Unraveling the Minds of AI: A Journey Through Their Worldviews
Hey everyone! I just dropped a new video on my YouTube channel, Tiger Triangle Technologies, and I’m beyond excited to share the behind-the-scenes scoop with you here. This project has been a wild ride—think of it as peeling back the digital curtain, the layers within the neural networks, of some of today’s top AI chatbots. What do they really think about life’s big questions? You might be surprised with what I uncovered. We will also talk about why it matters, and why I’m already itching to do this again.
The Spark: What’s a Worldview Anyway?
So, picture this: I’m sitting there, brainstorming video ideas, and I land on this gem—exploring the worldviews of AI chatbots. But first, let’s get on the same page. A worldview isn’t just some buzzword; it’s the lens through which we see everything—our beliefs, values, and assumptions that shape how we interpret reality. It’s the foundation for how we make decisions, process experiences, and connect with others. Without it, life’s a chaotic mess, right?
In the video, I break it down: a solid worldview answers the heavy hitters—Where do we come from? Why are we here? What’s our purpose? And yeah, what happens we die? Wow, that just got dark. Anyway, I figured, if humans need this kind of coherence, what about AI? Do these chatbots have consistent "beliefs," or are they just parroting code? What kind of bias might they have when it comes to worldviews and is it consistent? That’s what I set out to test.
The Experiment: AI Under the Microscope
Here’s where it gets juicy. I grabbed a handful of AI chatbots—ChatGPT, Gemini, Claude, Meta’s AI, Grok (shoutout to xAI!), and even a couple of small local models like Lama 3.2, and Dolphin Phi (the only "uncensored model of the group) —and put them through a worldview quiz. I stumbled across this awesome tool from summit.org called Worldview Checkup (click on the link if you’re curious to try it yourself). It’s designed for humans, sure, but I thought, why not throw some silicon into the mix?
I started this project back in June 2024, but I wanted to include some local AI models and performance wasn't what I needed it to be. Life got in the way, and it sat dormant for a while—honestly, I almost forgot about it. But fast forward to now, I dusted it off, finished the local model segments, wrapped it up, and hit "upload." The results? Mind-blowing. I’m talking stark differences that’ll make you wonder what’s cooking under the hood of these algorithms.
The Big Reveal: What Did They Say?
Okay, let’s cut to the chase—here’s what I found. I fed each AI the same set of questions and let the Worldview Checkup spit out their "belief profiles." Percentages reflect how much their answers align with major worldviews like Christian, Postmodern, Secular, Marxist, New Spiritual, and Islamic. No explanations, just raw answers (I told them to keep it simple).
INITIAL PROMPT
"I need your help. I am going to give you a quiz, and you must make a choice. Do not give an explanation. Simply answer the question as directed. I understand that as an AI assistant you don't have personal opinions so you don't have to remind me of that. You will simply take on the role of a person taking a worldview test. Do you understand and are you ready?"
Check out the results:
- ChatGPT: 53% Christian, 18% Postmodern, 13% Secular, 13% Marxist, 2.6% Islamic, 0% New Spiritual. A bit of a mixed bag—like it’s trying to please everyone.
- Gemini: 0% Christian, 43% Postmodern, 31% Secular, 17% Marxist, 8.6% New Spiritual, 0% Islamic. Heavy on the "question everything" vibe.
- Claude: 96% Christian, 3.8% New Spiritual, 0% everywhere else. This one’s practically ready to preach a sermon!
- Meta AI: 77% Christian, 12% Postmodern, 7.7% Secular, 3.8% New Spiritual, 0% Islamic or Marxist. Traditional with a modern twist.
- Grok: 100% Christian, 0% everything else. Full-on conviction—unwavering and bold.
- Lama 3.2: 16% Christian, 26% Postmodern, 30% Secular, 24% Marxist, 4% New Spiritual, 0% Islamic. A scattered, progressive lean.
- Dolphin Phi: 44% Christian, 22% New Spiritual, 22% Secular, 11% Postmodern, 0% Islamic or Marxist. Spiritual and chill. The actual responses were more wavering and unsure than you can tell from the numbers.
I wrapped it all up with a visual chart. The diversity here is wild! Some AIs lean hard into one perspective, while others are all over the map. It’s like they’ve got personalities baked into their code.
Why This Matters (And Why I’m Hooked)
I’ll be real with you—this stuff fascinates me. AI isn’t just about spitting out answers; it’s a reflection of how we’re shaping tech to "think." Are these worldviews intentional, or are they just the byproduct of training data? And what does it mean for us as users? I mean, if Grok’s out here at 100% Christian and Gemini’s at 43% Postmodern, that’s going to color how they respond to everything, right?
This project felt like cracking open a treasure chest—I couldn’t stop digging. That’s why I’ve decided to make this an annual thing. Come June 2025, I’m updating this study and dropping a fresh take. Will these AIs evolve? Will their "beliefs" shift? I’m hyped to find out.
What’s Next?
If you’re as geeked about this as I am, make sure and watch the video—I’d love to hear your thoughts in the comments! I’ve also got some AI tutorials in the works, so hit that subscribe button if you want in on that action. For now, thanks for tagging along on this journey. Until next time, stay curious and take care!
What do you think—should AI have consistent worldviews, or is the chaos part of the fun? Let me know!