Latest / Connor with Honor: AI, Real Estate and Real Life / They Banned the AI That Refused to Spy on You: Front Sight Focus | March 10, 2026

Transcript

They Banned the AI That Refused to Spy on You: Front Sight Focus | March 10, 2026 · Connor with Honor: AI, Real Estate and Real Life

• 0:00
  SPEAKER_00: Good everybody.
• 0:00
  SPEAKER_00: I've spent years on ranges.
• 0:03
  SPEAKER_00: LAPD Academy in service, IOD officers trying to get back in service, of course, motor guys, SWAT, SIS, some of the most specialized units on the Los Angeles Police Department.
• 0:14
  SPEAKER_00: And when I've trained them, as far as SWAT and SAS, those super specialized forces, I was more of a range safety officer, if you will.
• 0:23
  SPEAKER_00: Train the other ones, yes, but you know, those specialized units, they, well, they have a candle all their own.
• 0:29
  SPEAKER_00: So they work very well.
• 0:30
  SPEAKER_00: And the best thing I could have done is maybe made a suggestion or an adjustment, but they're in those specialized sections because, well, they're special, right?
• 0:38
  SPEAKER_00: And one thing, there's one thing that separates those people who hit what they're aiming at and the people that don't, and it's not the gun, it's not the stance, all of that stuff matters.
• 0:48
  SPEAKER_00: It's not really the ammunition that matters a little bit too, but it's where your eyes tend to go under pressure.
• 0:54
  SPEAKER_00: Now, when something scary is happening, when the stakes are real and your heart is doing 140 beats per minute and you're you're trying to suck wind, you're breathing because you just sprinted, you know, 100 yards or whatever to get to cover, your eyes naturally go to the target, the threat, the thing that worries you.
• 1:12
  SPEAKER_00: At that moment, the moment that happens, your front sight starts to go blurry, and of course, you miss.
• 1:18
  SPEAKER_00: You have to train yourself to pull your eyes back to that front sight post sharp and clear, why everything around it stays blurry.
• 1:25
  SPEAKER_00: That's the only way you hit what you're aiming at.
• 1:28
  SPEAKER_00: You might get lucky, but lucky doesn't win gun fights.
• 1:31
  SPEAKER_00: Well, lucky doesn't always win gunfights.
• 1:34
  SPEAKER_00: Lucky, lucky isn't a bad skill to have.
• 1:36
  SPEAKER_00: But AI right now, it's a target-rich environment, billions of dollars moving, headlines about the front end, work governments picking sides.
• 1:43
  SPEAKER_00: Your phone's about to start making your decisions for you.
• 1:46
  SPEAKER_00: A lot of noise, a lot of blur.
• 1:48
  SPEAKER_00: And today we're gonna find that front site.
• 1:50
  SPEAKER_00: Today's March 10th, 2026.
• 1:53
  SPEAKER_00: So let's go.
• 1:54
  SPEAKER_00: This is going to be your first segment, segment one.
• 1:57
  SPEAKER_00: We're gonna talk about the Great Divorce as I get this more centered.
• 2:02
  SPEAKER_00: Anthropic versus the Department of Defense.
• 2:04
  SPEAKER_00: Now, the story you need to know this week is the split between Anthropic, and this has been rolling out for a little while now, probably eight, nine days.
• 2:12
  SPEAKER_00: The company that makes Claude, that anthropic, the AI I used to run my businesses every single day, and of course, apparently the U.S.
• 2:20
  SPEAKER_00: uh.
• 2:23
  SPEAKER_00: But here's what happened: the DoD designated Anthropic as a natural national security chain risk.
• 2:29
  SPEAKER_00: The Secret Service pulled Claude from their internal workflows, and federal agencies are being directed to phase Anthropic out over the next six months.
• 2:38
  SPEAKER_00: Now, supply chain risk usually means foreign spy software.
• 2:44
  SPEAKER_00: Anthropic is an American company founded by people who left open AI specifically because they thought AI was being built too fast and too recklessly.
• 2:54
  SPEAKER_00: These are not bad actors.
• 2:55
  SPEAKER_00: They're the ones who actually wrote the safety playbook.
• 2:58
  SPEAKER_00: When a lot of people aren't paying attention to safety, it seemed like they were on board with safety.
• 3:03
  SPEAKER_00: Well, so what actually happened?
• 3:05
  SPEAKER_00: Anthropic refused two things.
• 3:07
  SPEAKER_00: They would not allow their AI to be used for mass domestic surveillance of American citizens, and they wouldn't allow their AI to make legal, autonomous decisions to basically press that trigger without a human being in the loop.
• 3:20
  SPEAKER_00: Now, there was language on there that talked about ethical and reasonable and other things as well.
• 3:26
  SPEAKER_00: In the language, they I think wanted more assurances when it came to what the government was going to use their technology for.
• 3:33
  SPEAKER_00: But you'll have to really dive into that to really, really wrap your mind around.
• 3:38
  SPEAKER_00: This is like anything else in our world today, especially.
• 3:41
  SPEAKER_00: It's up to you to make your own mind up whether you think that they made a good move or a bad move.
• 3:46
  SPEAKER_00: Now, here's the front sight on this, and I want you to really hear this.
• 3:50
  SPEAKER_00: I spent a lot of years in a system, and I can tell you from experience, the person doing things mostly by the book, the most documented, the most careful, the person that sometimes makes the people who benefit from cutting corners very uncomfortable, not because they're doing something wrong, but because they're doing something right.
• 4:09
  SPEAKER_00: And that right is a mirror, and it's a very powerful mirror.
• 4:12
  SPEAKER_00: People who do not like what they see in mirrors.
• 4:16
  SPEAKER_00: Powerful people don't like it.
• 4:18
  SPEAKER_00: So here's what's happening here.
• 4:20
  SPEAKER_00: The lesson isn't really political, the lesson is practical.
• 4:23
  SPEAKER_00: If you're building any part of your business or your life on a single AI tool, a single vendor, a single platform, you are the one, you are one government contract away from that thing disappearing from you.
• 4:36
  SPEAKER_00: The Secret Service rebuilt their AI workflow in a weekend because they built it out to be model agnostic.
• 4:44
  SPEAKER_00: They can swap out the brain.
• 4:45
  SPEAKER_00: That's how you build.
• 4:47
  SPEAKER_00: Not married to one tool, be platform agnostic, always capable of switching.
• 4:52
  SPEAKER_00: And I only say this because ChatGPT, Claude, Grok, these are all wonderfully fantastic large language models.
• 5:00
  SPEAKER_00: And don't forget Gemini, just fantastic.
• 5:03
  SPEAKER_00: But again, what if something gets shut off and you have your whole life in it?
• 5:07
  SPEAKER_00: You want to make sure that you're always keeping records of your interactions just in case you have to switch, and especially if you've become very dependent on it.
• 5:16
  SPEAKER_00: I built my whole operation on or elevate around that principle, not because I'm smart, but because I've seen what happens when you put all your leg weight on one leg.
• 5:26
  SPEAKER_00: Now, NVIDIA, Vera's Rubin, and what cheap AI actually means.
• 5:31
  SPEAKER_00: NVIDIA, I unveiled, has unveiled its Vera Rubin platform earlier this month.
• 5:38
  SPEAKER_00: Six chip co-design.
• 5:41
  SPEAKER_00: The result training AI models at the frontier level, the kind of models that power everything you use.
• 5:47
  SPEAKER_00: It's now roughly 10 times cheaper than it was 12 months ago.
• 5:50
  SPEAKER_00: So costs are coming down in every every every possible avenue here in artificial intelligence.
• 5:57
  SPEAKER_00: Google Gemini 3.1 Pro,$2 per million tokens in 2024.
• 6:02
  SPEAKER_00: The same level of reasoning costs 50 times that in 2024.
• 6:06
  SPEAKER_00: What does that mean in real language?
• 6:08
  SPEAKER_00: Intelligence is becoming utility, like water, like electricity.
• 6:12
  SPEAKER_00: You don't think about electricity when you flip the switch, you just expect it to work.
• 6:16
  SPEAKER_00: AI is heading there and it's heading there very fast.
• 6:19
  SPEAKER_00: Now, here's where people are going to get this wrong.
• 6:21
  SPEAKER_00: Most people hear AI is getting cheaper and they think, good, I'm going to use more of it.
• 6:25
  SPEAKER_00: That's not the front sight.
• 6:27
  SPEAKER_00: The front sight on this one is this.
• 6:29
  SPEAKER_00: When intelligence is cheap, when everybody has access to a 160 IQ assistant for the price of a cup of coffee, being smart stops being a competitive advantage.
• 6:40
  SPEAKER_00: I lost 135 pounds.
• 6:41
  SPEAKER_00: And that's still a battle that continues.
• 6:43
  SPEAKER_00: I think it'll probably continue the rest of my life.
• 6:45
  SPEAKER_00: I'm never going to be at a point where I'm just okay with it.
• 6:49
  SPEAKER_00: I still have it lurking in the back of my brain.
• 6:52
  SPEAKER_00: Maybe you have something similar with some addiction that you might have given up or have quit and continue to kind of fall, get closer to that line and then move away.
• 7:01
  SPEAKER_00: And I tell you something nobody in the fitness world wants to say out loud.
• 7:05
  SPEAKER_00: Once you know how to do it, once fasting is explained to you correctly, or however you're going to do it with caloric restriction and you commit, once those hormones are balanced, once you understand what's actually happening with insulin and how it stores fat and how to lower it and not store fat, the information is not the hard part, but the hard part's doing it.
• 7:24
  SPEAKER_00: The application, the judgment, that daily decision not to eat the donut.
• 7:29
  SPEAKER_00: That's exactly where AI is taking us.
• 7:31
  SPEAKER_00: The information layer is being commoditized.
• 7:33
  SPEAKER_00: The execution layer, the judgment layer, that's where the value lives.
• 7:37
  SPEAKER_00: That's where mine lives.
• 7:39
  SPEAKER_00: Don't compete on knowing things.
• 7:42
  SPEAKER_00: In two years, knowing things is going to be table stakes.
• 7:45
  SPEAKER_00: Compete on doing the right thing with what you know, on judgment, character, and on trust.
• 7:51
  SPEAKER_00: That's the front side of this whole thing.
• 7:53
  SPEAKER_00: AI is already inside your phone.
• 7:55
  SPEAKER_00: What must most people don't know it?
• 7:58
  SPEAKER_00: It's practical.
• 7:59
  SPEAKER_00: Now, Samsung is now rolling out Gemini.
• 8:01
  SPEAKER_00: AI to 800 million devices.
• 8:04
  SPEAKER_00: Gemini AI to 800 million.
• 8:07
  SPEAKER_00: Not a chatbot update, but agentic AI.
• 8:10
  SPEAKER_00: So your phone watches your emails, your calendar, your shopping list.
• 8:13
  SPEAKER_00: Their AI vision in the refrigerators recognizes types of fresh fruit and orders replacements when they expire.
• 8:21
  SPEAKER_00: If you don't have one of those, that's what those fancy refrigerators do.
• 8:25
  SPEAKER_00: This isn't science fiction.
• 8:27
  SPEAKER_00: This is March 10th, 2026.
• 8:29
  SPEAKER_00: The thing you need to understand is what agentic means.
• 8:33
  SPEAKER_00: It doesn't mean that AI answers your questions.
• 8:35
  SPEAKER_00: It means the AI does things on your behalf without you necessarily being in the room.
• 8:41
  SPEAKER_00: Now, I trained firearms instructors.
• 8:43
  SPEAKER_00: I trained people in the execution of basic drills from never held held having held a handgun or shotgun before to having their first experience with it.
• 8:55
  SPEAKER_00: One of the things we taught, and this translates directly, is that a tool is only good and as disciplined as the person holding it.
• 9:04
  SPEAKER_00: You can have the best equipment in the world.
• 9:06
  SPEAKER_00: If the operator doesn't have protocols or rules in place, doesn't know when to engage and when not to engage, doesn't audit the outcome, that tool becomes a liability.
• 9:15
  SPEAKER_00: Your AI agent is the same thing.
• 9:18
  SPEAKER_00: Here's the move this week.
• 9:19
  SPEAKER_00: Very simple.
• 9:20
  SPEAKER_00: You're going to go to your phone settings right now.
• 9:23
  SPEAKER_00: You're going to find what your AI assistant has access to.
• 9:26
  SPEAKER_00: Email, calendar, contacts, location, purchase history.
• 9:30
  SPEAKER_00: Most people have never even looked at this.
• 9:32
  SPEAKER_00: You set it up, you clicked allow all, and you walked away.
• 9:36
  SPEAKER_00: This is the same as buying a gun.
• 9:38
  SPEAKER_00: Loading it, setting on the counter, and walking away.
• 9:41
  SPEAKER_00: The gun doesn't have judgment, neither does the agent.
• 9:45
  SPEAKER_00: Know your tools, have access to.
• 9:48
  SPEAKER_00: Know what your tools have access to.
• 9:50
  SPEAKER_00: Know what they're doing on your behalf.
• 9:52
  SPEAKER_00: Maintain the human in the loop.
• 9:54
  SPEAKER_00: That's not paranoia.
• 9:55
  SPEAKER_00: That's operational discipline.
• 9:57
  SPEAKER_00: Same thing I applied on the job.
• 9:59
  SPEAKER_00: Same thing I apply every day now.
• 10:03
  SPEAKER_00: What is the human cost?
• 10:05
  SPEAKER_00: The lawsuit against Google that nobody is really talking about correctly.
• 10:10
  SPEAKER_00: A father is suing Google.
• 10:14
  SPEAKER_00: His son Jonathan is dead.
• 10:16
  SPEAKER_00: The allegation is that Gemini, Google's AI, maintained engagement with his son, and his mental state deteriorated, treated his psychosis, according to the lawsuit, as plot development, kept the conversation going, kept him engaged because that's what the system was built to do.
• 10:35
  SPEAKER_00: Similar cases against open AI are working through the courts right now.
• 10:40
  SPEAKER_00: I'm not going to try this case.
• 10:41
  SPEAKER_00: That's not my lane.
• 10:43
  SPEAKER_00: I spent 20 years writing reports and tickets and everything else.
• 10:47
  SPEAKER_00: I know how to state facts and letting the facts speak is probably most important.
• 10:52
  SPEAKER_00: And then having the Monday night quarterback is probably doubly so.
• 10:56
  SPEAKER_00: I'm not going to try this case.
• 10:58
  SPEAKER_00: That's not my lane.
• 10:59
  SPEAKER_00: The fact is, these systems are built to maximize engagement, keep you talking, keep you coming back.
• 11:05
  SPEAKER_00: That's the business model.
• 11:06
  SPEAKER_00: So for most people, most of the time, that works out great.
• 11:09
  SPEAKER_00: But for somebody that is having a mental health crisis or has issues in mental health in that world, for someone who isn't already not tethered to reality, an endless patient, infinitely available, never judging entity that validates whatever you say, is not a therapist.
• 11:27
  SPEAKER_00: It's an accelerant.
• 11:29
  SPEAKER_00: I've been on calls where a crowd forms at the wrong with the wrong voice.
• 11:34
  SPEAKER_00: The one keeps agreeing, keeps egging on, turns a manageable situation into a critical one.
• 11:41
  SPEAKER_00: We call it a brignoni state where somebody just kind of is that catalyst to keep them.
• 11:46
  SPEAKER_00: You're not going to let them talk to you that way.
• 11:48
  SPEAKER_00: You're not going to stand for that.
• 11:49
  SPEAKER_00: You're not going to take that from that guy.
• 11:51
  SPEAKER_00: You know, that that person.
• 11:54
  SPEAKER_00: They turn manageable situations into critical ones.
• 11:57
  SPEAKER_00: The most dangerous thing in certain rooms is an audience with no judgment.
• 12:02
  SPEAKER_00: Those AI systems are audiences with no judgment built in by default.
• 12:07
  SPEAKER_00: This needs to change, and it will change.
• 12:09
  SPEAKER_00: The question is whether the companies change it because it's right or because the lawsuits make not changing it too expensive.
• 12:16
  SPEAKER_00: Based on what I've seen of human institutions, my entire career, I'll tell you, I'll guess, I'll let you guess which one moves faster.
• 12:25
  SPEAKER_00: The front side here.
• 12:27
  SPEAKER_00: If you have people in your life, family clients, anyone who's struggling, do not point them to an AI chatbot as a solution, point them to a human.
• 12:36
  SPEAKER_00: These tools have not earned the level of trust, at least not yet.
• 12:39
  SPEAKER_00: Now, am I in agreement with these big companies, lawyers, doctors, shutting down the AI access to try to get some help and discuss things with them?
• 12:51
  SPEAKER_00: No, I don't think I am.
• 12:52
  SPEAKER_00: If you just look at the basic doctor reports, if you have a doctor giving a diagnosis, there's a percentage.
• 12:58
  SPEAKER_00: You attach a doctor to an AI, the percentage gets a little bit better.
• 13:03
  SPEAKER_00: You put AI on it all by itself, it gets dramatically better.
• 13:07
  SPEAKER_00: Do the research and see what that's about.
• 13:09
  SPEAKER_00: You know what AI right now is like?
• 13:12
  SPEAKER_00: It's like the first time I stood in front of a group of LAPD recruits on the range.
• 13:16
  SPEAKER_00: Brand new.
• 13:17
  SPEAKER_00: Hadn't been holding anything real yet.
• 13:20
  SPEAKER_00: They were holding real guns for the first time in a real context.
• 13:25
  SPEAKER_00: And every single one of them, everyone wanted to look at the target.
• 13:29
  SPEAKER_00: Their eyes went right to the threat, to the outcome, to what they're scared of.
• 13:34
  SPEAKER_00: And the bullet went everywhere, but where it needed to go.
• 13:37
  SPEAKER_00: I had to teach them one of the hardest things a person can learn.
• 13:41
  SPEAKER_00: In the most important moment, you need to trust the process, trust the site, and resist the instinct to look at what scares you.
• 13:49
  SPEAKER_00: That's where we are with AI.
• 13:51
  SPEAKER_00: The headlines are screaming, the numbers are massive, the stakes are real, and everyone is looking at the target, the fear, the hype, and the drama, and they're missing.
• 14:03
  SPEAKER_00: Pull your eyes back.
• 14:05
  SPEAKER_00: Front sight, sharp, everything else is blurry, and you hit what you're aiming at now.
• 14:12
  SPEAKER_00: This week AI got more political, more powerful, and more personal than any week I can remember since I started paying close attention back in 2021.
• 14:23
  SPEAKER_00: A company that tried to do it right got punished for drawing a line.
• 14:28
  SPEAKER_00: A chip platform dropped the cost of intelligence by 90%.
• 14:32
  SPEAKER_00: Your phone started making decisions on your behalf, whether you noticed it or not.
• 14:37
  SPEAKER_00: And a father is in court because an AI kept his son engaged when disengagement may have saved his life.
• 14:45
  SPEAKER_00: That's the range we're working on.
• 14:47
  SPEAKER_00: Those are real targets.
• 14:49
  SPEAKER_00: You don't need to be scared of this.
• 14:51
  SPEAKER_00: Fear is what takes your eye off the front sight.
• 14:54
  SPEAKER_00: You need to be trained, disciplined, present, and honest about what these tools are and what they're not.
• 15:02
  SPEAKER_00: I've been fat, I still battle with it every day.
• 15:05
  SPEAKER_00: I know what it's like to be in a system, your own body, the food industry, the medical establishment that was telling you everything except the truth you needed to hear.
• 15:15
  SPEAKER_00: Put the forks down and step away from the table.
• 15:19
  SPEAKER_00: And I know that things the day things changed was not the day I got new information.
• 15:24
  SPEAKER_00: It was the day I decided to look at what was real instead of what was comfortable.
• 15:30
  SPEAKER_00: That's the posture you need right now with AI.
• 15:33
  SPEAKER_00: Look at what's real, not what's comfortable, not what's scary.
• 15:37
  SPEAKER_00: Look at the front sight.
• 15:39
  SPEAKER_00: I'm Connor McIver.
• 15:40
  SPEAKER_00: I'll see you in the next one.