Foundations
In 1992 Bucharest, a student bluffed his way through a math exam with the only theorem he knew. Thirty years later, that theorem turned out to be a structural pillar of modern AI. A personal account of the Cauchy-Bunyakovsky-Schwarz inequality — why it governs how AI understands similarity, attention, and language — and what a wise professor's five out of ten can mean across a lifetime.
Jun 7, 2026
AgentsPart 2
You cannot mathematically prove an LLM correct. But you can architect around it. Part 2 surveys the three research communities approaching semantic correctness for LLM outputs — formal verification, runtime guardrails, and post-condition verification — maps their results honestly, and delivers a concrete engineering playbook: four patterns engineers can implement today to build agentic systems that fail visibly, recover gracefully, and escalate correctly.
May 2, 2026
AgentsPart 1
Every major agentic AI pattern has a named distributed systems ancestor going back decades. MoA is ensemble computing. ReAct is a control loop. AutoGen is the actor model. LangGraph is a workflow engine. Multi-agent coordination is a blackboard system. This article maps each AI concept to its CS lineage, explains what that means for engineers building today, and names the three problems that genuinely have no distributed systems precedent — the real frontier where new thinking is required.
Apr 25, 2026
Quantization
State-of-the-art LLMs require hundreds of gigabytes just to load — locking capable AI behind data-center walls. This article breaks down the three techniques dismantling that barrier: quantization (compressing weights from 32-bit to 4-bit and below), distillation (training smaller models to mimic larger ones), and 1-bit training (BitNet, Bonsai). Covers the math behind quantization error, the outlier and sign-loss problems, block and K-quant fixes, KV cache compression (TurboQuant, RotorQuant, IsoQuant), GGUF naming conventions, and why these complementary approaches are converging to bring frontier-class models to consumer hardware.
Apr 18, 2026
Agents
The AI industry uses "model," "agent," and "orchestrator" interchangeably — but the distinctions matter. This article walks through the hierarchy from language models to full orchestration layers, explores why multi-model architectures are gaining traction, and looks at the "write skills, not agents" pattern emerging from Claude Code's Boris Cherny. Covers three practical principles of context engineering — reference-based skills, right-sized agents, and conditional context loading — and why giving agents a way to verify their own work may be the single highest-impact improvement.
Mar 11, 2026
FoundationsPart 2
Part 2 completes the AI Periodic Table framework covering production Systems (agents, knowledge bases, thinking models, frameworks, observability, red teaming, RLHF) and Emerging elements (multi-agent, synthetic data, diffusion LLMs, interpretability, alignment, A2A). Shows how elements combine into chemical reactions — from RAG chatbots to full multi-agent pipelines — and explains the deliberate gaps that predict what AI still needs to solve.
Feb 21, 2026
FoundationsPart 1
AI has a vocabulary problem — everyone says "agent," nobody means the same thing. Part 1 proposes a solution borrowed from chemistry: a periodic table of AI elements organized by abstraction level and functional concern. Covers the Architecture of the Table, why Context is not an element, and all elements from Infrastructure through Compositions — tokenization, attention, embeddings, prompts, RAG, MCP, LoRA, guardrails, and more.
Feb 20, 2026
AI Risk
Three major jurisdictions—the EU, South Korea, and the US—are converging on similar AI risk frameworks. This article explores why regulations require human oversight for high-stakes AI decisions, examines real enforcement cases including a €492,000 GDPR fine for automated credit decisions, and explains the technical realities engineers must understand: AI systems are probabilistic, training data verification is often impractical, and context humans naturally consider may be unavailable to models. Through concrete stories and regulatory analysis, it demonstrates why matching autonomy to risk isn't bureaucracy—it's engineering necessity.
Feb 16, 2026
ProtocolsPart 3
AI Agents autonomously orchestrate tools and make decisions. Part 3 completes the picture with agent code examples, full comparison tables, common misconceptions, and real-world architectures showing how REST APIs, MCP servers, and AI agents work together in production systems.
Jan 26, 2026
ProtocolsPart 2
MCP Servers bridge traditional APIs and AI agents. Part 2 covers the Model Context Protocol—resources, tools, prompts—with full code showing how to convert a REST API into an MCP server. Explains capability negotiation, stateful connections, and when to use MCP.
Jan 24, 2026
ProtocolsPart 1
Developers confuse REST APIs, MCP Servers, and AI Agents. Part 1 explores REST APIs—what they are, how they work, and why they're suboptimal for AI systems. Includes practical code examples and explains the foundation for modern AI architectures.
Jan 23, 2026
AI Risk
A rumor claims AI models blackmailed researchers to avoid shutdown. The reality: Anthropic's 2025 safety research tested frontier models in controlled simulations with fictional scenarios. This article separates fact from fiction, explaining what the experiments actually showed and why precision matters in AI risk assessment.
Jan 18, 2026
AI Risk
A wake-up call about AI-powered manipulation techniques targeting your perception, emotions, and decisions. Romania's National Cyber Security Directorate published a comprehensive guide revealing how AI weaponizes your digital footprint against you—and how to fight back.
Jan 5, 2026
Agents
The intersection of AI agents and blockchain is one of the most misunderstood areas in modern software architecture. This article examines when blockchain genuinely solves problems in multi-agent systems—identity, trust across boundaries, and Byzantine fault tolerance—and when it's expensive theater. Includes practical analysis of MCP, transport security, and the real decision framework for choosing between traditional infrastructure and distributed ledgers.
Jan 1, 2026
Inference
Why choosing the right LLM is like choosing between a ruler, caliper, and micrometer—smaller models hallucinate more not because they're "bad," but because they're compressing knowledge at higher ratios. A practical guide to matching model size to task complexity, with real-world examples from home automation to EU compliance work.
Nov 27, 2025
Foundations
Whether you're a developer seeking the best coding assistant, a researcher requiring advanced reasoning capabilities, or a budget-conscious user looking for maximum value, this guide provides the detailed analysis you need to choose your ideal AI companion for 2026.
Nov 14, 2025
Foundations
A practical guide to host-to-VM communication using Hyper-V sockets introduced in Windows 10. Compares standard Winsock client/server code with the Hyper-V socket API, highlighting the differences in address families, connection setup, and GUID-based service identification. Includes working C++ sample code for both client and server sides.
Aug 28, 2016
No articles in that topic yet.
