All this talk about "outsourcing our memory to machines" sounds like a load of zero-sum thinking. It's not that we're shrinking our biological memories and pushing our memories into machines; instead, we're using the internet as another level of the memory hierarchy, allowing us to work with more information, faster. Why would using the web make it harder for us to store information in our biological memory? If anything, it lets us choose to memorize just the important parts, because we can look up trivial details easily.<p>To use a computer analogy, you could think of our short-term memory as CPU cache, our long-term memory as DRAM, and books as floppy disks. The internet, then, can be compared to a hard drive: much faster than floppies, with a lot more capacity per dollar. (I wonder what would correspond to solid state drives in this analogy. If you can figure out what it is, you can probably make a lot of money by building it. If Google doesn't beat you to it.)<p>A meta-criticism I have of this article is that it doesn't talk about actually measuring the supposed hurtful effects of the internet on creativity. I know this is a hard thing to measure objectively, but there's got to be <i>some</i> way to elevate the discussion above the level of bluster-laden hand-waving about the Scary New Thing. What testable predictions does the article's thesis make? If anybody has ideas here, I'd love to hear them.