The introduction of "Mobile Usability: How Nokia Changed the Face of the Mobile Phone" by Lindholm, Keinonen and Kiljander features this interesting excerpt:

"the only way to get a working assumption of what the technology enable us to do and how they are likely to be used is to be involved in these projects long enough. Even then, educated guesses and developed intuitions are only approximate. Something that was supposed to be easy to implement turns out to be practically impossible. Sometimes, the opposite occurs. Solutions that were originally postponed to allow technology to catch up are suddenly realized in unexpected ways"

Why do I blog this? This quote is an interesting summary of what I believe as it covers different aspects:

The importance to have a long-term involvement in an organization which design something: I personally work with a french video game studio for 7 years and it strikes me how much I learn in the long run and not through short gigs on their projects. For example, it's been almost from the beginning that we discuss the usability test and user experience field study ideas. It took us approximately 4 years to turn what was "user research as a R&D project" into "user research in the production pipe-line". The time to convince people, to show the value of user research, the importance to insert it in the production process, and finally to get some funding to make it accepted...
The notion of "educated guess" and "developed intuitions" is important. For that matter, I like how Jan Chipchase frame the results form his work: not facts but "informed opinions". Although the quote does not refer to user research, I find an interesting pattern here in the sense that knowledge construction about the evolution of technology is rarely absolute. There are contingencies and idiosyncrasies that plays an important role.
The difficulty in forecasting results because the world is a complex system.
The importance of time: innovation is slow, change takes time and as foresight researchers say, we always tend to overestimate the short term and minimize the long term (tail).