For some reason, the Friedman book linked in Tribe.net: Tribe Discussion: Alternative Money and Economics stimulated this outpouring from me.
I think the Blairite, education first policy is totally busted.
It basically offers the following diagnostic : "Sure business isn't very responsible to the people who work for it. But that's really the fault of the workers. Business has real and justifiable *needs*.
Sadly, the workers don't fulfil those needs right now. But if they only take responsibility for educating themselves better, and are willing to learn and adapt faster, then they'll match the requirements of business, and things will be fine.
Now, we understand, that maybe the workers require some help to achieve this level of flexibility and productivity. Perhaps they were let down by previous education regimes. But now we have the technique and the will to make that OK, and help people out of their predicament."
But is there a natural *limit* to the increasing demands that business will make on the worker? And is that limit lower than the human breaking point?
There's almost no evidence that business regulates itself to respect the idea of natural human limitations it won't try to go beyond.
Will business put workers lives at risk (and stochastically kill them) to save money on safety features? Sure, does it all the time.
Will it employ 8 year old kids in factories? Yep.
Will business employ teenage girls 15 or 16 hours a day without break? Check!
Will it feed them amphetamins to keep them awake beyond their natural sleep patterns? Happened in Guatemala. 
Again, what time-scales does business respect? Here we have a great example from international currency trading. Theoretically, currencies are bought and sold to represent the currency markets' faith in the underlying economy of countries. In fact, as we all know, it's speculative gambling. The dealer in New York who's about to take a punt on the Brazilian "Real" isn't waiting to see if the Brazilian farmer can get up to speed on new technologies before he makes his assessment.
Business has NO sense of time-scale except to try to compress it. And now we're seeing wonderful deeds in supply chain management. Dell put's a computer together to order. It can switch most of it's component suppliers in an afternoon.
The quote that an engineer's knowledge is out of date in three years is interesting. But that may soon look sluggish. If it suited business that product lines evolved so fast that the engineer's knowledge was out of date in 3 weeks, or that suppliers changed several times an hour, that's the way it would be.
Now how is the educated worker protected from this acceleration? Not at all. In fact, in some ways, she's *worse* off than her less educated sister.
1) If the job lasts the same length of time, she's made a bigger investment in her skills for the same return.
2) If the job involves more information processing, then it can probably be moved around the world, really easily.
Basically, the Blairite line is "give business what it says it wants today, and hopefully it won't want more tomorrow." But that's looking increasingly implausible.
 According to Naomi Klein