Eseandre Mordi
⬤ Token pricing for large language models is in absolute freefall right now. The smartest, most capable models are getting 900x cheaper every year. Mid-range models? They're dropping 40x annually. Even the basic models are sliding 9x per year. This isn't just a price adjustment—it's a full-blown collapse that's tracking similar to Moore's Law, which predicted computing costs would keep plummeting over time.
⬤ The data breakdown shows this price crash isn't hitting all models equally. Advanced models with the highest intelligence levels are experiencing the steepest drops in cost per million tokens. What used to be prohibitively expensive AI is now within reach for way more developers and businesses. This affordability shift is cracking open use cases that were financially impossible just months ago.
⬤ We're watching Jevons paradox play out in real time here. When something gets radically cheaper and more efficient, demand doesn't just tick up a bit—it explodes. High costs used to gatekeep who could use powerful AI models and how much computing power they could throw at problems. Now that barrier's crumbling fast, opening the floodgates for applications nobody even thought about before. It's the same pattern we saw with Moore's Law transforming computing from massive mainframes to smartphones in everyone's pocket.
⬤ This price collapse matters because it's setting up AI to follow the same explosive growth curve that computing took decades ago. Affordable AI means more experimentation, more integration into everyday tools, and entire industries getting rebuilt around what's suddenly economically viable. The question isn't if this reshapes how we work and build anymore—it's how fast it happens.
Eseandre Mordi
