Insights
Analysis that goes past the announcement
New AI capabilities are announced every week. The harder questions come after: what the research actually makes possible, what it takes to build and run, and whether it will deliver value you can measure. That is what these insights are for.
Selected analysis
Curated analyses are being added
Selected articles and posts will be featured here, each with a link to the original piece.
Want a topic covered? Suggest it by email, or bring it to a monthly roundtable.
Topics
- Enterprise AI and Agentic Systems
- Architecture and Governance
- Physical AI and Robotics
- Industrial and Manufacturing AI
- Research to Real-World Implementation
How each piece is written
- What does it enable?What the underlying research really makes possible, and what it doesn’t.
- What does it require?The architecture, data and integration needed to implement it.
- What could go wrong?Practical constraints, failure modes and risks.
- Will it hold up?Reliability and performance once it’s running in operations.
- Is it worth it?Where the measurable business value is, and where it isn’t.