It’s able to do such things not only because you told it to, but also because it has access to that imponderable store of data generated by the online activity of you and your friends over the past years. It’s early days yet, but the intention is that such IPAs (as I’m going to call them) will eventually supplant the web browser and the search engine as your interface to the Internet. They will manifest on your phone, your tablet and your desktop, and they will become increasingly indispensable as they, and the data that drives them, learn more about who you are, what you like, and what you believe.

It is early days, and the technology still has a way to go. When Apple’s Siri first appeared, for example, it was accused of having a ‘pro-life’ bias because, when asked to find an abortion clinic in Manhattan, it responded: “Sorry, I couldn’t find any abortion clinics.” It also exhibited a tendency to direct people towards Crisis Pregnancy Centers which were set up by pro-life groups to counsel women against abortion. However Siri’s response has more to do with the limitations of its software, rather than any ethical considerations. Crisis Pregnancy Centers present themselves as standard health clinics so they can attract women who may be thinking of aborting. The popular Planned Parenthood organisation, on the other hand, does offer abortions, but doesn’t describe itself as an ‘abortion clinic’.

Nevertheless there are dangers. The addition of an IPA ‘filter’, for want of a better word, makes it harder for us to judge the veracity of the information we are presented with. We have a semblance of control, but behind the scenes faceless corporations are making countless decisions based on the thousands of terabytes generated by our daily activities. Their agendas may not worry us, but our experiences with the like of Rupert Murdoch, for example, suggest we are naïve to think they do not, or will not, have one.