
In this solo “Manny’s Reflections” episode of the Get Paid podcast, host Manny Medina dives into one of the most consequential shifts in the AI market right now: the reversal from tokenmaxxing to rigorous cost optimization. He also spotlights a LinkedIn post from Shriram Sridharan, CTO and Co-Founder of Rox, who shared a framework around measuring Dollars Per Action (DOPA) and demonstrated a 10x reduction in token spend even while scaling usage. ● The great reversal from tokenmaxxing to “what am I actually getting for this AI spend” ● Shriram Sridharan's Dollars Per Action framework and why it didn't get the attention it deserved ● Dollars per token vs. tokens per action, and why only one lever is actually moving ● Foundation models with too much tenacity and the token burn nobody asked for ● The momentary respite from open source models and why it won't last ● Data center permitting backlash, aging grids, and the Texas governor saying no more ● NVIDIA chips aren't getting cheape...
Loading summary