August 29, 2008
Dubious statistic of the decade
In a 2006 white paper, IBM claimed that “just 4 years from now, the world’s information base will be doubling in size every 11 hours.” This week, that statistic was passed on — utterly deadpan — by the Industry Standard and Stephen Arnold. Arnold’s post actually reads as if he takes the figure seriously.
Now, I’ll confess to not having seen the argument in favor of that statistic. But color me skeptical that, by any measure of “information”, it will grow by a factor of more than 2^730 in a year, or 2^7300 in a decade …
Categories: IBM and UIMA
Subscribe to our complete feed!
Comments
5 Responses to “Dubious statistic of the decade”
Leave a Reply
I’m the guy who passed the statistic on deadpan – my editor edits out the rimshots when I joke. 🙂
The fact is that there is never any justification show for these knowledge doubling estimates. My point was that the rate of knowledge is increasing and we’d better find a way to filter information better. The IBM quote was available and just helped to show that the problem is getting bigger, though every 11 hours is probably a bit of a stretch.
Also, the definition of “knowledge” is never clarified. If people actually mean “data” then the output of polling companies alone is probably already doubling every 11 hours.
Thanks for commenting!
Innumeracy is epidemic, even in IBM and the press…
Larry,
Please think about what doubling every 11 hours would entail. It’s an increase by a factor of over 1,000 every five days. Hence, a factor of greater than 1 million every ten days. Greater than one billion every 15 days. Greater than 1 quintillion every month. In under 4 months it would require more mass for storage than exists in the entire universe.
You have my full sympathy about editors who change one’s meaning, but IMO you should have dug in your heels about this one …
Best,
CAM
[…] least the whole line of reasoning isn’t as bad as another one I recently discovered on the subject of information overload. Share: These icons link to social bookmarking sites where […]
As you (Curt) point out, the actual rate as cited is silly. But there’s also the more general issue of extrapolating future exponential growth from previous exponential growth. On its face, such extrapolation also seems silly. Yet Ray Kurzweil, who is highly respected by so many people, seems to believe in that kind of extrapolation so much that he makes it a core principle. (I recently heard him speak near MIT.) Very strange.