Jean-Louis Gassée Says Human Curation Is Back
Joyce Park stashed this in Tech biz
Only human can select content AND PROVIDE CONTEXT for other human!
Yes! And this is so well written!
With search engines, we see a different kind of curator: algorithms. Indefatigable, capable of sifting through literally unimaginable amounts of data, algorithms have been proffered as an inexpensive, comprehensive, and impartial way to curate news, music, video — essentially everything.
The inexpensive part has proved to be accurate; comprehensive and impartial less so. No matter what their proponents (and sellers) say, algorithms aren’t intelligent. They’re dressed up in rich-sounding names and euphemisms such as “Machine Learning”, or oxymorons such as “Affective Computing”, but however they’re decorated, algorithms don’t understand meaning.
Certainly, algorithms can be built to perform specialized feats of intelligence such as beating a world-class chess player or winning at Jeopardy. Like many, I have personal reasons to hope for the continued development of algorithms that sift through huge amounts of information (a.k.a. Big Data) to pave the road to cures for DNA anomalies. And on an even more selfish note, the business of algorithms — the tech world — has fed me and my family quite well for the last half century.
But ask a computer scientist for the meaning of meaning, for an algorithm that can extract the meaning of a sentence and you will either elicit a blank look, or an obfuscating discourse that, in fact, boils down to a set of rules, of heuristics, that yield an acceptable approximation. As times goes by, the rule book gets thicker, but meaning remains elusive. Google Translate, an algorithm that many consider to be the most prominent Machine Learning engine on the planet, stumbles on a simple sentence such as “les poules couvent au couvent” (“hens hatch in the convent”), tripped up by a grade school word equivalence.
I like the observation that Apps are bigger than Hollywood but Humans can curate music better:
But should we be shocked when we hear that search algorithms aren’t fair and balanced? Algorithms are designed and built by humans and they reflect the biases of their makers. To paraphrase an old adage, It’s Humans All The Way Down. Humans are born with the gift of guessing, of divining the rules of the game — that’s how we learn speech. This Gift From The Genes allows us to smell manipulation behind the pretense of impartiality, hence the disenchantment and possible legal actions when we find that the sanctimonious representations of fairness are clumsy fig leaves covering human shenanigans.
As the limitations and misrepresentations of algorithmic curation become more obvious, the marketplace turns back to humans. As Business Insider’s analysis of the Apple Music announcement put it:
“Most online music services rely on computer programs to recommend songs and build playlists. But Apple is placing a big bet on human editors with a strong knowledge and love of music. The idea is that these human editors, like the radio DJs of yesteryear, will help turn Apple Music into a great way to discover new music.”
It’s too early to judge the success of the new Music service, I’ll need a while to get to my Third Impression, to see how the novelty wears off, or how the confusion clears up. But I’m encouraged to see a more human touch openly applied to the curation of an art form.
With this in mind, let’s look at one of Horace Dediu’s graphs from earlier this year where he shows apps to be Bigger than Hollywood:
For a while now, music downloads have paled when compared to apps – hence Apple’s move to a streaming service. But there’s another idea lurking in there: If it’s a good idea to use human curators to navigate 30 million “songs”, how about applying human curation to help the customer find his or her way through the 1.5M apps in the Apple App Store? Apple bought Beats for $3B and spent a good chunk more to build its Music product. Why not take another look at the App Store jungle and make customers and developers even happier?