Writing
Articles on modern AI
Plain-language explainers on how AI systems work, what the terms mean and where the limits are.
Can AI Really Forget What It Learned?Deleting a file, stopping an answer, and removing the influence of training data are three different tasks.
Can We Tell Where an Image Came From?Content Credentials can help us inspect an image's recorded history. Understanding what the image proves still takes judgment.
Does AI Need the Cloud?What can run on your own device, and why the best answer depends on the task.
What Happens When AI Learns from AI-Generated Data?Why synthetic data can be useful, why recursive training can go wrong, and why the recipe matters.
When AI Agents Talk to Each OtherA simple guide to agent-to-agent interaction, the A2A protocol, and what makes collaboration useful.
World Models: How AI Predicts What Happens NextAn AI system can describe a scene. Helping it anticipate the consequences of an action is a different challenge.
A Practical Map of Modern AI: What All Those AI Terms Actually MeanScalable, modular, agentic, federated, sovereign… A connected guide to 17 terms, and the different questions they answer.