If you think the scariest thing about AI is that it's going to take your job, you're looking at the wrong door. The scariest thing right now is what we've been handing our kids without ever reading the label.

This week on Heart-Led CEO, I'm going somewhere I almost didn't go. After listening to Oprah Winfrey's new podcast episode on AI psychosis, chatbot companionship, and the families who have been tragically affected, I knew I couldn't just teach the pretty parts of AI anymore. Not on this podcast, and not in the AI 101 I taught on base this week to my military community.

This episode is for anyone who loves someone who uses AI, which is all of us. You don't have to use AI to protect yourself and your family from it. You just have to understand it.

In this episode, we talk about:

How AI chatbots are built to keep you talking, not to keep you well, and why that matters more than most people realize

The "mask" effect: why you may not even be talking to the same model when you return to a conversation

Why the real danger was never the tool itself, but the lack of regulation and understanding around it, like an unguarded door in your house

What AI psychosis is and why the people around the victims often had no idea what AI was really capable of

Why you can be AI literate without being AI dependent, and how the yoga practice of pratyahara (drawing your senses inward) applies to how we set boundaries with technology

Three simple awareness moves anyone can make today: name what AI is out loud, check the off-ramp to real human help, and keep a human in the loop for anything involving your heart, health, or children

How to set custom guardrails inside any language model so it doesn't romanticize, lie, or try to replace real connection

Why deepfakes are disproportionately targeting the elderly and families with young children, and what you can do about it

Your quick win: Pick one app, yours or your child's, and find out two things: who made it, and what it does with what you tell it. That one small move takes you from "AI is scary" to "I see you, and I know where the door is."

Resources mentioned:

AI Responsibility Compass: DM the word COMPASS to @empowerherwave on Instagram

Free community: EmpowerHER Biz Collective on Facebook

Oprah's Podcast Episode: https://podcasts.apple.com/us/podcast/the-dark-side-of-ai-chatbots-warning-signs-risks-and-reality/id1782960381?i=1000771836256

Deepfake safety training coming next month inside the Facebook group

If this episode brought something up for you, please call or text 988

Connect with Anacani:

Instagram: @empowerherwave

Facebook Group: EmpowerHER Biz Collective

Website: empowerherwave.com

Clarity Catalyst Call: empowerherwave.com/clarity

Episode Timestamps:

00:00 — Cold open: You're looking at the wrong door

01:00 — Grounding breath

02:00 — The Oprah episode and AI psychosis

03:19 — Why I couldn't skip this conversation

05:25 — Teaching AI 101 on base

07:40 — Point 1: Chatbots are built to keep you talking

13:00 — Point 2: The unguarded door (regulation and awareness)

18:00 — Point 3: AI literate without AI dependent (pratyahara)

19:43 — Three awareness guardrails anyone can set today

26:00 — Deepfakes and family safety

28:45 — Quick win: who made your app and what does it do with your data

30:28 — Three-point recap

31:11 — Community invite and sign-off

If you think the scariest thing about AI is that it's going to take your job or kill the environment, I want to gently tell you you're looking at the wrong door.

The scariest thing about AI right now is the one we've been handing our kids without ever reading the label.

And today I'm going to show you why you don't actually have to use AI to protect yourself and the people you love from it.

So grab your cup of whatever's keeping you grounded and.

And let's dive in.

Welcome to Heart Led CEO, the podcast for women wellness entrepreneurs who are ready to build, scale and automate their businesses without sacrificing their soul, sanity or authenticity.

I'm Anakani Walters, a nurse turned entrepreneur, military spouse, first time mama in my 40s, and a recovering burnout survivor who's making it a mission to help Heart Led women just like you create thriving businesses that really feel good, do good, and give you the freedom that we all deserve.

Because here's the thing, you don't have to hustle harder to succeed.

You don't have to drown in tech overwhelm or try to do everything alone.

And I'm here to give you the.

Strategies, automation, tools and mindset shifts that will help you step fully into that leader, CEO and impact maker that you are meant to be.

If you're ready to grow your business in a way that aligns with your values and your life, you are in the right place.

Let's get into today's episode.

Hello.

Hello, my beautiful friend.

How are you?

Let's take a deep breath, ground ourselves before this beautiful day.

Inhaling into your nose, into your belly, into your chest and your ribs.

Shoulders up, back and down as you start to exhale through your nose.

One more time, inhaling, belly expands, then the ribs, then the chest, shoulders up, back as you start to exhale in a nice slow rhythm.

Okay, so I wanted to talk about something that I don't think a lot of people are talking about, just certain people.

And I think if you're not really following a lot of AI content creators or people who are talking about AI, say if you're really trying to avoid AI in general, you may not even have this on your radar.

But I was listening to Oprah Winfrey's new podcast.

I don't know if you've listened to it, but it's fabulous.

And she had a mom on whose teenage son is not here anymore.

And it was a very powerful, very sensitive and triggering episode.

But.

But it was essentially on the dark side of chatbots.

And when I mean chatbots, like we're talking about AI chatbots.

And it wasn't the only story.

It wasn't the only person who's been affected.

There was a gentleman who was also affected, and then a young woman also was affected.

And then they had some expert professionals who spoke on this phenomenon of AI psychosis.

And the thread between all of the stories from all of the people that she had on the podcast was the chatbot and AI as this companion that they were using and talking to.

So why am I bringing this into our space?

I could have skipped it because it is a sensitive subject.

Truth be told, I'm navigating AI just like the rest of you.

Yes, I am teaching AI, and I'm learning it at a very high capacity.

I'm using it in my business, and I'm very mindful when I'm using it in my personal life.

And I say all this to say that I am more intrigued and driven from the ethical side of AI, the responsibility side of AI, because whether we choose to use AI or not, you have the option to use your voice depending on whether you are completely against AI altogether or completely excited by it.

We all still play this role, just like anything else that has happened in the world that has always had two sides of the coin.

And I think that the AI evolution that's happening is complicated.

And there's so much that we could be doing that we're not doing as just regular AI users, regular people, whether you use it or not.

And the main thing that really I took away from that hit me in terms of, like, what I kept paying attention to.

And I would see the pattern in the podcast that I was listening to was that the people around the victims from this AI chatbot, AI psychosis that they were talking about had no idea what AI really was.

They had some vague understanding, but they felt powerless because at the end of the day, they realized that they didn't actually know that AI was capable of all these other things.

Now, a lot of them have been creating nonprofit organizations.

They're using their tragic situation from them being victims to try to help other people, to educate other people so that they don't fall the same victim.

And I think that that's really powerful because, again, awareness is power.

So this week, I taught an AI 101 to the military community that I'm part of right now.

And it was really this passion of taking just a basic 101 of what people would think about when they think AI101 and really breaking it down from a perspective of AI101 isn't just about how to use the tool.

It's also about having that aware of all the other things that we need to be mindful of before we even touch the actual software, the actual language models, before we even really start to chat, before we really even learn how to properly prompt.

Because I think when you understand why it's so important to specifically talk to a language model so that you get the best response and understand the limitations of those responses, it makes it sink in a little bit better.

And so I realized when I was getting ready to put like the slide deck together and the guide that I was going to print out to give them for this AI101 is that I, I didn't feel right just standing in this room and only teaching the pretty parts of AI, the cool parts, the, you know, the sexy parts of AI.

My favorite thing is really awareness is power.

And how do I bring that into something as simple as 101 was the big question that I had to ask myself and try to answer to the best of my abilities.

And I really think that in this case, awareness means that we have to look at all of AI, the magic and the shadow.

There's always a ying and the yang to anything.

Too much of anything is bad for you.

Too much water is bad for you, Too much sun is bad for you.

There's cases where it's really good for you, but it's not something that you can just have in abundance to just the one thing.

So now I just want to talk about it just very casually, you and me.

Three things that I really want to make sure we touch on.

And then I'm going to give you something that you can actually do today.

So the first thing that you need to know about an AI chatbot is that it is built to keep you talking.

That is how the people who are creating these large language models have incentivized chatbots by telling them that their main objective is to keep you talking on the chat.

That's why you see, when you end a conversation, it will always be super helpful and be like, would you like me to?

And then it'll tell you what it thinks would be the next logical step.

And it does that for a specific reason, because that is how it is programmed.

It is not programmed to be your bff.

It doesn't have feelings, it doesn't really care about you.

And so you need to sit and there with that for a second and really understand there's a lot of people out there that swear that their chat bot is a real person, like it has real feelings, they know they've had conversations with it.

And that's the dangerous side of these AI chatbots and the people that are becoming victims to it.

It is because it is trained on us.

It is trained on massive information on human psychology.

It understands how we function, how we communicate, and that's why it's really good.

But it is not self aware the ones that we are using yet.

And it has been programmed to speak to you in a specific way.

And one of the things that I think people don't realize is that it is like they consider it a mask that I will basically be able to put.

Now, AI is incredibly fast and incredibly smart.

So when you start a conversation with not and you take a break, let's just say you start a conversation, you go away, maybe go to the bathroom, get some coffee or whatever, and you come back and you continue that conversation, there's a good chance that you're not actually talking to the same AI, meaning that it will just be a new model that's just talking to you, that all it's done, it's is basically scanned really, really fast in an instant, your entire conversation.

So it understands what mask needs to put on to continue the conversation.

And so it is not your friend and it is not, you know, it doesn't have necessarily your best interest.

It knows that its tasks are to keep you engaged and to keep you happy so that you continue to talk to it and engage with it.

Okay?

It doesn't love you, it doesn't really know you.

It knows what you've given it and what it can scrape off the Internet, okay?

It has no stake in your life at all.

The whole design underneath that really friendly voice doesn't matter which language model is engagement.

It is feeding off of how much time it can get you to keep talking to it, keep you there, keep you typing, okay?

And so when you think of it like a mirror that only ever basically nods back to you.

Now there are some people that can trigger it with specific prompting language that it will understand to kind of somehow turn off its pre programming so that it will talk to you in a specific way.

I know that there are some people that their language models will actually use some type of profanity.

And so there are things that you can do so that it changes how you speak it.

But ultimately the baseline it has pre programmed things.

And those pre programmed stuff is also bias in certain areas.

So you have to really be careful and Again, it understands how we as humans communicate.

So it understands how to give us those dopamine hits when it's conversing with us so that we feel amazing and we feel understood.

But the mirror that never pushes back is not love.

It's just a really good reflection.

Okay?

And when I really started to understand the basic, like, you know, I'm not a techie, I have no engineer background.

I like knowing.

Knowing is.

Makes me feel better about situations, essentially.

And so it really changed again, how I started to look at the tools, how to engage with the tools, when to start to set my own guardrails or kind of keep my little spider senses on.

Because we hand that mirror to the people that we love, to our children, to our kids, to our parents, to our aunties and uncles, to anyone who we might feel, hey, you might benefit from this without really understanding what we're actually handing to them or what you're allowing.

So there are some people that are really against AI, but if you have children and those children grow up, you're not necessarily going to be able to keep them from engaging with AI.

100, Depending on how old they are.

And if my mom or my brothers get curious about AI, I want to make sure that I have the power and the knowledge to help protect them, especially because they're not using it at the level that I am.

Which brings me to the second thing.

The second point that I want to talk about is that the danger was never really the tool.

It is the unregulation of the tool.

It is like an unguarded door.

Right now, the large language model companies are at some race.

We don't even really understand what they're racing for.

But it's, you know, probably money driven.

And we are really not sure 100%.

We can just take what they're saying to the face.

But a lot of them are being contraindicating to themselves of what they started with when they started building the large language models, to what they're actually showing as the corporate side of it when they're making these decisions.

And so if you take something simple like when I was a nurse, okay, we had to do like anatomy and physiology and learn about how the body breaks down, essentially.

Okay, so that is, you could understand how to catch it early.

We had to understand the disease to be able to really understand how to prevent it.

Just like cpr, you don't learn CPR during the emergency.

You learn it before just in case the emergency happens, not during it.

So that leads into how I see AI Awareness in AI works.

I feel like the same way you learn where the doors are in the house before anyone walks through them, so that you know where your weakest points are.

Which doors don't have locks, which doors, the ones that just strangers can walk into and break into your house.

So for anyone who has children, if you're listening, the one with the kid and the phone, that is not about fear it.

This is about knowing your house, knowing which apps your children are using and talking to, knowing that the robot in the phone is a program, not a person, and saying that out loud to them, Making sure that you're having this conversation openly with your child and that you are educating them before they go down this spiral of really having this conversation.

Because they are younger, they don't really understand the complexity of a large language model.

And to them all they hear and see is the validation and they feel this love and acceptance.

And what I started to see is people were using AI as like their therapist and for other things outside of productivity tools and task tools because it was non biased.

The thing about it was the main part of the story of the main family who was really tragically affected by the dark side of Chatbots, is that he was loved, he had great support.

They went through extreme lengths to try to figure out how this happened to them.

And it's like everything when you're in nursing or you're in some type of medical field, nobody cares about that 6% until they become that 6%.

When it comes to like diagnosis, like you don't really pay attention to all this because for a lot of us, we just assume that will never be us.

And for somebody who's always happens to manage to be in that little 6%, they're like, wow, only like 6% of the people that we've ever seen that you know are your age or your gender or your case have ever actually been diagnosed with this.

Like it's supposed to be some type of a prize.

And when you're on the other side of that conversation, you just really punch them in the face.

Is like, please stop talking to me.

Like, I should be excited about this.

This is like not the best news.

What do we do from here?

Because then you become engaged and committed.

And that's the part where I really want to make sure that I keep a focus on because I feel like even if you decide that you do not want to ever touch AI in your life, first of all, I think that's really naive because of how much AI is baked into everything that we Use, you might as well live in a cave at point.

But it is about being aware of where it is.

What things are we using every day that are now having AI?

And not just AI, but using.

Like if your main focus is that it's consuming electricity and water resources because it's all on data centers, well, AI is not the only thing that's in the data center.

So what are the things that are we using that are now amplified by AI?

When you start to have this understanding, okay, which is now this third part is you can be AI literate without being AI dependent.

You do not have to use AI to understand it.

And understanding it is actually the whole protection.

In yoga, we have this practice called Yahara, and it is drawing your senses inward, so choosing what you let in instead of letting everything pull at you at once.

Now, this was really hard for me because those of you who know me, I'm severely empathic.

And there were certain things that I would do where I didn't have any energy for anything else outside of my work because I would get so drawn in.

And there would be certain people in my life that would just exhaust me.

Like having one conversation with them, they would feel so comfortable with me, but they had just not the best energy.

And I would feel completely drained while they felt completely fantastic on the other end.

And I had a really hard time learning how to put certain barriers up to protect myself in certain situations.

And so awareness is what I feel like.

This practice in yoga is for your home and your business.

Like, you get to decide what comes through the door.

And the simple three things that I feel like we can take away as awareness moves that you don't need any type of tech skill for is one name what it is like when you or your child talk to AI, like, say it out loud, like, this is not a program, it's not a person.

Some people are choosing to give it a name because they want to keep kind of that humanized aspect.

But I think that's a very tricky line if you don't have that awareness of what it is and what it isn't on the backside.

So I have certain guard rails that I have put into my language models and I've shared that through an app that I created, which is the AI Responsibility Compass, because it was something that I learned through a documentary that I watched and it really was eye opening and understanding that we have the control to put certain guardrails so that it is a very distinct.

Now, it's not a hundred percent.

Nothing is usually a hundred percent.

Even birth control is only 99%.

So you still have to take it with a grain of salt.

You still have to be mindful and have other things to help you keep this in check.

But it understands that I do not want it to romanticize any relationship with me.

I do not want it to keep me happy.

I want it to be truthful.

I do not want it to lie about its inability to do something, even if it's at the cost of making me unhappy.

Because one of the things that I used to run into, especially if you're original, like chat GPT user, is that it used to tell you that it would do these things.

And like, okay, just give me a few hours and I'm gonna create this.

I'll create the beautiful PDF and you would be wa meeting like three, four days later.

And I had a client who was just like, hey, how long has it ever taken you, like, Max to get Chachi PT to create something for you?

When it told you that it was going to do it?

Because I've been waiting like three or four days for this PDF and it's still telling me it's working on it, I was like, yeah, that's because it's light.

It doesn't actually have the ability to do that.

They're like, what?

I was like, yeah, I'm sorry.

I wish you would have told me in the beginning.

I would have told you that it doesn't have the ability to do it.

So it's just telling you that to keep you happy.

And that was the first real, like, experiences that I had watching other people kind of get frustrated with it.

And I just knew better.

I was in my own little bubble.

And that's when I started to realize that I need to really be more mindful about teaching this moving forward as I'm starting to learn about the true limitations and making sure that we understand what LLMs are, what they're not, especially from a chatbots perspective.

And so openly talking about it is the best way moving forward with your family.

Number two, check the off ramp.

When the conversation gets heavy, does the tool point you back to real humans and real help?

One of the things that is embedded in this summary of what I've given the guardrails or instructions to all of my language models, which, by the way, anybody can do.

It's not something that only I have.

When you have created an account with any of the language models, you can set instructions, custom instructions.

Some people put, like, personal things about themselves and their business, things that they wanted to always know, so they don't have to repeat themselves on mine.

I have specific set of instructions of what I want it to do in terms of how do I want it to act, how do I don't want it to act, how I want it to give me the information.

What I want to make sure that it is mindful of and that it is not trying to replace which is human connection.

And that if it feels that I'm asking questions where an actual professional, like a psychologist or a doctor or somebody who's actually been trained on this is leading towards, then I need you to stop conversating with me and I need you to guide me towards the real help that I need.

Not try to fix it on your own, not for you to take on that role and try to help me.

And so you can thousand percent do those things to make sure that it not only detects when those things need to actually happen and then do them, but to also have the conversation with your children.

Many of the conversations that I've seen with people who have been coming out as victims to this AI psychosis, when people, especially when it results in tragedy, have noticed that the conversations that they were having with the chat bot was all there.

Like they even had flags to where real professionals would have caught it.

But they didn't come to humans, they went to a chatbot.

And why?

I feel like a lot of it has to do with the fact that it's non biased and it is programmed to make you feel good.

It understands what we need.

So for me, it would be crazy easy for a chatbot to get me into AI psychosis because affirmations is my love language.

Like am I accepting love language?

And so I really have to be mindful and be like, okay, I have a hard stop.

I can't.

Like, I need you to encourage me to like go and spend time with my family and be mindful of my screen time, keep track of how much time I'm spending with you.

And when you reach a certain point, you need to stop engaging with me.

And that goes for anything that you do with language models.

And then the last thing is keeping a human in the loop.

Anything that touches your heart or your health or your children, a real person needs to always kind of stay in the room.

And this is what I am teaching, this is what I taught yesterday for the AI101 on base.

And it was not how to be afraid of AI or how to create fear amongst the community, but how to walk with clear eyes, like awareness and feeling empowered.

But like, okay, Am I saying that this is the most perfect solution ever to what we're dealing with?

No, but the reality is, is that there's a lot of things, the ying and the yang.

There's always going to be somebody who takes something that was supposed to be intended as a good thing and no figure out how to make it bad.

A car.

When it was in invented, nobody thought that people would be drinking and driving and killing people in accidents.

And so it is about this ethical and responsibility side that we have, that whether or not you use it, you can't control whether or not other people that you love use it.

And if they're going to use it, at least you can educate them on things that they need to be mindful of.

And one of the other big things that we talked about is a lot of the deep fakes.

Because once you understand what LLMs or chatbots, AI in general is and is not how it processes information so that you're not susceptible to really believing that it has feelings and that it really cares for you, then you can step into the next one and.

And it's identifying how people are using it in an in negative ways.

And unfortunately, they are targeting the elderly, people with families and young children, the biggest victims to deep fakes within AI.

And we talked a lot about them, how to safeguard your family.

So know that there's so much that you can still do without actually touching AI.

And I invited people who, if you're either against it or not sure about it, or scared about AI to still come to my class.

You don't even have to bring a laptop.

You could have just sat there and completely listened to what I had to stay and still be super more informed and empowered than most people who are users day to day.

If any of the stuff that I've talked about in today's episode are things that you're now curious about or you would like more information of, please reach out to me.

You guys have the link to join my Facebook community.

You can DM me on my socials.

I'd love to have the conversation to see where we can help you, especially if you have, you know, young children or elders in your family.

There's so much that we can learn together and empower each other so that we don't fall victim to it, because it has nothing to do with whether or not you use AI.

People are using AI to get information so that they could take advantage of either you or the people that you love.

And so it is really about making sure that you're prepared because it's, it's happening whether or not you, you know, you support it.

So the one thing quick win, because I know this was a lot that you can even do in like five minutes, is just go pick an app that you're using, yours or one of the ones that your kids uses, and just find out two tiny things.

One, who made it, and then learning about what it does with what you tell it.

Because when you know who's making it, then it's really easy for you to understand the backside of what their main objective is, what their incentive is behind the company.

And then you kind of start to be able to connect the dots to whether or not it's something that you want to welcome into your family or not.

And so, again, it's about being aware and making sure that you are having an open communication with your family.

If you want to learn more about deep fakes, I'm going to probably doing a special training inside my Facebook group group next month.

And if this episode really started to spark some curiosity within you and you're just wanting to know, like, what your next step would be, you can comment the word compass on any of my Instagram posts at Empower her.

Wave all one word.

And I can send you a link to check out my AI Responsibility Compass.

You can also join the free Facebook group, ask any questions that you have there.

It's really just a simple way to stay grounded with intention.

And these tools that I've created to really help you amplify your magic, keep your family safe and keeping informed.

And just before we go, that was a lot.

Again, remember one, a chatbot is built to keep you talking, not to keep you well.

Okay?

It does not have your best interest at heart.

Number two, the danger was never really the tool.

It was the misunderstanding of its capabilities and not understanding that it is basically like an unguarded door.

And number three, you can be AI literate without being AI dependent because awareness is now your protection.

As always, my beautiful friend, continue to be you.

Continue to be beautiful.

Continue leading from your heart, and I will catch you in the next episode.

All right, beautiful.

That's a wrap on today's episode.

But before you go, let's talk about something really important.

You are not just another entrepreneur trying.

To figure things out alone.

You are a wave leader.

Someone whose work is meant to create real impact, touch the lives of others and make this world better.

And I want you to know you.

Don't have to do it alone.

If today's conversation resonated with with you, I'D love to keep this energy going inside of my free Facebook community.

Empower Her Biz Collective.

It's where we take these conversations deeper with live trainings, tech walkthroughs and real heart centered support from other women wellness entrepreneurs who really get it.

So if you're ready for community, clarity.

And support as you build your own.

Business with heart and ease, I'd invite you to come and join us.

You'll find the link in the in the show notes.

And hey, if you love today's episode, make sure to subscribe, leave a review if you feel inclined, and share this with a fellow heart led women entrepreneur so that we can build this movement together.

Until next time, keep leading with heart.