Sure, there's so much more than tokens. Context, compute, whatever you want to name. But I'll say it anyway, because tokens are simply how AI works.
And to be clear: I'm not asking anyone to just hand us more tokens. I'd take ChatGPT over Gemini any day (I still miss Bard), so "more tokens" is not the wish here.
The point is different. No matter which setup you run AI in, tokens = money. And I see extreme, almost absurd potential in getting the same result for much, much, much fewer tokens — just by creatively combining old-school and new techniques. That's the whole idea behind my multi-purpose professional-model approach.
So I treat tokens as money. Honestly, more like capital.
And hey — it's June 21
One month on Claude Code Pro-Max x20 is over.
Did it perform? Beyond my imagination. It was mind-blowing — but more importantly, it was mind-opening. Watching this thing work, especially after some of my own tweaks, felt like looking straight at the future.
My conclusion
Again, the same line, because it keeps being true: soon is now, and the only limit is tokens.
Nearly everything is scalable now. With a bit of creative design and a good architecture, we can digitize and scale almost any process — today, not someday. And it only gets better and easier from here.
