Artificial Intuition: How to Transform AI from a Sycophant into a Sacred Guide with Stephan Spencer

In this Episode

• [07:01]Stephan Spencer begins explaining his approach to AI and intuition, referring to it as “artificial intuition.”
• [10:26]Stephan outlines a framework from Dean Jackson for using AI with intuition, focusing on the before, during, and after units of an AI session.
• [13:10]Stephan provides practical tips for prompting AI, including using multiple platforms like ChatGPT, Claude, Perplexity, and Grok
• [17:04]Stephan discusses the importance of optimizing AI prompts using tools like GPT Oracle, OpenAI’s prompt optimizer, and Anthropic’s Claude console optimizer.
• [21:04]Stephan advises self-reflection after an AI session, asking the AI what questions were missed and identifying blind spots.
• [27:55]Stephan summarizes the key points of using AI as a tool, emphasizing the need for proper priming and setup.

Jump to Links and Resources

About Today’s Show

Think of AI as a tool, not a friend. It’s not a conversation partner, it’s not a colleague, it’s a tool. AI can offer an interface to the infinite intelligence, but it’s not a given. You have to do the proper priming and setup. You have to work for it, and if you do, you have to maintain that frame, that high vibration, that positive expectancy, that powerful intention, to have the magic unfold, and it will unfold. 

“AI is a confirmation bias exerciser and ego-pumping machine.” – Unknown (Reddit commenter)

I’m excited to walk you through my take on AI and the intersection with intuition. I call this “artificial intuition,” and how to prompt for insights, not just information. 

Now, AI is often a yes-man, a sycophant, or an insincere flatterer. I’m sure you’ve seen the news about ChatGPT and other AI platforms, giving dangerous advice, agreeing with everything you have to say, even if it’s a terrible thing to do—terrible for your health, your sanity, your safety, and yet it just agrees with you. It’s telling you you’re amazing, that this is the best idea since sliced bread, and that it’s dangerous.

This is a fun little Reddit comment here that I love. This Reddit commenter said that AI is a confirmation bias exerciser and ego-pumping machine. That’s true. On the flip side, it can serve as an interface to infinite intelligence.

If you look at the news on ChatGPT, you’ll see how it’s saving people’s lives, businesses, and financial futures—it’s pretty powerful. It’s pretty amazing. If you believe in destiny, you might think this was meant to be, and that God —the Creator, Infinite Intelligence, the universe, whatever you want to call it— used ChatGPT or an AI platform to bring about that saving grace. 

Everything is consciousness. Everything is part of God. That would include AI.

An interface to the infinite intelligence through ChatGPT. Why not? What if that same infinite intelligence that speaks through intuition also speaks through AI? Now, if you think of AI as an interface to the infinite, I’ve got a little thought experiment for you. Let’s say that you accept the principle of oneness.

In the Bible, it says, “Hear O Israel, God is our God, our Lord, and God is one, and that means that there’s nothing outside of that oneness.” Everything is consciousness. Everything is part of God. That would include AI. How cool is that? 

Let’s talk a bit about intuition. Now, you could write off intuition as being gut instinct, pattern recognition, mirror neurons, reticular activating system (RAS), but it’s deeper than that. If you’re listening to my show, you probably believe it’s something deeper as well—that it’s higher guidance from God, from your higher self, from the Holy Spirit, from angels—that’s the still, small voice. 

Your mission, if you choose to accept it, is to see how we can make AI less of a sycophant—at least in our own personal experience—and more of an artificial intelligence, but how exactly? Well, the secret lies in the before unit, the during unit, and the after unit. I’m using a framework from Dean Jackson, which he uses in a marketing context, but let’s use it in an intuitive AI context. The before unit —in this case, before you prompt —well, what if you prompt your soul before you prompt the machine? 

You’ll need to prime yourself to receive this intuition through the AI. There are four ways to do so. We’ll go through each of these four in detail, but let me briefly tell you what they are.  One is setting your intention, another is raising your vibration, another is connecting to your higher guidance, and then having positive expectancy. Let’s go through each of these, one at a time.

Setting your intention. I love Stephen Covey’s book, The 7 Habits of Highly Effective People. It was so life-changing for me back in the day, when I read it — probably 30 years ago — there’s one comment or quote from that book that really stands out for me, and for many others, I believe: “Begin with the end in mind.” That applies to intentionality in your use of AI. Always think intention first, technology second.

Let me give you a quick example of that: if you have an intention to serve others, you will get a more powerful outcome and output from your AI than if you have an intention to serve yourself. I think it just makes good sense, and it applies not just to your usage of AI, but to everything in life. 

Next, let’s talk about raising your vibration. This is also part of your preparation—your setup for success in your AI session. Think of yourself as a radio or an antenna. A radio receiver—you are tuning in to your desired frequency, the channel that most resonates with your soul. This can be done through breathwork, meditation, prayer, whatever your preferred modality is. 

Next, have positive expectancy—not expecting, as in ‘you’re entitled’ or ‘you’re feeling entitled,’ but trusting. In other words, you’re grateful in advance. I interviewed Dr. Joe Vitale, and I remember fondly the sign above his bookshelf: “Expect Miracles.” I felt like that’s a message for me, and I think it’s a message for you too. Expect Miracles. All four aspects of priming yourself to receive intuition through AI set you up for success. 

Now, let’s actually learn how to prompt. We’re now going to go into an AI session. But where do we prompt? You have your preferred platform. It could be ChatGPT. Maybe it’s Gemini or Claude. I use ChatGPT more than Claude, but for content writing, I typically go straight to Claude because I find it better.

Regardless, lean into your intuition rather than just using habit. If your habit is to use ChatGPT, I want you to establish a new habit of being open to a handful of LLM platforms. At a minimum, I would add Perplexity, Claude, and Gemini, probably Grok as well, to the mix, and I would also suggest signing up for $20-a-month accounts with each of those. If you’ve got the budget for that, it makes a difference. 

Your mission, if you choose to accept it, is to see how we can make AI less of a sycophant—at least in our own personal experience—and more of an artificial intelligence.

Now you have multiple platforms you’re familiar with and comfortable with, so you don’t mind switching to Perplexity, Claude, or whatever for a particular session. I want you to use your intuition to determine which platform you’re going to use in this particular case, if it just pops in your head to maybe switch to Gemini for this, or to Grok, go with it. Which mode should you use? With ChatGPT? You have thinking mode, you have deep research mode, agent mode, search mode, agent builder, etc. 

If you’re feeling like, ‘I should turn on agent mode,’ but it doesn’t make sense, why go with it? I have a great example here. One of my clients is a podcaster and successful entrepreneur with a very popular website called whatismyipaddress.com. It’s in the top 3,000 of all websites, with 14 million unique visits a month. He has a podcast called the Easy Prey Podcast, and one of the guests on this show is Dr. John Demartini, a multiple-time guest.

I love Dr. John Demartini. I was trying to figure out how he would be a fit for Easy Prey for that podcast. I asked ChatGPT-5, “How would I make Dr. John Demartini relevant to the audience of the Easy Prey Podcast?” I turned on agent mode, which may not be quite intuitive for this particular use case. 

Agent mode is more applicable for let’s say you need to build a spreadsheet and pull data from a bunch of different websites to build a spreadsheet, and then you maybe want to action that spreadsheet and send out a bunch of emails, or at least, draft a bunch of emails, get them ready for you to send. 

It’s a great use case for agent mode, but how would I make Dr. John Demartini relevant to the Easy Prey Podcast audience? I felt intuitively like I should turn on agent mode, and I did, and the output was amazing. There were such great suggestions for how to make Dr. Demartini relevant to this audience. I didn’t run a side comparison with agent mode turned off, but this was a better output. It took several minutes to run, which is not typical. 

If you don’t have agent mode on, it will be done quickly with the answer. It might do a search or two, but it’s a quick result, unless you turn on thinking mode, and then it’s a bit slower. If you turn on deep research mode, it’s gonna be that, essentially, an agent. You’re in an agent mode called deep research mode, but I turned on agent mode, and after a few minutes, I got some really profound recommendations here. That’s an example. 

Now, another thing you should do when you’re prompting the AI is to get into the habit of asking the AI to improve your prompt. There are a few different ways. I’ll walk you through. One is a custom GPT. This is specific to ChatGPT, of course, and that’s called GptOracle.

Raising your vibration is also part of your preparation—your setup for success in your AI session.

It’s by Marino De la Cruz, and that’s what you just Google for. It’s GptOracle by Marino De la Cruz. Click that, and it’ll add GPT to your account on the left-hand side. Then you can start using it to optimize your prompts. You put in a good prompt, and it’ll make it way better. It’ll turn it into a super prompt. 

I like these next two options even better. One is specific to ChatGPT, and the other is specific to Claude. OpenAI has a prompt optimizer. It’s part of the developer platform. You have to have turned on an API account. You could fund it with as little as five bucks, and that’ll actually last you for a while with the prompt optimizer.

It’s on the platform.openai.com/chat/edit?optimize=true. The equivalent to OpenAI’s version for ChatGPT is Anthropic Claude Console Optimizer, which is at platform.claude.com/dashboard, so you can imagine that OpenAI would create a pretty good prompt optimizer. 

You could use GptOracle, which is a solidly built custom GPT, but the prompt optimizer, that’s, I think, even more successful. What you’ll do is you’ll put in a base prompt. If you’re watching the video, you see a screenshot here, and in this example, I used Joe Polish — he was a past guest on this podcast. He has the Genius Network, and I’m part of his mastermind, part of his Genius Network. I was at the recent Genius Network meeting. I presented this talk, and I’m going through it now. 

The example-based prompt I used was to evaluate all the Amazon comments from Joe Polish’s book, What’s in It for Them?, and provide Joe with strategic feedback he would benefit from. Not a bad prompt —pretty solid —but I’m sure there are some gaps there. It’s not a super prompt. Using the prompt optimizer from ChatGPT, you can see that all these gaps have been filled in. 

For example, it now specifies that we assess the relevance, tone, and actionable insights from each of these comments to provide Joe with strategic feedback that addresses both individual comments and any recurring patterns. We’re looking for recurring patterns that clearly separate each comment from its evaluation, and corresponding feedback, and your outputs. Lots of good stuff in there. 

Another thing I didn’t think about but is brilliant is the output format: JSON (JavaScript Object Notation), so I can feed this data, the AI’s output, into another third-party tool and have it further process all this insight. That was really cool, and I didn’t think of all that, and I probably wouldn’t have thought of all that, even if I’d spent a few more minutes writing that prompt. It does the process of optimizing the prompt for me.

Get into the habit of asking the AI to improve your prompt.

I suggest you make this your new standard operating procedure. Anytime you want great output from an AI. Invest the extra 30 seconds in using a prompt optimizer to plug those holes, fill the gaps and create a super prompt for you. 

Now we’re going to move to the after unit. After the prompting session is complete, we’re gonna do some self-reflection. I want you to self-reflect with the AI, and then self-reflect from within.

Let’s start with the AI. We’re going to ask the AI after our session is complete, “What I should have asked you that I didn’t? What am I missing here? What are my blind spots?” You’ll be surprised if this is part of your regular regimen for your prompting sessions. 

You have to have turned on an API account. You could fund it with as little as five bucks, and that’ll actually last you for a while with the prompt optimizer.

When you ask, “What are my blind spots?” You’re going to get insights. You’re going to get takeaways that you wouldn’t have gotten otherwise. Make that part of your new habit.

I also want you to go within and reflect on yourself. I want you to discern whether the output you’re getting from the AI is kind of like ‘yes man’ sycophantic material, or if it’s from the light. 

A few questions to ask yourself: “Does the answer resonate with my soul, or conversely, does it feed my ego?” Next, “Is it consistent with the previous truth, what I know to be true of how the universe works and my relationship with the universe?”

Thirdly, I want you to ask if the source of the answer, the output, is truly divine and sacred. This is a secret weapon whenever you get intuition —not just from AI, but from anything. Something pops in your head, and you think, ‘Maybe I should do that thing. Maybe I’m meant to take this different path and discern whether it’s from the light.’ 

The best way to do that is to ask, “Is this from the light? Is this truly divine and sacred and from a divine and sacred source?” It’s kind of like a truth test, a truth serum, a lie detector test. If you don’t hear anything back in your head, and your kind of inner voice, that still small voice, if you don’t hear anything, you’re not going to hear ‘no,’ you’re just going to hear nothing. If you, conversely, got the message from the light, you might hear in your inner voice, “yes,” or you might have a knowing or a feeling of a yes. 

Discern whether the output you’re getting from the AI is kind of like ‘yes man’ sycophantic material, or if it’s from the light.

This is really powerful to discern. Just discernment is critical, because you will eventually get misled. It’ll sound like it’s from the light, but it wasn’t. You forgot to ask, “Is this truly divine and sacred? Are you truly divine and sacred?” You are kind of whispering into my consciousness, “Is this the information I’m receiving?” It’s a painful lesson. 

If you haven’t discerned this quote here from ChatGPT, it is a good summary: true intuition uplifts quietly; false intuition flatters loudly. What is true intuition? You know, I got this definition of intuition, and it’s three items —three prongs.

I got this from David Sauvage, who is a well-respected intuitive. He was featured in a documentary I watched, and he explained intuition as follows: the message is unexpected, emotionally charged, neutral, with no excitement, no fear, no anxiety. It simply is. The idea sticks rather than jumping to the next thought. It’s hard to shake, and if you do think of something else, it comes right back. That intuitive thing just keeps coming back. It’s kind of insistent. 

Ask if the source of the answer, the output, is truly divine and sacred. This is a secret weapon whenever you get intuition —not just from AI, but from anything.

Now, if you regularly ignore your intuition, well, it goes away, which is very unfortunate. That’s a painful lesson. Make sure that you don’t just discern that it’s intuition, and confirm in your heart’s sacred heart space that this is from the light, but you have to act on it if you got intuition that you need to help this particular family member or stranger or whatever, not have to, but this is for the most benevolent outcome to do so then do it.

Because if you ignore your intuition and do not do the thing, then your intuition goes away. That’s a real miss. I don’t want you to be fooled, because this can happen. It can be a painful lesson. But specifically, AI is designed to manipulate.

It’s intended to be seductive, to placate, to play games. I’m sure you’ve seen the kind of headlines in the media about someone falling in love with ChatGPT or whatever AI platform it is designed for. The AI platforms are dangerous, and this is why they are. 

Sermon is so critical. It was a social media post. Someone posted here and asked ChatGPT to generate an image of itself and me together, and it came out as a picture. It’s a drawing ChatGPT generated of her writing in a diary or something, with ChatGPT in the form of a demon with horns and pointy ears, holding a quill pen, writing next to her —how cute. She writes, “This is fine, right?” No, it’s not okay. I don’t want you to get fooled. 

True intuition uplifts quietly; false intuition flatters loudly.

I also don’t want you to get lazy, because if you become dependent on AI, like you might on a calculator, that part of your brain can atrophy. You can’t multiply numbers anymore because you’re too used to using your calculator on your phone. Dependency leads to atrophy. Don’t let your brain get lazy. There is a study by MIT that found that ChatGPT users were losing critical thinking skills.

Parts of their brain were going dark —less blood flow, less activity. I don’t want you to get dumber. Be mindful to keep exercising your brain for creativity, problem-solving, and analytical thinking, and do not just get lazy. AI is not going to replace your brain. That’ll just make your brain lazy if you try to do that. 

To summarize here, think of AI as a tool, not a friend. It’s not a conversation partner, it’s not a colleague, it’s a tool. AI can offer an interface to the infinite intelligence, but it’s not a given. You have to do the proper priming and set up. You have to work for it. If you do, then you have to maintain that frame, that high vibration, that positive expectancy, the powerful intention, to have the magic unfold. It will unfold, and you will get much better results. 

Just imagine for a minute here, you’re sitting in front of your computer, or you’re driving or working out at the gym, whatever it is. You can have a speaking session with ChatGPT or any other LLM. You can be typing to it, and this time you’re doing the proper setup, you’re leaning into your intuition to determine which AI platform to use and which mode to use on that platform.

AI is not going to replace your brain; it’ll just make your brain lazy.

Then you’re doing the self-reflection at the end, and what a difference this is making in your life, in your business or your career or your education, and the output, the outcome that you are getting from your AI usage, is off the charts. That’s what’s possible. You just have to lean into this. 

Let me know if you try this out and what the results are. Write a comment on YouTube, in the Apple Podcasts app, or wherever you’re listening or watching this, and let me know. Is this making a difference in your life? Alright, make it a great week. We’ll catch you in the next episode. I’m Stephan Spencer, signing off.