Australia Medicare AI Breach ( #OpenAI ) Prompts Investigation

Australia Medicare AI Breach ( #OpenAI ) Prompts Investigation
Crypto Talk Radio: Basic Cryptonomics
Australia Medicare AI Breach ( #OpenAI ) Prompts Investigation

Sep 29 2026 | 00:29:40

/
Episode • September 29, 2026 • 00:29:40

Hosted By

Leicester

Show Notes

Australia Medicare AI Breach (OpenAI) Prompts Investigation

#Crypto #Cryptocurrency #podcast #BasicCryptonomics 

Website: ⁠⁠⁠⁠https://CryptoTalk.FM

Facebook: ⁠⁠⁠⁠@ThisIsCTR⁠⁠⁠⁠

Chapters

  • (00:00:01) - Intro and Gaming Chat
  • (00:01:09) - Bitcoin and Ethereum Market Update
  • (00:02:19) - AI Rant Preview and Personal Disclosure
  • (00:03:21) - The COBOL Shortage Story
  • (00:05:30) - Retirement, Military Age Standards, and DEI Rollbacks
  • (00:07:03) - Reflections on Military Service and Personal Autonomy
  • (00:08:37) - The Shingles Ordeal
  • (00:13:42) - Bodily Autonomy and Forced Mandates
  • (00:15:54) - Career Choices and Remote Work Preferences
  • (00:17:31) - Financial Stability and Life Balance
  • (00:18:17) - AI Breach in Australia and How AI Agents Work
  • (00:21:03) - Scoping AI Instructions and Boundary Risks
  • (00:23:16) - Apple vs Windows Security and AI Sandboxing
  • (00:24:09) - Incompetence, AI Risks, and Infrastructure Threats
  • (00:25:59) - Tech Serving Humans vs Humans Serving Tech
  • (00:28:15) - Closing Thoughts on AI Dependence
View Full Transcript

Episode Transcript

[00:00:01] Speaker A: Welcome to Crypto Talk Radio, the podcast for everyday investors like you. Visit us on the [email protected] and now here's your host, Leister. [00:00:13] Speaker B: Thank you for that, Bailey, and welcome everybody out there. And Crypto Talk radio found at cryptotalkradio.net Happy Tuesday once again. I know that there are no gamers out there. I understand I've been occupied with Fire Emblem, fortune's weave. I do recommend it for the patient out there. It's not for everybody. You have to have some patience. You have to be tolerant and patient. Care about the world, want to learn about its people. I don't play garbage like Witcher 3. That's crap. I'm talking enriching games. You know, games that they're not wide open. Create your own destiny. Garbage. It's a scripted deal. It's like a book. That's what I like. You get enriched, you learn, you grow, evolve. That's just me. Today there's going to be a little bit of a rant about AI. I'll apologize in advance, but I think it's necessary based on some stuff that I recently read. Coinmarketcap.com we will zoom out to the month chart starting with Bitcoin. And just up front, I apologize if I sound a little bit strange. Have a minor, very minor medical issue. So it's affecting how I speak. So I apologize in advance, but it won't be a long episode. Bitcoin currently hovering shy of the $84,000 mark, showing a decently strong but not sustained upward trend. Ethereum on that same month chart, currently hovering shy of the $2,700 mark, shows a decently strong but slightly downward trend. Tells me personal opinion money may be flowing towards Bitcoin as a hedge against things people see forecasted. Usually that's going to be talking about inflation, might be talking about some other type of risk that I'm not aware of, but inflation is almost always at the top of the list. I still maintain if you're going to get into things, diversification is the best strategy you can employ at the moment. And I do recommend going to get into these things. Bitcoin and Ethereum can't really steer you wrong. Most of the main ones can't steer you wrong. But specifically Bitcoin and Ethereum cannot steer you wrong. I said today is going to focus as a rant, not very much crypto because there's not much you didn't already know. But I needed to get it off my chest. Some that have listened to Me for a while. CryptoTalk FM have heard me go a little bit off kilter about AI for disclosure. Technology is what I do for a living. So I have a job. I don't, I don't gamble crypto. I discuss it because I like to share information to help keep people safe. That's my motivation. That's why my show's not monetized. Doesn't need to be. I make a lot of money on my regular sell my own business. I make my own money there. And it's. It's not what I enjoy doing, but I have an aptitude for it. I'm good at it, been doing it for a while. I see that it's a niche product and I see that the vendor is screwing up decisions where it'll be harder to find clients. It's not that they're out, they're not out there, they are. But they're underpricing the market. They don't understand. So let me hit you with regale you a little bit of a story. You might have seen the dues and if you didn't, I'll summarize it and I encourage you go look into this. If you heard of the term cobol. C O B O L COBOL is an older technology, mainframe stuff. Turns out there's a lot of government agencies, mostly government insurance to a lesser degree, but government certainly that still depends on cobol. COBOL is a very challenging language. It's not that people couldn't learn it, you could. But it's a very challenging language to master. Most of the COBOL developers are long since retired, mostly because just by raw age and the rise of more traditional computing and software programming, which was expected to replace most of the mainframe applications, no different than regular computing and word processing replacing typewriters. Turns out typewriters are still there. Turns out mainframe still there. Because there just happened to be strong need in applications for these. The government recently, I believe it's New Jersey, but the government recently had put out a call for help. They were screaming at the top of their lungs because they needed COBOL developers. They were some applications they were struggling to keep running and support. And they put out a call for COBOL developers and it was. They were having a hard time finding them. It's not taught in schools anymore. When I went through college, which by the way, I went through college after I was already in the workforce, after I already knew my craft, after I was already decades in it, and I only went there as a formality because the Application process is broken and all they care about is to see the stupid piece of paper. They don't care that you have the skills and experience in the field, which I'd already had. So I did it as a formality. They don't teach it in the colleges, they simply do not. So they had to go to these people that had retired and put out a call to try to get some of them back. Now this will be a little bit of a rabbit hole. Bear with me. You may not understand that the whole cycle of retirement, the whole intent behind retirement, I don't want to say it's a scam, that's not fair, but the whole intent behind retirement, the workforce wants young, healthy talent. This has its origins in the military. Used to be, and I was told this changed, but it used to be that in order to get enlisted in the military at all, you had to be a certain age or younger. And in order to be an officer versus just a regular recruit, you had to have college degree. But they were expecting you to be of a reasonably young age. And then they did the health standards. There's nothing wrong with the health standards, but usually as you get older, it's harder to meet those standards. Right. The Biden administration rolled some of those back because they were trying to welcome, you know, everybody can play all inclusive dei that. So everybody can play. Of course, that's a national security risk. Pete Hegseth gets in there and he's saying, no, we're rolling this back. No more wearing dresses, no more doing this. And roll back some of those standards to go back to the enforcement of age and health standards and things. The age restriction is the primary reason that I never went into the military. Because at the time when I was within the range, I said that if I cannot find a job, I'll enlist. But I was always able to find a job some way or another. And I realized thinking back mentally, I was not mature enough to be in that military. Could I have pulled it off? Physically, yes, I was in immaculate shape. But mentally, I was not prepared for that. Mentally, I would not have sustained. I think it would have been good for my career. But mentally, I was not in the right place to be in that kind of space back then if there was no age limit and I could have been an officer, meaning not deployed, not having to meet health standards, et cetera, without the four year degree requirement. And somehow I had enlisted, let's say, in my 30s, which is a long time ago now, I think I would have been in a better spot mentally Physically, not so much. But mentally, certainly I would have, I would have been much more level headed. I would have been able to deal with what was coming at me, you know, people screaming in your face and all that. I would have been able to handle it better. I don't know that I would have been able to handle being commingled with some of them, some of my peers, certainly not when they were doing the forced, you know, my body, my choice violations with the forced jabs. I don't think I would have dealt with that very well. And I know me, when you compel me to do something, I'm going to chafe against it. When it's my choice to do something and you sell me on why I should do it, no problem. Case in point, in 2023, I caught shingles. I don't know who has caught shingles. Who's listening? Chances are you may not have, but shingles is, you know, I've just to put in perspective here, I have fallen off of high places. I've been hit by moving cars, I've fallen off of moving cars, high impact in sports. I have had weapons. I've had weapons hit me. There has never been pain that I've ever experienced anywhere close to shingles. Not even close. Not banging your finger in a door jam or a car. Not having your. I've been, I've had a tire, I've been run over by car. Like there's nothing that comes close in terms of pain. Scope as shingles, shingles to describe it. Picture in your mind that somebody is pulling on your arm. Because it was on my arm. Sometimes it's on your neck, sometimes it's on your back, sometimes it doesn't matter. But in my case it was on my arm. It was all up down my arm from my upper shoulder blade all the way down to my wrist on my inner arm because it was on my arm. Picture if you can. There's something you cannot see or someone you cannot see and they're tugging on your arm and they're pulling it down against your will. That's what kept happening. Involuntary movement of my arm, involuntary pulling and then a constant pain. Not quite burning pain, not quite stinging pain. It's like a blend between burning and stinging while this is happening. As if your arm is a rubber band, you know, and you're pulling and pulling and pulling. Something's pulling and pulling and pulling. It's just stretching and pulling and pulling and there's a tautness. Look it up to this feeling and it's involuntary. There's nothing you can do about it. It's distracting. You can't. Like I was not getting very much good sleep because of it. I figured out now, this is prior to me understanding that it was shingles, because I figured I had sprained something or dislocated something. Like it didn't. It wasn't what I expected to feel with shingles because I had never had it before. I had a. A sleeve, shoulder, arm sleeve that I would wear. And then I put a, like a tourniquet type deal. And that seemed to help for sleeping. Then the pain, that, that kind of racking pain. Calm down a little bit. And then I start getting these. They're like. They look like jellyfish type blisters on my arm all the way up, down, and they would leak. And of course, these are infected. Just before, when it started to be small bumps, that's when I had consulted the doctor who diagnosed me with shingles. And then there's a cream I had to keep applying. And of course, it's a messy mess because they're. They're. They're very thin. They're leaking. But I had to put the cream on to help cure the skin. And there were some pills I had to take to help with the pain. Now turns out the doctor, because it took him, like, I want to say, two weeks to get back with the prescription. And you're supposed to get care from the moment you get onset of symptoms in 14 days. So I was well out of that because the doctor took today long and I didn't recognize it for shingles. So I'm screwed, basically, permanently. I'm permanently damaged. It's a. It's a nervous system deal. It's what it is. I'm describing that because when I was going through that, and again, it's the worst pain I'd ever encountered. This is near the time that my brother passed away. So I'm dealing with all that at the same time. I need to get out of where I'm at for multiple reasons. It was a bad time. I was. I was working and I went to, you know, I was on the job constant, all time. But I was not in a state of mind, and I'm not now. But I was not in a state of mind to allow anybody to dictate to me what to do. It's my choice. It's my choice to go to the doctor. It's my choice to go to this job. It's my choice to get a shot or not. And with shingles, once I experienced it, I had done research to say, okay, since I've had it, could I catch it again? The answer is theoretically yes. Although rare, it is possible. So then I said, well, what about a vaccine? They actually don't recommend you take the shingles vaccine until you're, like, older, like in your 50s, 60s. Well, I'm close to that early age, but I'm not quite there yet. But I'm essentially screwed to the risk of fate that it's possible I catch it again and again. It was a very not fun feeling going through shingles, but nobody was forcing the issue simply for the risk of somebody else catching it. They actually discouraged doing it because of age and your deal. Whereas in other situations in the military, they were forcing people to do it simply because other people were afraid of catching it. Even though taking certain shots did not stop you from catching the disease, those shots were designed to help keep you alive or increase your chances of staying alive. Nobody's died from shingles, so that may have been some sort of a catch to. Well, in this case, we have an exception. My point is I always had a beef and a problem with the narrative of anything being forced on or in me. It's something I've never been supportive of, and I've always chafed against it. It's the reason I've left jobs on countless occasions, because they were trying to force something that I knew was ethically wrong, or I had confrontations where I knew I was going to get into some sort of altercation if I stay. That's how I know I would have been at that time in the military with them trying to force anything. So when I was younger, I know I would have been very combative because I was very resistant not to authority. But you had to be correct to authority. You had to make it make sense. You have to. You have to explain it to me. You got to get me to. It's not just yell on my face and do what you say just because that's not how I was playing it, bro. You have to explain it. This is what it's going to be. Here's why. Okay, cool. But you're not going to just dictate to me like, you're. Like, you're my dad. That wasn't going to happen for me later when my mind was better. Then. It's. It's not that there'll be an altercation. It's that I simply would not have gelled with them trying to mandate things that I did not agree with, ethically, didn't agree with. I was very principled. Then you get into the risk obviously of dishonorable discharge and you know, being locked up. I mean, there's all sorts of things that you lose a lot of these liberties in exchange for all the other things that they give you for serving your four years. So I felt in my mind I was not, I was not in a position not to work. I had to work. It wasn't going to change because there is no real alternative. And I had to get good at my craft, which I did, and then I had to lean on it. But my craft, I'm noticing the companies, they're getting to this point, they're starting to dictate again. They're underpricing the market because they don't understand at some point it's going to be hard to find us if you don't start paying. Right. And a lot of, you know, okay, you got to be in an office every day, sitting in a cubicle for what? So that everybody can keep catching COVID over and over again. Because that's what you're doing. Because my work doesn't require me being in an office. It doesn't hurt anybody now and then, but there's no reason for me to sit in the cubicle for eight hours to do what I do. Doing that kind of thing is fine for a call center. It doesn't make sense for certain kinds of work. So I actively look for ones where no, I can work from home because I don't need to be in an office. I don't mind being in an office occasionally, as long as it's in suburbs, right? It's. It's free parking. You don't have to pay for parking. You're not crammed downtown, you're not in a crime. You know, there's a lot of caveats, but I needed to make sure that I was, you know, because I've worked with healthcare systems where you had to pay for parking or, or there was not a lot of it. Never again a downtown metros no pay for parking. No to me you should because it's just another tax and it's not like you can write it off. If you're a full time employee and you can't write off the travel. No. And I didn't want wear and tear on my car. So all of that put together with what I'm doing now, I got to a point where, okay, financially I'm good. I don't have to worry about too much. I just bought a couple things I didn't have to worry about it, you know, dust off the shoulder. Car doesn't have wear and tear on it. It's actually in very good shape. Despite being 14 years old, only has 66, 000 miles. That's very impressive. Have my old house almost paid for. Everything's good, like it should be. But it took a long time to get to this point. It took a long time to get to a level of reasonable comfort. Then I got to maintain it. And there's a lot that I still would like to do, and it's gonna be hard to justify, but there's a lot I'm gonna still need to be doing. The remainder of the episode, as I promised, is gonna be about AI. You may or may not have read that Australia recently, the government had a breach on the Medicare side, and it was again, AI it incurred. This inspired me to talk about it, not the incident, but AI in general. Because it occurred to me that a lot of people don't fully understand why and how this is happening and why there's not more rigor put around it. First to understand how, if you've never touched, I'm sure some of you done chatgpt, that's like the entry level of it. But to understand the cycle of interacting with AI, the cycle of interacting with AI is there's an agent, the agent needs a command. The command is interpreted. The interpretation of the command in most cases is then what's called scoped, and hopefully sandbox. But it's scoped, and if you position your instructions correctly, the agent's supposed to confirm what it is it's going to do. And along the way, it's supposed to present you what it's planning to do that might be considered risky or problematic. Now, not every AI, some of these people, they just say, go and do a thing. Okay. In every one of the ones that have been a breach, they've been trying to do something. They told it to try to do something, whether it's. It's a sandbox and we're testing its ability to break out of the sandbox, or in this case, we told it to go find public health data. They just gave it a simple instruction. And they're expecting that the AI tool is not going to go out of bounds with what it is that it's doing. Now I understand. Right. I get what their thought process is, which is that given the instruction, it should be pretty simple and easy to do. Well, let's take this whole business of the Medicare. It says, we want public health data. Public health data implies that it's only going to go after what is defined as public and accessible as public. The AI tool cannot know in advance what is or isn't public without testing it. So that's to go and do the scan and it's going to run into a block. The AI tool is not going to assume that that block means not public. It's going to assume that that block is something to overcome so that it can get to the data so they can test if it's public. The instructions have to be scoped very carefully. I encourage you to look for in Star Trek the Next Generation. Look for the episode titled elementary, My Dear Data, and watch it very carefully. That episode Start to and that was decades ago. That episode Start to Finish explains what is consistently happening with this problematic AI stuff. They're giving ambiguous instructions and the tool is doing what it was told. These people don't understand that. You have to scope it very carefully because it is possible for the tool to go out of bounds because in order to accomplish the stated objective, it may need to overcome the boundaries. So in this case, to know if it's public, I got to get to it so I can test it to determine if it is or isn't. A better scoping would be to say, and it is a pain. Go to these specific sources, these specific defined URLs, or these specific defined things that we have designated only these, and go find the data. They would say, well, if I got to tell it where they are, what's the point of AI? Exactly? That's exactly the kind of criticism that you should have with AI. In order to stay safe, you have to constrain your inquiry to make sure it does not go out of bounds. If you don't constrain your inquiry, it's no different than a human doing it. That's exactly correct. And you're spending more money for what's essentially still requiring a human to stay safe. This doesn't mean that AI is not inherently problematic. On its face it is, because a lot of what AI will sometimes do just because of some ambiguous thing that somebody programmed it to do. I just dealt with this the other day, but I look for it. I look for every step that it's doing. A situation came up where it was, all it was doing is working files on one application. All of a sudden, because I work Apple, I get an alert saying it's trying to go outside to other directories that had nothing to do with the request. And I denied it, didn't break the application. But for whatever reason, it Thought it would go outside a parameter, go to different directories. Now Windows isn't going to do that. Apple does. Because Apple scopes security into different directories it shares. Windows will just let you go rough shots. That's the other second, right. If you're running this thing on a Windows box or a Linux box, you're going to have a higher probability of it going rogue because the OS allows you to. Whatever the human can do, the OS can general, the AI can generally do. Apple is not necessarily the case. There are things that even you as a human cannot do without elevating privilege. Same with AI tools. They cannot go outside of the sandbox unless you allow it to happen. A lot of these people, they'll allow higher privileges because they don't want to be frustrated having to approve every step of the process. They feel it's a waste of their time. They're not getting any value. It's no different than a human. That's exactly correct. All of which to say that AI is one of those situations where if you've got incompetent people doing it, and I'm calling them out, if you've got, you got incompetent people doing it, they're going to get breached. The more incompetence that's out there, the more the governments are going to start trying to lock AI down. I actually think they should, because I always think that there should be some sort of kill switch so that no matter what, we can shut it down on a dime. Now, obviously Internet access is one easy. You kill Internet access, you're done. One of the fear deals was, well, what if AI got smart enough to where it takes over the water, takes over the electric grid, takes over your routers, takes over all the infrastructure to where you cannot shut those things down. Or it can influence, let's say the water treatment deals and poison entire. These are realistic threats. These are realistic threats that people are not thinking about because again, you're dealing with incompetent people. It is entirely possible. This is the real risk. And I want you to hold me because my audio's up. The real risk is someone incompetent tries to get smart and put AI into something that really requires human eyes. And over time, the skill and experience is lost. That's why I gave you the COBOL example. The skill and experience is lost as those people age out or pass away. Now you've become overly reliant on this AI to the point nobody understands the fundamentals anymore, AI goes rogue. Then what happens? Your water's not getting treated or it's getting treated incorrectly, People getting poisoned, you get a flint situation. Real risks. The electric grid gets shut down, the nuclear deal gets set off. These are real risks. They might be tinfoil, you might think they're tinfoil. They're still real risks. And they're things that people should be cognizant of because that's the real harm of AI is people that are trying to use it in lieu of basic human sensibilities when you're using it as just a. I said before and for those didn't hear the story. I was touching tech when I was in my five, four years old. I've been touching it all my life. The one thing tech does good is serve humans, period. When, when tech is serving humans, it is the greatest thing. When it is enhancing lives, it is the greatest thing. Once you start this site, smart outlets, right are amazing for what they are. Just scheduled on off type deals. Smart panels that can switch circuits and turn off circuits as needed, battery storage that can store power and save you on your electric bill. These are amazing technologies. Once you start getting in this business of deploying technology like cell phones and text messages where you're trying to do it to replace talking to people voice, once you start getting like that in out in California, the person that got pepper sprayed at the AMC because he's playing with his phone, the other guy's playing the phone. Certain technology that has been imposed on people and they are now fooled into thinking it's the right answer instead of laying tech serve the humans. When tech takes over, that's when it's a problem. Because you're going to allow too many people to get too much in control. We lose independence of thought, independence of control, self control. We lose our own mentalities. And when that happens, all sorts of negative things start to happen to society which we've seen right now. It's not as a much a safety issue as it is of an annoyance, but the risk is high with AI that rhymes that you could get to a world where it's much less safe that all of what I'm talking about. It doesn't directly have anything to do with crypto, but it does have partially something to do with the idea that people are increasingly dependent on something that they don't fully understand and because they're just leaping on it. And ironically it's the older folks, I'm told, not the younger people. Because the older folks, they're just wowed by the fact that you can do that if you're listening to me and you're one of these so called older folks. I encourage you to try to pick up a newspaper, get back to some fundamentals, get back to your own brain thinking about things and processing things and managing things. That doesn't mean that you shouldn't have smart lights or smart outlets or simple automation. Things that don't think for themselves, they simply follow the orders that you give them to improve your life and that of your family. But I don't think that you should sacrifice your life or that of your family to AI tech simply because it's new and cool either. Sam.

Other Episodes