Can anyone tell me why the new automated announcement system at Kings Cross-St Pancras Underground station sounds like a female Stephen Hawking?

Of course I understand that it’s a lot cooler to have a computer generate alerts as the need arises, rather than imprisoning Mrs. Flibbledy-bit from Dagenham in a small room and forcing her to record all the possible syllables one might possibly need while traversing the bowels of London. We all know where the latter leads: the sheer desperate boredom that causes the tannoy to squeal out that your train will be terminating at “Wiiiiiiiilllllsden Green!”

But surely the technology has moved on: why can’t these systems get the vowels right? Why can’t the computer be taught that the end of a sentence involves a falling intonation, or that sometimes you need a pause between words? Until it’s sorted, I say bring back Flibbledy-bit et al. She may sound a bit stilted, but at least she doesn’t resemble a circa-1970’s science fiction film.