http://www.zdnet.com/amazon-shows-interest-in-texas-instruments-mobile-chip-division-7000005755/?s_cid=e539
and:
http://thenextweb.com/insider/2012/10/15/amazon-in-talks-to-buy-texas-instruments-smartphone-tablet-omap-processor-activity-report/

TI recently announced that it was exiting the consumer focused microprocessor business to focus on embedded applications. A big part of that was the OMAP: variants of which power the KF and KFHD. Amazon could have turned to one of the other ARM purveyors, but it looks like they are exploring this option, which is not unlike what Apple does with its custom ARM designs for the iPhone, iPad, etc.

So if this story bears fruit, then maybe Amazon WILL get into the H/W business alongside existing content and service businesses.

More generally, with regard to other competitors, I think Bezos wants the Kindle to be #1, a subset of his desire to make Amazon #1. He feels it would be strategically dangerous to let the Kindle be eclipsed, even without knowing what form the threat might take. That’s already proved true with Apple’s organizing the agency model against him, in order to make their iPad #1. What might Google do with its Nexus, once it finally settles with publishers and can start selling its millions of out-of-print books at its own bookstore? What might MSFT do with its Surface, if it reaches a deal with B&N and/or Google or European publishers? Etc.

The thing is: I think Bezos has only ever been interested in the content side of things: the blades not the razor. He got into the device business initially to prime the pump for ebooks. When Apple came out with the iPad, he was pretty clear he did not want to get into competition with them. From day 1 I think the kindles have always been sold at cost. What he told investors from time to time may have been something different, however. I think he would have been comfortable letting others develop EBRs, but things like agency, ePub, DMCA, DRM, etc got in the way.

In a sense we are all still dealing with the shockwaves of what digital and internet technologies did to the music business. Media publishers ( of all kinds in the broadest sense) are now quite paranoid.

The other thing to keep in mind is that those up front engineering and development costs are always amortized over some number of forecasted unit sales. If more units than forecast are sold, then the device becomes profitable. Also the manufacturing costs of electronic devices tends to fall off rapidly as component cost reductions, economies of scale, and manufacturing learning curve efficiencies come into play.

So when Bezos says he is selling devices at cost on day 1, that doesn’t necessarily mean he will be doing so at end of product life. I suspect the unit profit on an original KF is higher now than it was at introduction. If you look back over the history of the eink kindles, there were several times along the way where Amazon chose to reduce the price (often in response to BN pricing moves), which suggests that they had some margin headroom.

I think Amazon has been pretty consistent with their vision of being a retailer and service provider, and not a hardware manufacturer. They have always consistently tried to provide apps and tools to access their retail and AWS ecosystem from a wide variety of H/W platforms. I think their H/W developments are somewhat analogous to what Microsoft is doing with their Surface tablets: providing a superior product target that their OEM partners can strive towards.

There have been (as you say) some tune-ups along the way to deal with responses from competitors like the big6, BN, and Apple, but I don’t think their corporate vision has changed all that much.