Topics
All Categories
35 topics across AI foundations, models and labs, building with AI, and applied AI.
Foundations
Artificial Intelligence
The core concepts behind modern AI — what these systems can genuinely do, how they are built and where the real limits sit.
AI News
Model launches, research releases, funding rounds and policy shifts — the AI stories that actually change what you can build.
LLMs
How large language models work — tokenisation, context windows, attention, training and the behaviours that follow from the architecture.
Machine Learning
Machine learning fundamentals — training, evaluation, overfitting, embeddings and the maths that underpins every model you use.
Generative AI
Generative systems across text, image, audio and video — how they produce output and what separates a demo from a product.
Prompt Engineering
Prompt design that survives contact with production — structure, examples, evaluation and the patterns that reliably improve output.
AI Ethics
Bias, transparency, consent, labour and accountability — the governance questions that arrive with every AI deployment.
Models & Labs
OpenAI
OpenAI models, APIs and product decisions — GPT releases, tooling changes and what they mean for people building on the platform.
Anthropic
Anthropic research, the Claude model family and the safety-first design choices that shape how these models behave.
ChatGPT
ChatGPT features, limitations and practical workflows — getting consistently useful results out of the most widely used AI product.
Claude
Claude models in practice — long-context work, tool use, coding performance and where Claude outperforms the alternatives.
Gemini
Google Gemini across the model family — multimodal capability, context handling and integration with the wider Google stack.
Google AI
Google DeepMind research, Vertex AI, and how Google ships AI across Search, Workspace and Cloud.
Perplexity
Perplexity and AI-native search — citation quality, research workflows and how answer engines change information discovery.
Meta AI
Meta AI research and the Llama family — open-weight releases and their effect on the wider model ecosystem.
Mistral
Mistral models and the European open-weight scene — efficient architectures, mixture-of-experts and self-hosting economics.
DeepSeek
DeepSeek releases and the reasoning-model wave — training efficiency, benchmark results and open-weight availability.
Open Source LLMs
Open-weight models you can actually run — licensing reality, hardware requirements and where they close the gap on frontier models.
Building
AI Agents
Agentic systems that plan, call tools and act — architectures, failure modes and what it takes to make them reliable.
MCP
The Model Context Protocol — how tools, resources and prompts are exposed to AI clients through one open standard.
AI Coding
AI-assisted software development — where these tools genuinely accelerate work, and the review discipline they demand.
Cursor
Cursor as an AI-first editor — codebase indexing, agent mode, rules files and how it fits an existing workflow.
GitHub Copilot
GitHub Copilot across completions, chat and agents — enterprise controls, model choice and measurable developer impact.
Windsurf
Windsurf and its agentic editing model — Cascade flows, context handling and where it diverges from other AI editors.
Claude Code
Claude Code in the terminal — agentic coding, subagents, hooks, MCP servers and repeatable engineering workflows.
RAG
Retrieval-augmented generation — chunking, embeddings, reranking, evaluation and why most RAG systems fail on retrieval, not generation.
Vector Databases
Vector search infrastructure — index types, hybrid search, filtering, scaling and choosing between managed and embedded options.
AI APIs
Working with model APIs — authentication, streaming, tool calling, caching, rate limits and controlling token spend.
Applied AI
Automation
AI-driven automation for real operations — where to apply it, how to keep a human in the loop and what it actually saves.
No-Code AI
Building AI workflows without writing code — visual builders, integration platforms and the ceiling you eventually hit.
AI Security
Securing AI systems — prompt injection, data exfiltration, model supply chain and the controls that limit agentic blast radius.
Image Generation
AI image models in practice — prompt control, consistency, editing workflows, licensing and production-quality output.
Video Generation
AI video generation — model capabilities, shot control, cost per second and where it fits a real production pipeline.
Voice AI
Speech synthesis, transcription and real-time voice agents — latency budgets, cloning ethics and deployment architecture.
Robotics
Embodied AI and robotics — vision-language-action models, simulation-to-real transfer and where physical AI genuinely stands.