Unfortunately, I would imagine that UCAV's would replace pure fighters first...some of the modern SAMs and air to air missiles can perform maneuvers that pull high-teen G's (The Russian SA-10 travels at six times the speed of sound and maneuvering at 100 - one hundred - g's) while maintaining structural integrity, whereas human pilots obviously cannot undergo such physical stress. And while our top-of-the-line aircraft could potentially escape unscathed after such acceleration forces, the pilot would be pretty far into a blackout (probably death) by the time he pulled through the first avoidance maneuver.
Even disregarding this though, it seems to me we are pretty close to a pilotless cockpit when the F/A-18 does carrier launches and landings virtually hands-off (or at least has the proven capability to). It seems to me that modern fighters have pilots for the mere appearance of a human face in air combat.
Clearly, there is a place for pilots in CAS (think A-10) when split-second decisions have to be made, but even here, how many of us have played fighter/attack computer games where we strafe troops, destroy tanks and shoot bombs through windows? I would bet many of us...The combination of Wi-Fi and aircraft capabilities seems to inevitably trending towards more automation for every aspect of air warfare.
If we want to be a capabilities based military, clearly UCAVs are the way to go in terms of cost savings, efficiency and battlefield effectiveness. and I say this as one who has dreamed of becoming a naval aviator for years. The hard reality for me is that while I would personally hope aviators could still fly, it makes more sense from a strategic standpoint to build stealthy, highly-maneuverable UAV's that can avoid the air defences of our most likely enemies. Plus, you dont have to spend $1.5 million training a 22 year old to fly when you just pop him in a simulator for every training hop.
This is the wave of the future, unless you want helo's real bad ;-)