Have you tried running your assessment assignments with AI? How did it go? If it failed, well done – you can stick to your ‘tried-and-tested’ methods (did you try setting the temperature to ‘0’? Just asking!)). If it didn’t, then it signals the need for some consideration.
Consider whether the assignment remains relevant and reliable, or if it’s compromised. You might need to ponder how to tweak it to strengthen its ‘human touch’ component and make it more personalised and contextualised (though, let’s face it, this won’t make your assignment 100% AI-proof – nothing can). Alternatively, you can fall back on the good old ‘I’m watching you’ testing environment, explore a mix of alternative assessment tools, or try to introduce a balanced integration of human and AI components (if you can’t avoid it, make the most of it). Here’s a simple decision tree that could be a good starting point.
[…] Assessment and AI […]