> <i>Adds Auguste: “We can use technology to do more than automate tasks. We can use it to accelerate learning, optimize talent, and guide people into better jobs and careers.”</i><p>IMO, this phrasing is slightly misleading...the primary use of computational technology (well, most machinery, actually) is to automate things. And we should not forget that, and we should use that to the best of our advantage.<p>What use of computers <i>doesn't</i> involve some kind of automation at its heart? One of the earliest private sector uses of computers was to arrange airline seating; something that was such a monumental and tedious task to do by hand (and by phone) that the airlines that won the battle for building a computerized system had a huge advantage over their rivals. Google is basically the automation of counting links (and other "signals") and evaluations of the resulting graph. Facebook started out by automating the connection of faces and biographical information to identities of your fellow students and then later, your entire social network -- in a fashion much more efficient and navigable than a comparable rolodex. Its newsfeed is an automated collection and curation of timestamped events and posts and profile diffs...again, something you could've done manually on Myspace by visiting all of your friends' pages and manually noting down updated info...and this is partially the reason why Facebook is huge and Myspace is largely dead. And the latest crop of startups, like Uber: the automation of connecting driver to passenger, including handling the financial transaction and geocoordination of a trip.<p>I think thinking of computational technology as nothing more as automation (this includes the design of intellectual tasks so that they can be offloaded to a machine, even something as simple as a todo list or the sort/filter function of a spreadsheet) is perfectly fine, because that means humans still wield the ability to <i>think</i> of the ways to apply computational work and how to scale and control it...something that we're a long ways of automating at a higher level (i.e. SkyNet).<p>It never fails to surprise me when teaching people something as basic as routine web scraping, of something like Craigslist listings. Technically, it's just a loop and some string parsing and HTTP requests...but the number of ways that someone can make that task <i>their own</i> -- such as deciding how many pages to go, or how frequently to do the search, or the ways to join the results of different search terms, or, later, how to automate the joining of one data source to another -- even simple programming assignments can have a near infinite number of permutations personal to the individual programmer.<p>In the OP's quote, all of those things that he mentions as being not-just-automation: "accelerate learning" and "optimize talent" -- those all have automation at their core...such as the delivery of content via Khan Academy, or the automated grading of Codecademy. And using technology "to guide people into better jobs and careers"...that's a job board...and Craigslist would never have become successful if Craig Newmark, instead of eventually writing the necessary automation programs, continued to hand-collate interesting links by hand to email his friends.