Is there a consensus in the CS community that Kay is an important figure who deserves this kind of adulation? Do any of us really understand or care about his achievements? Just asking.
It's true that his specific outputs there (Smalltalk, Squeak, etc.) have never become that mainstream. But: 1) they have become much more widely used than a lot of other research systems (Smalltalk is not just some PhD thesis used by 3 people), and 2) being even "highly influential" on two major areas of development that are mainstream-important is a pretty high bar to clear already.
http://online.wsj.com/ad/article/laptop-invented
https://en.wikipedia.org/wiki/History_of_the_graphical_user_...
First thing first is that he's a pretty great talker. That ensures anyone will even take the time to listen to what he has to say. But unlike some other great talkers/writers in our field with slavish devotion, Alan Kay was part of pioneering cutting edge research into personal computing. To me he's a breath of fresh air. He's not trying to sell hocus pocus methodologies at software conferences, he's taking a holistic, big picture view of computing and offers some crazy ideas - not all of which I agree with, but most of which have some merit.
His view that computing is an immature field and we don't yet know what we are doing really resonates with me. People too often try and treat it like it's something mature like the electrical engineering or mathematics that underpins it, but it's totally not. We're still in the stage where we are constantly forgetting our history because we're stuck in fashion cycles (ie, people calling Swift innovative because it takes ideas from 1970s programming languages). And he rightly shines a light on that and says hey, we can and should aim to do better. For that I applaud him.
He can be a bit esoteric and out there and what he says is often not directly applicable to everyday programmers. I totally agree. But from thousand foot view strategic perspective he's well worth listening to.
The Smalltalk VM led to Sun working on the Self VM which led to the JVM.
"Notes on the Synthesis of Form" (from Kay's reading list on the Squeak website) has some interesting stuff about trying to take graphs of nodes like this and figure out a good hierarchical decomposition given emergent discovery of edges.
Here's something I've been thinking about lately: what if the object edges were organized spatially like a cellular automaton, maybe we can use spatial intuitions then, and if we only allowed linear-typed resources to be passed between them we actually have conserved "flow" of these resources, allowing the use of tools from economics to study the flow. Then it ends up looking like https://mitpress.mit.edu/books/growing-artificial-societies except with programmable monadic behavior at each cell. The linear-typed resources will allow linear proofs locally at each object but the monadic programming of each will allow non-linear emergence.
(For me, Alan Kay is good publicist for other geniuses. He's talked about Ivan Sutherland, Neil Postman, etc. I was overlooking those people until I started watching his talks on Youtube.)
"For pioneering many of the ideas at the root of contemporary object-oriented programming languages, leading the team that developed Smalltalk, and for fundamental contributions to personal computing." [1]