Only interface that's likely to surpass keyboard right now, is the neural interface. If it will not fail like the voice recognition technologies.
Voice recognition could work when coupled with touch, but it has reached a plateau http://www.codinghorror.com/blog/2010/0 ... ition.html
Not to mention that PCs will go nuts when they'll start to figure out what "make this detail prettier" in autocad or "make it unboke" in word means. Humans are very bad at expressing what they want to do in logical, exact and definitive terms.
My friends with iPads have them thrown into corners because they are only good for bling and stuff. Once they need to start talking on facebook or anything even remotely serious, they beg to borrow a laptop from the people who have them with them. Because keyboard walks over all the fancy on screen input methods, even with all the auto-correct that turns well meant phrases into something you will be embarrassed later on.
Yes, I use my HTC phone for web browsing on the road, but that is limited with simple auto-complete navigation, super-light web browsing or calls. Once I need to view a document, just view, not edit, or manage files on memory card, I want to throw the device away.
Moreover, glass like, transparent device is not a holographic 3D projector, the projection will be off as soon as you move an inch from the dead center. See what happens with those 3D phones and cameras. And bright light makes the screen unusable unless some super-technology is invented like OLEDx5. The video is fail.
And, the problem with technology is humans, they are lazy to talk very clearly, they have accents, different languages, greasy fingertips and fatigue easily when working with vertical keyboards, or if watching down on a horizontal surface all the time.
Core 2 Duo E6300, MSI P45 NEO-F, Club 3D GTX 260, 4Gb DDR2-800Mhz, Audigy X-Fi Fatal1ty Champ1on ed., 0.5Tb+1Tb Seagate Barracuda 7200.12, 630W AXP, Samsung SyncMaster BX2450, ViewSonic VP171b