💼 US Corporations are Cutting AI Spending
Major American companies have begun reviewing their AI expenditures, moving away from the unlimited use of the most expensive and powerful models. Instead, businesses are transitioning to a hybrid strategy: using cheaper, specialized, or open-source models for routine tasks, driven by the sharp rise in token costs.
🌍 A paradigm shift from a "power race" to an "efficiency economy." This is stimulating the development of specialized Small Language Models (SLMs) and open-source solutions, while forcing providers to compete not only on quality but also on price affordability.
👤 Using top-tier models (such as Claude 3.5 Sonnet or GPT-4o) for simple tasks is becoming economically unfeasible. An increase in the popularity of local solutions and more flexible API strategies is expected.
Source 1: https://www.wsj.com/business/china-us-ai-model-costs-53a12e96
Source 2: https://www.democraticunderground.com/100221399449