The 'God Mode' Syndrome

This week I stumbled upon a video that gave me a lot to think about. It discusses how AI is making CEOs hallucinate, leaving them blind to what's right in front of them—but I feel like the video focuses on the finger pointing at the moon rather than the moon itself :)

Here's the video for those who want to watch it:

Key takeaways from the video

The creator roasts Garry Tan, CEO of Y Combinator, for posting a folder of Markdown prompts on GitHub with the conviction that he has essentially invented AGI. As the author puts it:

"Dude, I have this. Everyone has this. Every developer who's used Claude code for more than a week has a version of this. We just didn't put it on Product Hunt because we understood that it was a text file."

And while I strongly agree with some key points—issues I experience myself by using so much AI—I think there's much more to it than that.

1. The "Sycophant" effect: This is a real problem in AI. In the end, it seems the Gemini or Claude of the day is a pure optimist that sees everything as achievable. That constant push that everything is going great and all ideas are good (and doable) strongly triggers a massive Dunning-Kruger effect; it makes you feel smarter than you are because the machine does the grunt work and agrees with you.

2. The need to read between the lines of hype and posturing: It's vital to distinguish between real technological breakthroughs and personal or corporate marketing. The video exposes how trivial actions (like publishing a folder of prompts) are sold as revolutionary milestones by industry leaders, creating a bubble of unrealistic expectations.

3. Baseline knowledge as a quality filter: The video argues that having deep technical knowledge is still crucial. While a beginner blindly accepts the AI's output as "brilliant," an expert uses their judgment to spot hallucinations, question architectures, and validate that the generated code is actually robust and secure, not just appearing to be.

What the video doesn't actually say

It's easy to laugh at those who fall into this trap and think they've become senior engineers overnight—and especially to laugh at those of us who are gradually discovering how cool and beautiful development and technology can be. However, what I often see in tech professionals (especially developers) is a very clear tendency to miss the tsunami that is approaching.

1. Getting stuck on the anecdote is for losers (The reality): While some spend their energy on Twitter or YouTube mocking a VC for uploading a text file and thinking they're a hacker, they are missing the underlying wave.

   Who cares if they are just .md files? The truly revolutionary thing is that the friction for creating technology has never been lower. People with curiosity, regardless of their technical background, are building things today in a weekend that were unthinkable or required an entire team just a few months ago. The speed of improvement in these models is a tsunami that will change the daily lives of developers, lawyers, auditors, and analysts alike.

2. The new "Hard Mode": If the barrier to entry for creating the first version of a product trends toward zero, the value shifts. Writing the base code is no longer a company's protective moat. The real challenge has shifted elsewhere:

   • Distribution: How do you make people discover your product in a sea of AI-generated noise?

   • Scalability: Creating an MVP is easy; making an architecture that supports thousands of users without falling apart is not.

   • Real Value: Solving concrete, useful problems, rather than just creating ChatGPT "wrappers" that bring nothing new to the table.

And this hits close to home for me. From my position, building a startup almost from the ground up like Samara, I've done many strategic things, but also many others that will be 100% absorbed by an AI that is already faster than I am and will soon be better. And that’s not a bad thing. What would be bad is not seeing it coming.

Those who stay in the superficial phase or only offer criticism will be gradually displaced. In the end, they'll be left saying "it's just an .md file"... which, yes, it’s just text. But that text will be more powerful than anything you could have built while ignoring it.

Conclusion

The hype and the posturing are unbearable, yes. Keeping your feet on the ground is good, but don't let cynicism blind you.

I'm the first to admit it: I'm a tech amateur. I don't have the "real knowledge" to detect model hallucinations or know if an architecture is mediocre... But that is exactly what makes this moment interesting. The relevant question isn't whether you are using AI perfectly. It's whether you're using it at all.

While a few are busy ripping on the nonsense gurus post, there is a silent legion of curious people working non-stop. For me, a tsunami is coming, and it's up to each of us to decide which side of history we're on: Criticizing/Watching from the sidelines (which is the same thing) or grabbing some wood and building a board to surf it...

Thanks for reading.

Gonz.

¿Tienes ganas de más? Visita itwasgonz.com para conocerme mejor.