John Gruber used the analogy of automobile automatic transmissions for the iPhone/iPad operating system versus the Mac OS and Windows classic operating systems. Given my car purchasing experience recently, I really resent that analogy. The "old world" versus "new world" naming in Steven Frank's article, while still obfuscated to some extent, is a lot closer to the truth.
My main issue with using car transmissions as the analogy for computing experiences is that the automatic transmission in its current form is the pinnacle of marketing, not technology: it is the end of a long line of technological band-aids to make it possible to operate an inherently complex and dangerous apparatus (the automobile) with the minimum of skill possible. The average automatic transmission makes it possible to drive a car in some fashion with relative ease and efficiency -- it does not make it easy (or possible, in some cases) to drive it well. There are counter-examples to this, of course, and I would argue that they are more representative of what Apple is doing with the iPhone/iPad than the general "automatic transmission" category as a whole. For example, the BorgWagner DSG transmission used by VW and Audi (soon others like Porsche) removes the need to skillfully shift gears but it keeps the option for the driver to have as much (or little) control over gear selection as they want. Good automation removes drudgery without continually forcing inferior choices on the user.
Going back to Apple and these new devices, the thing to remember about the company is that, while secretive, evil, etc. they do understand that success comes from offering fewer, better choices. If it were a transmission, it would be automated -- yes -- but it would not be any "automatic transmission." The genius of Apple in these "new world" devices is that they offer few choices that generally excel at what they do -- and don't offer choices where they don't excel.
The automotive world has clearly proven that you don't need to be that good to succeed in the long run, but I don't think that mere success is Steve Jobs' dream here. I think it is all about setting the bar, defining what computing should be like for common users and, at least for the short term, owning this brand new market lock stock and barrel.
Of course, if they succeed, one thing will be exactly like the "automatic transmission" analogy: the general-purpose computer will become as marginalized as the manual transmission is on cars in the US these days (which, as a software developer, worries me: I don't like the idea that the tools of the trade could become once again major investments). However, the default choice is likely to suck a lot less than it does for cars.
