Shazam is mighty impressive. I've been playing with it recently, and it's true: in a louded, crowded pub I can hold my phone on my hand for twenty seconds and I'm sms'd with the name and artist of the music, and the album it's from. What's more, it's consistently fast enough to be a part of a conversation, not just a thing on it's own. And for only 50p. So. I'm impressed with the audio-processing technology. But this is only the start, right?
Lie detector. "So you're really not seeing anybody else?" No. [tap tap tap tap tap, put mobile phone on table.] "Sorry, could you say that again, but string it out over 20 seconds?" Blah blah blah. "Sorry, my phone says you're not telling the truth".
How good in bed collaborative filtering. You put your phone on the table and it listens to the chat-up line, you get a predicted 4/10 score back. Sleep with them anyway. Rate them. The information you've provided is fed back in and used to (a) make the score more accurate; (b) provide you with more accurate scores in the future by grouping you with people who enjoyed shags with similar chat-up lines (assuming you enjoyed this one). People who enjoyed shags like your also shagged...
And an always-on phone could operate as a permanent lie detector, giving a discrete buzz when the conversation dropped below a certain truth threshold. Now that would change the world. Especially if it was accompanied by a little Jimmy Nail avatar saying "she's lyin'" every time.