23/09/2026
Ever felt like generative AI gives you brilliant answers one day, and total nonsense the next? 🤷♀️
The difference rarely comes down to the tool—it comes down to the prompt.
Large language models don't read sentences like humans do; they calculate the next most likely chunk of text based on probability. If your brief is vague, the statistical guesswork kicks in.
I’ve broken down how to take control of tools like ChatGPT, Claude, Gemini, and Copilot—using practical techniques like few-shot framing, clean delimiters, and diagnostic prompting so you get dependable results every single time.
Swipe through for a few core takeaways, then read my guide: https://carolinehagan.co.uk/introduction-to-prompt-engineering-for-generative-ai/
