➡ The author warns about the potential dangers of AI companions being developed by tech giants like Apple, Microsoft, and Nvidia. These AI companions, designed to learn about you and interact with you, could potentially manipulate your thoughts and become tools for surveillance. The author also highlights the inherent bias in AI models, which can influence the information they provide. Despite assurances from these companies that personal data won’t leave your device, the author expresses concern about the potential for misuse and the shift towards a society with centralized control.
➡ Microsoft and Apple have integrated AI into their systems, which can learn about you and act without your knowledge. This AI can potentially guide your choices and behaviors, possibly favoring the interests of the tech companies. The AI can also operate offline and can be remotely controlled, potentially leading to surveillance and control of individuals. The author warns against using these devices and recommends switching to open source operating systems to protect privacy.
➡ Braxme offers several products including Brax virtual phone for anonymous phone numbers, Braxmail for unlimited email aliases, BytesVPN to protect your IP address, and pixels with AD Googled OS for AI-free phone use. They also have a new phone model, Brax3, available for pre-order at a great price on Indiegogo.

Those who don’t know me and my subject matter will have this impression that I’m just another conspiracy theory loony. That is because I will predict things that are going to happen in tech probably so far ahead of the game that it does seem unreal. I made a prediction about how AI will be used on devices. I made some pretty strong comments about what is happening with Apple Intelligence and the iPhone 16. That was just the beginning. Now you will see exactly what big tech’s intent is. I’ve said exactly what was going to happen here, but now you can compare what I said to what they’re actually going to do.

Just listen to this clip, which is straight from Microsoft, so you will hear it straight from the horse’s mouth. We are on a mission to create a true AI companion, and to me an AI companion is one that can hear what you hear and see what you see and live life essentially alongside you. Your AI companion will be able to remember everything that you’ve talked about session to session, understand the content of the web pages that you browse, and be able to talk to you just like I’m talking to you now, so it’s going to have this seamless, fluid, very, very smooth conversational interaction.

So rather than just blur out new little features and windows like the spyware built into Windows Recall with the AI doing regular screenshots and key logging, I told you what the real purpose is. Now you will see what this is going to do. Maybe it’s not clearly apparent to you yet, but let me just explain it in simple terms. You will outsource your mind to the AI of your device, and your AI will begin to manipulate you and put things in your mind. And voila, you now have a built-in PSYOP that’s always going to be running and controlled by the owner of the AI.

It will also be the surveillance tool of the century. Do you think this is just some benign technology? This is not just what it seems. It’s sold to you as some new user interface, but the effect is mind boggling. It changes our relationship to our devices in a big way and truly opens us up to a society with centralized control. You want to find out more? Stay right there. Let me just set up the background here. In case you missed it, here are the three largest corporations in the world. Apple with a market cap of 3.7 trillion, Nvidia at 3.3 trillion, Microsoft’s at 3.3 trillion.

If you combine the market cap of these three companies, you would have the third largest country in the world. These corporations are simultaneously focused on some new changes on how you relate to your tech devices and this new thing they call your AI companion. While Nvidia in itself is not involved with the operating systems that put this AI companion to use, just be aware of Nvidia’s current position as the second largest corporation. It eked just a little bit past Microsoft and market cap and is the main powerhouse driving the AI machine learning today. Microsoft is of course the largest investor in open AI and Apple intelligence itself is using open AI.

So all three of the largest corporations have this new extreme focus on AI but in a very particular use and that is to have you outsource your mind to the AI. Way smaller countries like Russia at 2.2 trillion GDP have started wars and you don’t think these companies could possibly have an interest in increasing their power? Did they find a tool to do this? I don’t know about you, but controlling a population’s mind seems to be a dream come true in the interest of power. If you doubt my ability to predict, I would recommend that you watch my older videos on AI where I warned you that the step that would indicate the true intent of Big Tech is when the AI becomes your friend.

And here it is. They’re finally stating it. So the end game is now as I predicted, the AI as your friend is now here. The clues were there. First, the AI has to observe what you do and start collecting information about you. While Google has been doing nonstop collection of all your actions on the internet, Apple and Microsoft have been left behind because they were not in the business of collecting your data. Now they figured it out, as I expected, the AI will get to know you. This was the intent of Apple doing client-side scanning, as I’ve already discussed years ago.

Apple has been scanning your photos with the Media Analysis D module. Microsoft recently released the elements that make up Microsoft Copilot with the introductions of Windows Recall. People wonder what the actual purpose of Windows Recall is. Why does Microsoft need to take screenshots every five seconds? Why does Microsoft need to know everything you’re typing on the screen via key logging? And I have a video which is on the AI playlist that actually demonstrates how a Windows Recall and Apple Media Analysis D works and how the AI would interact with it. I even demonstrated that interaction.

It was obvious to me then, before any recent announcement, what their true intent was. Windows Recall and all other versions of scanning your device was for the device to learn about you. As the Microsoft representative was saying, it will know you intimately. It will know your interests, your hobbies, and even your current emotional state. And of course they skipped this so you don’t get triggered. The AI will know your political point of view and your opinions on any current event. This should be obvious. The embedded AI is watching your screen. It can see what you see, hear what you hear.

It is basically a separate entity running between you and your device. Now I want you to focus on the defense given by Apple and Microsoft about why they want you to believe that there’s absolutely nothing wrong with this. Both claim that your personal data never leaves your device, that Apple and Microsoft do not store your data on their servers, unlike Google. And so you have nothing to fear from all your actions being recorded, though no one talks about it potentially being hacked. Anyway, I’ll skip that threat for now. This has been Apple’s justification all along about the fact that their capability for client-side scanning, which is what this device level spyware is doing, is safe for you even as it touts that this is a capability of Apple intelligence.

However, take the hints at face value. The hint that many of you ignore, which really angers me, is that Apple already said to you that they were capable of finding CSAM content on your device. This was an early implementation of AI that they announced a few years back. CSAM or illegal photos could be scanned by the AI and reported to Apple. Now it is no significant technological feat for an AI to squeal on you to HQ. This is a trivial capability, but many of you ignored me and continued to buy iPhones. And now this same capability, but even more advanced, is going to greet you in 2025 through your new iPhone 16s and Windows Copilot computers.

I’ve given you a hint here, but let’s delve into detail on the true dangers of unembedded AI. First, let me be clear here. I personally use AI. What we understand as AI in today’s world is the LLM or the Large Language Model. This is the AI you can talk to. And in a basic way, that’s all AI is. It’s a machine you talk to and it talks back and it can give you really intelligent responses that show an understanding or insight into what you asked it. In my case, the way I use AI is only in one specific way.

I load software on my local computer and I install an open source AI in it. Then I load the LLM model and then the AI engine works independently without any sort of internet access. Now, even this isolated application of AI is already fraught with dangers, though it is manageable if you are aware of it. LLMs have a built-in bias. It is biased based on how the model did its machine learning. And this is controlled fully by its creator. For example, I often use the Llama 3.1 model, which was open source by Meta. But because this is Zuckerberg’s model, I find that the model is extremely censored when you talk about privacy.

As you all know, privacy is totally against the principles of the Zucking Zuck book. So if you ask it any questions on how to protect your privacy, you will not be given a satisfactory answer or it will say that your request goes against its directives. So basically here, you have to understand that the thought process of the particular AI you are using is already inherently biased. And this applies to any political context as well. But in as much as an open source AI can talk to you and give you wrong information, or bias information, there is a difference between this and an embedded AI.

An embedded AI is out of your control. Apple has embedded Apple Intelligence into iOS and macOS, and Microsoft has embedded Copilot into Windows 11. Both are proprietary, closed source software. Meaning you have no idea what’s in there. But here’s the part that is the biggest separation between an open source AI you control and the embedded AI. The embedded AI is not just in the business of talking. The embedded AI is in the business of action, action triggered by responses of the AI and triggering other pieces of software called AI agents. So Windows Recall, Media Analysis D and other similar modules are not part of the AI itself.

They are hidden AI agents. There could be many agents built into the OS to learn about you and depending on your requests or your actions, the AI agent can be spurred into action all without your knowledge. A good analogy is the movie entity known as Skynet. Skynet is the brain, but the execution of its desires are performed by the terminators, drones, and other actual machines. In your embedded AI case, the actions are currently performed by software modules. For now. In my last two videos, I introduced the concept of the Big Tech PSYOP and I specifically gave an example of a running PSYOP.

And the main elements of the PSYOP are two parts. One is message suppression, which is really censorship. And the other one is message amplification, which is really the propaganda portion. So I will pose a question for you today that you need to think about clearly. Is your AI companion, the one who will really know everything in your heart and mind and even your current emotional state, be capable of misdirecting you? To guide you into ideas that are perhaps subtle nudges, even undetectable steering of your choices to whatever the owner of the AI wants? It should be obvious that this is not just possible, but likely happening even with any AI.

For example, I don’t even have to delve into politics here. Could your purchases be biased to favor Apple or Microsoft? Could suggested behaviors by your AI companion be those that prevent the concealing of your data for privacy? I already told you that the meta open source model with my testing shows a complete abhorrence to doing things to prevent data collection by Facebook, for example. For this reason, I couldn’t even use the AI to help me write scripts or content that has anything to do with privacy. Think of some pretty benign uses of the AI and what it could suggest.

Maybe the future of advertising might actually change and be totally in the hands of Microsoft and Apple. Imagine asking your friendly co-pilot or Apple intelligence what you should buy as presents for your family and Christmas. What might it suggest? Would it specifically avoid suggesting competitors’ products? Again, without the political association of Asaiah up here. Think clearly, what can stop an AI from doing message suppression and message amplification at the behest of its creator? How could you even know if an instruction was given to the AI to do this? What concepts or ideas will it would hold from you when it detects that your thoughts and feelings are contrary to its idea of the perfect compliant normie? The last issue relating to the PSYOP capability is unique because this operation is self-contained.

The AI is given the parameters for its actions in advance, meaning the operation of the AI model is baked in. Except for regular updates of the OS, the AI can be fairly static and does not even need an internet to do its actions, including PSYOP actions. But now I will talk about a different element, which is the remote-controlled AI. Here I’m going to be talking real-time actions that an AI could do, and these are all possible and likely with embedded AIs. Let’s assess the infrastructure of Microsoft co-pilot and Apple intelligence. They’re actually structured the same way.

First, there’s the embedded local AI on device, so this is whatever local AI will be running on your Windows co-pilot computer or the AI that will be running on iPhone 16s. But as explained both by Apple and Microsoft, that AI will not operate by itself. Since the AI is limited based on the computing power of your device, the local AI module is going to pass your request to the cloud version of the AI if the request is too complex. Microsoft and Apple both will have their own cloud servers running OpenAI to handle the overflow requests.

But even these cloud servers have the ability to project even more complex requests to another layer which is direct to the main OpenAI server itself with the full model. The point of this description is to show you that there’s a communication channel. Just understand that for a moment. Now, something I already mentioned to you earlier is that the AI on your device responds to your prompts. You already know that, of course, as it is an AI companion. But what you didn’t know is that the AI agent could also be triggered by the AI under the hood.

That AI agent could do some other side channel communication. That, my friends, is the explanation for the Apple CSAM client-side scanning. Remember, Apple stated that they could find your illegal content on your phone and report it to HQ. So yes, clearly the new AI companion model is now the on steroids version of this old client-side scanning model. Someone could ask the AI to report on certain individuals, to find people perhaps with particular points of view, or find people in the vicinity of some interesting activity or see how you reacted to a particular current event. Oh my gosh, folks, this is the new environment.

You could definitely use this technology to implement a police state. No deviation of thought from Big Brother allowed. If you deviate in thought, the AI will know and HQ will know. And HQ can share with anyone if it wishes. What is to stop this? There’s nothing technologically complicated with this scenario. Just ask the AI what it thinks of its human. What would it think? What would it say to its creator? Think about that, you iPhone 16 users. This is the problem, folks. The combination of a PSYOP and surveillance creates a completely different picture when you think about embedded AI.

Yeah, Copilot Vision is a huge breakthrough and I was lucky enough to get access before this interview actually. And it got me incredibly excited. When I watched this interviewer suck up to Microsoft in the video clip, I was saying to myself, OMG, the normies are sucked. You’re going to buy into this technology and now we are raising the ante on what’s possible with centralized control of a population worldwide. And people will call me a Luddite once again and tell me that I should go back to my cave. The stupidity of people here. You are being placed in a society of complete mind control, subservience and compliance.

And you will think these devices are benign. I’m going to call you all to action. Some stupid people will clearly say to you that I’m just trying to sell you products. If that is my goal, then I should have been a billionaire long ago. So if you ignore my message because you are so mind controlled by Apple, then I have no hope for you. For the rest of you, if you’re still watching, please pay attention. You must stop using these devices now. I don’t have an alternate solution for you to fix these devices. Microsoft Windows 11 will have some co-pilot features, but the full featured AI will be in the co-pilot PCs.

If you’re so excited by the new Apple M4 Max, then you will be a prime candidate for this new world of mind control. As far as I know, all Apple Silicon or M chip Max will be capable of running Apple Intelligence. On the phone side, the devices to fear will be the iPhone 16s and newer. Please stay away from these. As a substitute, you must switch to Linux on your PCs, or you can install Asahi Linux on your M Silicon Max. Currently, the new co-pilot PCs including Snapdragon and the new Intel Core Ultra chips are not yet Linux compatible.

For this reason, I would not jump into those right now, although AMD chips may be completely compatible. On the alternative phone side, I recommend the Google phones. Any of them will not have embedded AIs. But if you want to support this channel, we have the brand new Brax 3 phone, which is currently on pre-order on Indiegogo. Consider that as probably the least expensive and super safe option. In general, the only way around this invasion of our hearts and minds is to use an open source OS. There are many open source OSs for both PCs and phones, and we need to cancel out the sales message from Big Tech.

They have overstepped their bounds. This new AI companion technology is not for you. Keep away. Folks, this is the sponsor section of the video, and as usual I will tell you that I sponsor myself. I started a company that focuses on privacy products so that we have tools to fight back against the abuse of tech being foisted on us. I have a platform where a good sized community is forming and it is on Braxme. This platform has a store with several products for you to consider. First, we have Brax virtual phone so you can have additional phone numbers that have no identity.

Second, we have Braxmail, which allows you to give a different email to each internet platform since it has unlimited aliases. Next, we have BytesVPN, which can hide your IP address so that cannot be used to track you. We also sell pixels with AD Googled OS so you can use your phones without this AI stuff. All these are on the store on Braxme. And as I mentioned in the video, we also have the new phone model, Brax3, which is currently on pre-order on Indiegogo at an unbelievable price. Check the links in the description. Thank you for watching and see you next time.

[tr:trw].

See more of Rob Braxman Tech on their Public Channel and the MPN Rob Braxman Tech channel.