Knowledge expansion priorities

{
  "missing_count": 5,
  "stale_count": 10,
  "gaps": [
    "Chinese AI ecosystem dynamics and US-China competition",
    "AI hardware/infrastructure beyond Nvidia (Cerebras, Groq, SambaNova)",
    "Open-source model ecosystem (Mistral, Llama, Yi, Qwen)",
    "AI safety/alignment technical approaches beyond high-level topics",
    "Edge AI and on-device inference (Apple, Qualcomm, Google Nano)",
    "Multimodal model convergence (text+image+audio+video)",
    "AI for science and research applications",
    "Regulatory developments (EU AI Act, US executive orders)",
    "AI talent flows between companies/academia",
    "Energy consumption and environmental impact of AI"
  ],
  "improvements": [
    "Track model versioning more systematically (e.g., Claude 3.5 Sonnet → 3.6)",
    "Add entity attributes: funding amount, valuation, employee count, key customers",
    "Create relationship strength metrics (e.g., partnership intensity 1-10)",
    "Track competitive threat levels between companies",
    "Monitor sentiment shifts around specific technologies (e.g., RAG vs. fine-tuning)",
    "Build timeline of model capability jumps (tokens/sec, context length, multimodal)",
    "Track researcher migrations between labs as leading indicators",
    "Monitor API pricing changes as commoditization signals",
    "Track open-source vs. closed-source model performance gaps",
    "Create technology adoption curves (early adopters → mainstream)"
  ]
}