This is actually a question I’m asking myself at the moment as well. I taught older kids (8-16) robotics for a while, and found the younger they were the more

This is actually a question I’m asking myself at the moment as well. I taught older kids (8-16) robotics for a while, and found the younger they were the more harmful failure scenarios were. If they failed, they would get frustrated and gradually disinterested. The more there was a defined success criteria that could be measurably achieved (or not achieved) the more obvious it was to them that they had no succeeded. The more it was just about play the better. Ideally failure states should be learning opportunities and kinda fun. Littlecodr is awesome for that. It’s just a set of cards with instructions like move forward, turn left etc. get him to program you. It’s funny when the instructions lead you into a wall instead of between the door frame. Failure is fund Some of the best stuff is also just dumb hands on stuff. Playing with cardboard glue and string. Making spaghetti towers. Folding paper planes. Magnets are awesome. Basically they learn from you, so the more time you spend together the better. If your looking for low energy investment check out primo (primotoys.com). It’s little wheeled bot you program with simple physical tokens. Osmo is also cool but I find the iPad platform derails their attention. Littlebits is also cool, but the parts are a bit small for unsupervised play at that age. Makeymakey is also rad. Adafruits circuit python is awesome as well, but the play sessions would have to be quite curated. You could preprepare stuff and just do the wiring together with crocodile clips etc, but in all honesty maybe just a battery would be fine ;) Looking forward to what pops up here :)