What if We’re at the Start of a New Context?

Hi,

There’s this idea that’s been growing in my mind for a while now — and although I’m still not sure how much of it is real and how much is just temporary excitement, I know it’s something I can’t ignore.

I’ve recently been realizing how deeply I vibe with the digital — not just as a tool, but almost as a language, a vibe, a space I feel at home in.
At the same time, I hate the actual process of graphic design — I don’t want to be the executor. I want to be the thinker, the visionary, and have someone (or something) bring it to life.

That’s when I realized: AI can be that something.
With skills like prompt design, we’re not just designing anymore — we’re directing.
And that changes everything.

It made me think: this isn’t just a new tool. It’s a new context.
Like how the internet in the 90s changed the world and became the foundation for platforms like Google, Facebook, and Amazon — I believe AI is now becoming that same kind of foundation.

But we’re still in the early days. Most tools and platforms today are just AI-powered.
They’re not native to the AI context.

And I keep thinking:

What will be the first truly native AI platform?
What kind of tools will define this new playground?

I know I want to be part of that.
Maybe even build something that stands on this new foundation — just like Instagram stood on the internet.

It’s still raw, and maybe I’m early. But I feel like the right conversations now could shape something real.
So I’m reaching out.

Would love to hear your thoughts if this resonates with you even a little.

Best,
Mehdi Arman

3 Likes

just thought I'd point out that with things like dynamic system prompts from Pydantic AI you can actually set up an Agent to write the prompt for another Agent, effectively using the message history to re-write the prompt every time it's ran.... Its an AI doing the prompt engineering of another AI, so technically, you don't even need that skill (although I prefer my AI and their tools to be, mostly, designed by a person and not a sleeper agent :joy:).

out of curiosity though, are you thinking of like a point-and-click/paint system? like you draw out a page, or part of one, then give it some text to describe interactions and it creates the html/css?