What Continual Learning Actually Is →
How models, agents, and agentic systems turn operating experience into persistent, verified improvement, with the challenges of generalization, adaptation, safety, and shared learning.
White Papers
Ideas explored in depth.
Published writing on AI, networking, and neuroscience. Earlier work remains available in its original context.
AI & machine learning
Compressing the World's Knowledge
A paper arguing that large foundation models should be understood as compact parametric stores of world knowledge, with major implications for retrieval, infrastructure, and local-device intelligence.
Don't Imitate, Innovate: The Case for Useful AI
A case for judging AI by usefulness, contribution, and problem-solving impact rather than by how closely it imitates human cognition.
Why AI Hallucination (Confabulation) is More Solvable Than Our Own
A technical argument that machine confabulation is a real issue, but one that can be engineered down more systematically than many human cognitive biases.
The Power-Hungry AI Myth
A paper arguing that current AI has no biological drive for control and that fears of autonomous domination misunderstand the architecture of modern AI systems.
AI for networking
Beyond Protocols: Why AI is Networking's Overdue Paradigm Shift
A paper arguing that networking's biggest opportunity is no longer another isolated protocol improvement, but a shift toward intelligent, distributed, AI-driven systems.
Time to Revisit the Internet Layering Principle: AI-Driven Cross-Layer Optimization
A case for augmenting strict layer isolation with AI-driven cross-layer intelligence trained on multi-layer telemetry and user experience signals.
Challenging the Norm: A Groundbreaking Experiment in AI-Driven Quality of Experience (QoE) with Cognitive Networks
A paper on using cognitive networking techniques to optimize quality of experience directly rather than relying only on lower-level network metrics.
Predictive Networks: Networks that Learn, Predict and Plan
The updated Predictive Networks white paper describing how networking systems can learn, forecast, and proactively steer around failures and degraded paths.
Neuroscience
AI and Neuroscience: the (incredible) promise of tomorrow
Exploring biological and artificial intelligence.