§The field guide
Understand the category before anyone sells you anything
Definitions, comparisons, costs, and safety — written in plain language from operating experience, with our commercial position disclosed on every page that needs it.
01Foundations
6 min read
What is an AI assistant?
A plain-language definition of AI assistants — what separates them from chatbots and apps, what they can actually do in 2026, and the vocabulary you need to evaluate one.
7 min read
What is an AI employee?
The 'AI employee' pitch decoded: what vendors mean, what's genuinely new, what the salary-like pricing actually buys, and how to tell a real one from a chatbot in a suit.
8 min read
What can an AI assistant actually do?
A concrete, hype-free inventory of what AI assistants genuinely handle in 2026 — organised by how much you can trust the output, with the failure modes stated.
8 min read
How AI assistants work: models, memory, tools, and permissions
The four-layer anatomy of every AI assistant — why the model matters less than you think, and why memory, grounding, tools, and permissions decide whether it's useful or dangerous.
7 min read
How much does an AI assistant cost?
From S$0 chatbots to five-figure 'AI employees' — what drives AI assistant pricing in Singapore, what each tier really buys, and the questions that expose overpriced offerings.
02Comparisons
5 min read
AI assistant vs chatbot: which do you actually need?
Chatbots answer; assistants act. The practical differences — memory, tasks, grounding, cost — and a simple test for which your situation calls for.
6 min read
AI assistant vs AI employee: where the line actually is
Same technology, different contract. How to tell whether you're buying help with your work or delegating a role — and why the price difference is about supervision, not intelligence.
5 min read
Assistant, agent, copilot: sorting out the labels
Three words the industry uses interchangeably, three genuinely different things. A field guide to decoding what a product actually is beneath its label.
03Decisions & safety
8 min read
Is it safe to connect an AI assistant to your email and business systems?
The real risk model for giving an assistant access to email, calendars, and business data — what can actually go wrong, and the safeguards that separate safe deployments from incidents.
6 min read
What you should never delegate to an AI assistant
The boundaries that don't move as models improve: where delegation fails for reasons of accountability, relationships, and risk — not capability.
7 min read
Build, buy, or hire: three ways to get an AI assistant working for you
Subscribe to a product, commission a custom deployment, or have someone build it in-house — the real costs and failure modes of each path, and a sequence that avoids the expensive mistakes.
++In preparation
- Best AI assistants in Singapore: a living comparison (after our first benchmark run)
- ChatGPT vs Claude vs Gemini as AI assistants (after our first benchmark run)