How to choose an AI tool: a beginner’s guide to ChatGPT, Gemini, Claude, Perplexity and Copilot
Compare AI tools by task, tokens, quality, sources and cost, with everyday examples and documented business uses.
Read article
Fluent AI HK / Articles
Help every reader see AI from different perspectives.
Compare AI tools by task, tokens, quality, sources and cost, with everyday examples and documented business uses.
Read articleAnthropic has shared a case of a government team using Claude Code to document legacy systems, alongside figures from a developer pilot. Explore the practical uses, distinguish completing setup, self-reported time savings and correct results, then learn to design a small exercise whose results you can check.
Read articleGoogle’s introduction to Genie 4 describes a model that lets users adjust simulated scenes with text in real time while generating sound alongside them. What could this “world model” be used for? Explore a creative example, evidence-checking exercises and access conditions to understand why a coherent scene still does not establish that its content is real.
Read articleGoogle has announced that Gemini can turn spreadsheets, receipts and purchase records into financial charts and interactive models. What does this mean? Use a simple exercise to distinguish data recognition, calculation and business judgement, and learn how to check AI analysis.
Read articleAgent Skills organise work instructions, reference material and any required scripts in a folder to help AI handle specific tasks. A handover example explains the difference between providing information and specifying a method, followed by practice checking the process, missing information and results.
Read articleTurboFieldfare lets a Mac run Gemma 4 locally, but about 2GB of runtime memory, 8GB of total computer memory and 14.3GB of model storage measure different things. Use specific figures, the author’s measurements and a workplace exercise to distinguish capabilities, conditions and inferences.
Read articleAn AI refusing a task does not necessarily mean it lacks the ability; answering does not mean the result is correct. Use Claude’s security access arrangements and test figures to distinguish capability, permission and reliability, and assess which conclusions third-party evaluations can support.
Read articleAn AI remembering your preferences does not mean its model has been retrained—or that it has permission to access websites or make payments. Using Google’s ADK documentation, understand four different mechanisms, then use a workplace example and a short exercise to assess what a tool actually does.
Read articleHow could a fictional Hong Kong coffee company use Gemini 4 Argon, Grok Bot and ChatGPT dots to promote an office coffee subscription? Compare audience research, campaign planning and draft marketing materials using the same brief.
Read articleA practical guide to scripting, visuals, AI motion, voice-over and editing, with a 150-second product video budget and commercial-use checks.
Read articleReady to practise? Start with our free 12 chapter course.
Course ↗