than a handful of friends.
• 31:59
  SPEAKER_00: Yeah, not a fan.
• 32:00
  SPEAKER_00: I'll be well.
• 32:01
  SPEAKER_00: You be well.
• 32:02
  SPEAKER_00: We all be well.
• 32:02
  SPEAKER_00: Connor with honor, over now.