It doesn't seem that, in practice, users want to modify their computer systems in that way, despite that that's what early programmers imagined (note that such manipulation of systems had been something their personalities had obviously been drawn to--as is mine and those of most of the people on HN, probably--and also that it was necessary in those days to get interesting results).
Besides, the only way you're going to get anybody but the tinkerers to be interested in modifying the parameters of the thrust calculation in their Spacewar game (I recall how in school we modified the trajectory of the bananas in QBASIC Nibbles) is to not give them anything else interesting on the device that already easily does something interesting and fun... I also recall that there wasn't much else to do on the 386s in my school other than Minesweeper or setting up new groups in Program Manager. Now we've got billions of web pages and thousands of free apps.
Necessity is the mother of invention -- except for when it comes to tinkerers. And I'm super thankful for tinkerers: they push us forward in software just as they do in other fields. But in a world where devices already do way more than their users can imagine (most people very much do NOT have good ideas for how software could be better or do something new) and offer a ton of other easily accessible options for activity, the vast majority of people just aren't a bit interested in tinkering.
Don't confuse core product with actual product. If, before the 20th century, you had asked people if they wanted to partake in mechanical flight, they would have most certainly said no. They couldn't distinguish a core product that would differ significantly from the current actual product of a bicycle with wire-wings.
Kay was right. And there are people out there right now, on the equivalent of a crisp morning at Kittyhawk, trying, and trying, and trying to prove it.
A computer is a tool that helps us wonder. If you can open it up in a clear, abstractable, safe way, there doesn't have to be limits to that wondering.
And flying isn't? Surely it is for some, just as for some it is "merely" a tool for transportation.
A computer is a tool that helps us wonder.
So is a book. You don't have to be able to program The Little Prince or The Way Things Work in order for it to help you wonder.
A computer is a tool that helps us wonder.
And my mom wonders why it keeps crashing when she tries to check her email. I tried to explain how important it was that she had the freedom to fix the bug herself, but she won't listen. She foolishly is only concerned with checking her email.
If you can X...
Which is to say X is not yet a solved problem, or even one known to have a solution. In the meantime: Forgive us all who dare persist contentedly with balloons and dirigibles.
Flying and books incite us to wonder. A computer can assist us in wondering. It can enable us to explore ideas. That's the potential of the computer and that's what Alan's Dynabook was promising.
Sure, we can rip out functionality until we have a computer that is so pretty and simple that a monkey can use it, but that's not fulfilling the promise that computers hold. That's not what Alan was talking about.
Forgive us all who dare persist contentedly with balloons
You're forgiven. (Just not tolerated well on a place like HN. :-)
So now my mother is a monkey? Come on now, you're exaggerating.
But seriously, I apologize if I went over the top. I understand and respect what you're saying. There's nothing wrong with getting the current state-of-the-art stable and usable; I shouldn't have intimated otherwise.
Edit: And don't tell your mom what I said or she'll tell my mom, and then I'll get it.