Sticker shock has execs rethinking this whole AI thing
Anthropic
OpenAI
GitHub
Business leaders are shocked by rising AI costs as usage-based billing replaces flat fees. A KPMG survey found 29% of senior execs struggle to understand operating costs, and nearly half are rethinking deployments. Meanwhile, an open source tool from a Netflix engineer trims redundant input to save on token costs.
A KPMG survey of over 2,000 senior executives across 20 countries found that 29% struggle to understand operating costs as they scale enterprise AI deployments, and nearly half are re-phasing deployments when costs outweigh expected value. Vendors like Anthropic, OpenAI, and GitHub have moved from flat-fee subscriptions to usage-based billing per token, causing bill shock. Gartner research indicated that AI-assisted coding costs per developer could exceed global average salary by 2028, especially in lower-wage regions. An open source tool created by a Netflix engineer trims redundant input (like database schemas and logs) sent to LLMs, saving users hundreds of thousands of dollars in token costs. The tool uses 'squashers' to remove unnecessary verbosity before submission.
- Сокращения
- LLM = Large Language Model — большая языковая модель
- MCP = Model Context Protocol — протокол контекста модели
Source: The Register AI/ML —
original