Julie Park.
← Julie Park Journal

The Contextual Frontier

AI has all but closed the information gap. The gap it hasn't closed is the one between information and action, and that's a different problem entirely. You don't cross it with a faster model, lower latency, or higher accuracy. You cross it with an AI that understands your context. I believe context-aware AI is the next frontier. This is where I track it.

01
Latest

Context-Aware AI: Why LLMs Alone Aren't Enough

On the empty space between information and action — and why closing it is a different problem from making the model better.

August 2026·9 min read
Read the essay →
In the works

Each one goes live here as it's finished.

02 Seeing the World: Egocentric AI and the future Project Aria is buildingInside Meta's Ray-Ban strategy: why glasses are the next platform. Drafting
03 Understanding a Person: Contextual AI and agentic memoryThe trap of the chatbot abstraction, and what real, persistent context looks like. Drafting
04 Taking Action: the frontier of Physical AI and VLANVIDIA Cosmos and the Physical AI stack: AI infrastructure's next layer. Drafting
05 Where the Four Layers Meet: the integrated framework for context-aware AIApple vs Google vs OpenAI: three roads to personal context. Drafting
06 The Privacy Paradox of Egocentric AIWhat we trade for an AI that sees what we see. Drafting
07 Telcos’ Second Chance: the carrier as context infrastructureWhy the operators that missed the platform era may own the context layer. Drafting
08 Four Months at HBS: a field report on the AI industryWhat looks different about the AI landscape from inside business school. Drafting
Egocentric AIContextual AIPhysical AIInfrastructureTelco