[Dan Russell]1 recently posted this TEDx talk on ‘the revolution in asking & answering questions’ (see [Dan’s blog]2 on it). It’s a fun watch, and interesting to see someone I’ve (briefly) met do this kind of talk – it’s a massive challenge. I really liked what Dan did (so this isn’t a critique at all!) and because of that, I started thinking about how I might do something similar: 1. I’d probably do some audience polling (popular in TED, but also educationally useful – in classrooms I generally used to go for a ‘left arm’ v ‘right arm’ voting method – harder to abstain then 😉 ), maybe something around “who’d rather search for this using offline v online”, or “who has/hasn’t had” some experience, etc. 2. Another nice example here might be giving some examples, and asking for each “Could you use a search engine to ask this?” – this could be done at the start and the end 3. I can imagine building a narrative structure to a cool talk with one side of a slide being ‘traditional’ (books, face-to-face, etc.) and the other web-based information seeking, making the visual contrast, and building a narrative around a particular task. 4. It might even be possible to start something (maybe with an assistant) asking two people on stage to do a particular task, one with a book, the other with a web-enabled device. See how long they take, what they find, etc.etc. 5. I can imagine doing something like the above, asking the audience whether anyone can ‘get this answer by the end?’, but this might just be a massive distraction! 6. One of the things Dan did here, was demo working out sunrise from a particular place, and then introducing a problem (the mountain shadow!) and how we’d address that. I’m not sure this works on scale but this is exactly the sort of search challenge it’s cool to set without context, and then guide students through a) what problems there might be, and b) how we might address them. I love search problems like this :-). 7. I also love the use of multiple language search (I’d probably demo manypedia! ), depending on the audience you can imagine using audience languages to demo some things around this