When a database of tagged images goes awry (sorta)
This morning I was browsing the Telegraph website (Britain’s No.1 quality newspaper website–or so says their tagline) and noticed a curious link at the bottom of the page. It was promoting some new online novel called ‘Corduroy Mansions’. The author credit was Alexander McCall Smith. The image however was of The Cure’s Robert Smith.
For a quick moment I thought, “Hey look, Robert Smith wrote a book. But what an odd pseudonym. Ya know, to only change his first name, add a middle but keep the surname.” Of course the ‘duh’ moment came quickly as this was obviously an error. The more websites rely on image tags or keywords rather than file names to search for their images, the more this will happen.
The lesson here is that, no matter how we try to pretend that our machines are so smart, they’re not. They’re stupid hunks of plastic, wires and such. They rely on us to be the smart ones (a losing battle at times, I know). No amount of programming can replace the need for people to pay attention.
I’m just poking a little playful tease here because the fine folks at the Telegraph just happened to be the ones to catch my attention. These kinds of mistakes are inevitable and happen to all of us who are blessed (or cursed) with churning out content upon content, day upon day.
See Also:
Telegraph
