We all know modern publications use any or all tools at their disposal in order to drive traffic to their websites. I don’t know what it’s like to write for Quartz, but this gaffe is nothing short of embarrassing for Quartz writer Christopher Mims. Let me elaborate.
The linked article, entitled “Intel’s voice recognition will blow Siri out of the water—because it doesn’t use the cloud”, initially exhibited a fundamental lack of understanding of pattern recognition.1 Granted, it’s impossible for one person to have an in-depth knowledge of all the technology utilized by our gadgets.
However, in the update he posted to the article, Mims posted a link to an article he wrote that completely invalidates the title of (and largely the premise of) the piece:
Currently, the everyday magic of understanding voice commands is carried out almost entirely in the cloud, because processing human speech is difficult enough that even a sophisticated smartphone doesn’t have the processing power to do it at a high enough level of reliability.
Ideally, either an editor or Mims himself would have noticed this inconsistency before publishing. That the headline remains unchanged after the update is disappointing. Rather than posting an update and leaving the piece otherwise unchanged, Quartz should take a cue from traditional newspapers and at the very least correct misleading body text and explicitly state the difference between the initial and current article. Specifically, the wording “Intel has a solution.” immediately following the first subheadline needs to be stricken. The tone of the rest of the article should be adjusted as well.
A pattern recognition system relies on either specific or immense training data to deliver accurate results. The Intel solution could depend on recording and processing speech data from each user of its technology (which would then be stored on the user’s mobile device for future reference) to achieve greater accuracy. This possibility is not mentioned in the article.
Instead, the original (and to a large extent, the current) article states that Intel has somehow magically solved the same problem currently addressed by racks of servers with a chip that can fit in your smartphone. ↩