Insights
Ideas on AI, built for the real world
Clear, practical takes on where AI is heading — and what it means for Canadian businesses.
AI agents: from a model that talks to one that does the work
How tool calling and a plan-act-observe loop turn a chatbot into an agent that gets things done — and how to keep it safe.
ReadExplainerMultimodal AI & VLA: models that see, read, and act
How AI takes in images and screens — not just text — and how vision-language-action models go from understanding a scene to acting in it.
ReadExplainerRAG: how AI answers from your data, not its memory
Retrieval-augmented generation, explained — how models ground answers in your own documents with embeddings, vector search, and citations.
ReadExplainerHow LLMs actually work: tokens, attention & next-token prediction
A plain-language look under the hood of large language models — and why understanding it explains both their power and their limits.
ReadAI FrontierWorld models vs. LLMs: Yann LeCun's billion-dollar bet
Why LeCun raised over $1B to argue the future of AI is learned world models (JEPA) — not bigger generative LLMs.
ReadAI FrontierTest-time compute: why AI now thinks before it answers
The next scaling frontier isn't bigger models — it's letting them reason longer at inference. What that means for local AI.
ReadExplainerHow AI evolved: from CNNs to modern LLMs & VLA
An interactive timeline of AI's journey from the discriminative era into the generative era powering today's tools.
Read