I like another responder, worked on PLATO in the 1970's. I worked and played on that system for over 10,000 hours. I've also been a Smalltalk consultant in Silicon Valley. Last, in Arizona, when its population was 1.25 million about 40 years ago, I placed first in the state in the state math contest which all high school math students in the state took.
My take is a little different from Alan's, but he got most of the important parts right, but he is still going have to do more to win the ball game.
Smalltalk is too difficult. A typical system has 1500 classes and when used on a project, the number of classes tend to bloom. Coming into a new Smalltalk environment can be intimidating. It is like coming into a large city, and as you know, anyone coming into a large city usually concerns him/her self with a very small part of it. I am no dumb bunny, but I would hate to tell you the number of hours I have spent roaming around chasing my tail in Smalltalk environments trying to find out this fact or that relationship. I, and a lot of other people who try, don't have full access or a full grasp to those top notch Smalltalk experts...and I wonder sometimes if they don't fake it some.
Alan is right that we need play for understanding, but we also need discipline, and well managed computer aided discipline and structure in the right context. This can take us far beyond where we usually go in this day and age. What we can achieve in education, industry, and life can be much more than it is today.
I went to India when I was 15, about 45 years ago. And at that time I met kids who were studying Calculus in the 8th grade. Forty years later we are seeing the result.
I am convinced that by carefully taking the right path, that there is no reason we can not reach that level and beyond well beyond by using computers. But certainly not the way we our society is trying to do it now. Why? Because it won't work. I know what testing is about, because I was one of the best. It is not the answer and is a way of doing thing which is stuck in a traditional and limiting past.
Like I said, I agree with Alan about play, but again that is not the whole story. There has to be a sharper focus on where we are going. We must not use computers as we have taught in the classroom, but must make a definite and clear cut departure from the traditional in order to take full advantage of computers. After all, class room teaching developed and is stable in a way which does not take into account the possibilities of computers. How can it possibly be the best way without a complete understanding of this revolutionary consideration?
We must blend what we are good at with what computers are good at. A computer can do things which are extremely difficult for us to do. I have no doubt that a day will come when a computer will keep track of every one of our belongings, and to find that book you misplaced 20 years ago, all you will have to do is ask the computer. We can eventually arrange things so that it can help us track our own educational development, so that if you want to master a new set of skills, it is just a matter of conveniently conveying the information to the computer and immediately having everything you need to do the necessary hard work right at hand, all personally organized for your consideration, with infinite remediation right down to “1 + 1 equals 2” if that is what you lack. To become a master of any intellectual skill, in order to be maximally effective, it must be simple as sitting down and pressing next to begin, with everything including play, exploration ,and the opportunities for initiative right there in front of you ready to go.