Amazon's CFO has increased the company's 2026 capital expenditure forecast to $220 billion, anticipating that this sum may still be insufficient to meet demand, particularly concerning memory chip availability for AI infrastructure. Concurrently, Alibaba has released Qwen3.8-Max, a model employing sparse mixture-of-experts routing with 2.4 trillion total parameters, though only 95 billion are active per query, which is expected to reduce inference costs. A smaller Qwen3.8-27B variant is also slated for release, with pricing set at $2 per million input tokens. AI
IMPACT Amazon's significant capital expenditure increase highlights the growing demand for AI infrastructure, while Qwen3.8-Max's efficient routing could lower AI inference costs.
RANK_REASON Cluster covers both a major capital expenditure announcement from a large tech company and a new model release from a significant AI lab.
Read on Mastodon — fosstodon.org →
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
