DigiNews

Tech Watch by Johan Denoyer

← Back to articles

tokens too cheap to meter

Quality: 8/10 Relevance: 9/10

Summary

The article argues that AI compute costs are dropping rapidly, predicting LLMs running on commodity hardware in the next few years and cheaper per-task economics. It covers hosted vs local AI, architectural improvements like Mixture-of-Experts and Jev/Laya, and the broader market implications including investor dynamics and future usage scenarios.

🚀 Service construit par Johan Denoyer