Clouds, autocomplete and tweets

I’ve been experimenting with re-publishing a few of my recent columns on Medium. If you’ve arrived here in hope of reading them, simply follow (and share!) the links below.

I am the algorithm – on language, thought and ten more years of Twitter

Is autocomplete evil? – how a machine’s whispers in my ears are changing the way I think

The tyranny of the cloud – why you should think twice before you hit “upload”

Why computers will become invisible – our extraordinary intimacy with unseen technology

1 Comment


Leave a Reply

  1. The issue, which is to say The Issue, is that data fusion is far more powerful than believed by, as far as I can tell, all but all commentators. Speaking from inside the puzzle palace, you’ve no idea — said not as a brag but rather as a kick in the shins if not teeth. Radio stations were once valued by Sarnoff’s Law, viz., that the measure was the number of listeners (N). The Internet is often said to follow Metcalfe’s Law, viz., that the measure is the number of possible conversations (N^2). If you will for a moment call a social network a body of data, then that follows Reed’s Law, viz., the number of subsets that can be defined (2^N). Ignore the math if you like, but the point of points is that every additional datum that is added does not contribute to a linear sum of all data yet revealed, rather it is an exponent in a power series, i.e., each additional datum is not less powerful than the one before but far more powerful. Again, you have no idea, nor will you if what you need is “transparency” before you make a decision whether to reveal all to your selection of counterparties. Privacy of data already revealed in some venue because the holders thereof can be trusted (and you do)? That isn’t trust, it is Pollyana.

Leave a Reply

Your email address will not be published. Required fields are marked *