Especially for Habr: interview with Alan Kay

    “Fortunately or unfortunately, I learned to read well at the age of three. So I managed to read about 150 books before the first grade. I always knew when teachers were talking nonsense. ” - Alan Kay



    Hello everyone.
    I got, then, the email of Alan Kay. And I conceived the idea to chat with this legendary IT pioneer. Well, in order to wind up the importance and significance, I offered Alan Kay an interview on behalf of the entire Russian IT community. He agreed. (And now Alan Kay knows that Habr is cool!)

    I recall the merits of Alan
    • He worked in the legendary Xerox PARC, Atari, Apple, Disney, HP.
    • He proposed the concept of Dynabook (in 1968), which determined the conceptual basis for a laptop, tablet computer and e-book.
    • One of the “founding fathers” of object-oriented programming (SmallTalk, 1969).
    • He participated in the creation of the first personal computer Xerox Alto (1973).
    • The initiator of the useful movement "Every child on a laptop."
    • In 2001, he founded the Viewpoints Research Institute, a nonprofit organization dedicated to children, learning, and cutting-edge software development.
    • In 2006, he challenged the industry - he announced the possibility of creating an operating system with a graphical interface of 20,000 lines of code.
    • In 2016, joined Y Combinator.


    I had a couple of weeks to reach the most "advanced" IT people of Runet. It turned out to be a rather fun and interesting quest. And a kind of test for one’s own / another’s, for knowledge of the history of IT, for adequacy, for the connectivity of the IT community, for the ability to communicate (I have minus three friends on Facebook), for the ability of organizations / communities to act as a whole. Many thanks to those who passed this "test".

    There were 61 questions. Alan did not answer all, but the numbering remained original, for the convenience of version synchronization.

    Special thanks for the translation to Danil Kornev, Alexander Kozlov and Sergey Danshin. If you have recommendations on how to translate some kind of semantic block better - write in a personal.



    Alan Kay:Wow, how many questions !!! Even for someone who loves to answer questions! Perhaps it will take me several weeks to answer all your questions, and even after that it is not a fact that I will answer all of them! There are some particularly good questions here, and the most difficult to answer (and I will answer them one of the last).

    Magister ludi

    (future thinker, sci-fi analyst, information and IT-revolution researcher, IT-journalist)

    Question 1


    original
    1) You read a lot of books (about 300 per year). How do you choose them? How do you organize your personal knowledge (dairies, mindmaps, or do you just rely solely on your memory)?

    These days I read fewer books a year - perhaps 2-3 per week. I don't organize anything because I want most of what I read to be connected and inside my head. (At some point as I age this idea won't work anymore.) I do write in notebooks, but as a way to get stuff back in my head. I don't do much re-reading of the written stuff in my notebooks.


    - You said that you read a lot of books (about 300 a year). How do you choose them? How do you organize your knowledge (diaries, mind maps, or just keep everything in mind)?

    - In recent years, I have not read many books - perhaps 2-3 per week. And I do not organize anything, because I want the maximum of what I read, remember and combine with what is already in my head (although at some point I will be too old for this). Periodically, I take notes in notebooks, but more simply to remember something and “save” in my head. I rarely read these notes.

    Question 2


    original
    2) What was (is) the most productive format to learn and generate ideas for you and great people from Xerox PARC (reading, individual thinking, playing golf, drinking beer together, mindstorms, hackathons)?

    First, most of us at PARC came from the ARPA-IPTO research community, which from about 1962 had generated many great ideas. Day by day, “most ideas are mediocre down to bad” so the two heuristics that are good are (a) to have lots of ideas, and (b) to have ways to keep them from preventing more to happen (the notebooks are good for this). The day to day “good ideas” may happen a few times a month. In our group at PARC there was a lot of beer drinking, biking, tennis, and “recreational coding”.

    - What is the most productive training and idea generation format that you and the coolest engineers from Xerox PARC used (reading, individual thinking, playing golf, grunting beer with the whole crowd, brainstorming, hackathons)?

    Firstly, most of the founders of PARC came from the research community that formed in the IPTO at ARPA, where a lot of great ideas were invented from 1962. For the most part, ideas are average, if not worse, therefore two heuristics are important - you need to (a) have a lot of ideas, and (b) have a way to prevent their repetition and uncontrolled growth (writing ideas into notebooks is a good solution). As a result, good ideas will appear somewhere a couple of times a month. At PARC, we talked to each other a lot - we drank beer, rode bicycles, played tennis, and did “health coding”.

    Question 3


    original
    3) Who inspired you in childhood (scientists, discoverers, philosophers)?

    Scientists, science fiction writers, musicians, artists, other thinkers. Examples are: Einstein, Robert Heinlein, Bach, Michelangelo, Vermeer, Bertrand Russell.

    - Who was your example in childhood (scientists, discoverers, philosophers)?

    - Scientists, science fiction writers, musicians, artists, and other thinkers. For example: Einstein, Robert Heinlein, Bach, Michelangelo, Vermeer, Bertrand Russell.

    Question 4


    original
    4) What are your favorite Sci-Fi books? Do you read books of Strugatsky brothers? (If yes, please share your thoughts on their books and ideas.)

    I don't know how to answer this one - too many favorites - so will come back to this.

    - What is your favorite science fiction book? Have you read the works of the Strugatsky? (If so, share your thoughts on these works.)

    - I do not know how best to answer this question - too many beloved writers - I will return to this question.

    Question 9


    original
    9) How much of your day do you happen to devote to thinking “in future” (10%, 40%, 70%?). Could you give some practical advice to people wanting to improve their ability to “think about future from the future” (maybe dive deeper into the Sci-fi books, or, instead, find inspiration from reading books on Historical figures, or playing old computer games)?

    I don't think about every kind of future (and don't know how to). But I'm always concerned with an number of “issues”, which include “what do humans need to do to be able to think better than most humans today?” And “how can we better understand and make systems of every kind?” - and, the way I tend to think - there is a lot of overlap between these two areas.

    I'm pretty sure that “pondering” is going on in the back of my mind pretty much 24 hours a day. I probably spend about 3-4 hours a day actively thinking and doing things that might help illuminate more.

    “Relaxing” helps a lot. Finding ways to be very critical without being immobilized by depression is an important set of skills to learn. Dealing with the many kinds of “blindnesses” we all have is another important set of skills. Most ideas are mediocre down to bad but this is the one time not to be critical - let the ideas flow, and write down ones that seem promising (generally speaking it is not a good things to work right away on seemingly good ideas - put them on the back of the stove for awhile).

    A metaphor is that our night vision is better outside of our fovea - if we want to see a dim star we must look to the side of it, not directly at it.

    - How much time do you devote to thoughts about the future / from the future (10%, 40%, 70%)? Can you give a couple of practical tips to people who want to pump their ability to “think about the future from the future” (maybe read more science fiction, or, conversely, more historical books, or play old computer games)?


    I do not reflect on all the possible types of the future (and indeed have no idea how to do this). What is really important for me is a list of certain “problems”. For example, “what needs to be done so that tomorrow’s people can think better than today's people?” “How can we better understand and create systems of any type?” These two problems are highly interconnected, in my opinion.

    I am quite sure that this process of reflection is going on constantly in my head 24 hours a day. I think I spend 3-4 hours a day actively thinking and working on things that allow me to better understand my list of “problems”.

    "Rest" helps a lot. One of the very useful skills is the ability to be very critical, and at the same time, not succumb to depression. Another skill is to be able to work with different types of so-called "Blindness." Most ideas are mediocre, if not worse. At the same time, if you are critical, then you can skip potentially good ideas - and you need to let ideas flow, and write down those that look promising (methodologically it’s better not to take good ideas right away - you need to postpone them for a while to come to him when it becomes clearer that they are really good).

    You can draw an analogy with night vision (the eye is more sensitive outside the central fossa of the retina) - if you want to look at a star, then it is better to look not at it itself, but at its edges.

    Danila kornev

    (CEO, Zet Universe, Inc. (a 5 yo Startup focused on building a digital work environment that facilitates both information creation and consumption on the Post-PC devices. Few keywords: dynamic ontologies, zoomable user interfaces, semantic data processing, semi- automated data organization, rich information visualization, spatio-temporal data organization, graph-based data stores). Previously worked at Microsoft, Google, and Microsoft Research. Contributed to the famous Microsoft's Productivity Future Vision 2019.)

    Question 10


    original
    10) Do you think there is still a space for innovation in the digital work environments, or are personal AI assistants and desktop metaphor the end?

    I think that “hardly anything has happened yet” so there is plenty of space for both innovation and also real invention (and big advances are really needed!)

    - Do you think there is room for innovation in the field of so-called digital work environments, or is the desktop metaphor and personal AI assistants the end of innovation in this area?

    - In my opinion, nothing breakthrough has yet happened - and there is plenty of space for innovation and new inventions (and these promising developments are very, very necessary!)

    Question 11


    original
    11) If not, what do you think of the zoomable user interfaces as the core of the alternative digital work environments pioneered by Jef Raskin given that we've got tablets / 2-in-1s on the rise, as well as AR where natural interaction and use of spatial memory are the foundational elements of interactions?

    Zoomable interfaces happened a long time before Jeff Raskin (he was not careful to document actual sources of ideas). For example, Sketchpad was not only the first modern interactive graphics system (in 1962) but it also had a zoomable interface. Another early really important zoomable interface was done by Nicholas Negroponte's “Arch-Mac” research group in the 70s for the “Spatial Data Management System”.

    - If you do not think that the innovations in the field of digital work environments have ended, then what do you think of zoomed user interfaces as the basis for alternative digital work environments? (zoomed user interfaces were promoted by Jeff Raskin himself, and today we have many 2-in-1 form factor tablets, and there is an expanded reality where natural user interfaces rely on spatial memory).

    - Zoomed user interfaces were born long before Jeff Raskin (he was not very attentive to documenting the sources of his ideas). For example, the Sketchpad was not only the first modern interactive graphic system (and this was in 1962!), But it also had a zoomable interface. Another very important example is the early zoomed-in graphical interface developed by the Arch-Mac team of Nicholas Negroponte in the 1970s for the Spatial Data Management System project.

    Danila medvedev

    (Applied futurologist, immortality, cryonics, Intelligence amplification)

    Question 14


    original
    14) Overlapping windows ... At some point it was a valuable invention, but today when we operate with huge amounts of varied information it's too hard and cumbersome to do. Do you see any prospects of solving these deficiencies of "overlapping windows"?
    If you have a big display surface - real or virtual - you don't have to overlap the windows. The initial idea was that overlapping windows would really help on screens that were too small. “Huge amounts of information” is not going to be solved by simple displays no matter how large (things have to be to fit to what human senses and brains can deal with). Another point is that “It's not Big Data that's important but Big Meaning!”

    - Once overlapping windows were a useful invention, but today, when we work with huge amounts of information, overlapping windows make it difficult to organize information in the workspace. What prospects do you see in solving this “overlapping window” problem?

    - If you have a huge display surface - real or virtual - then you do not need overlapping windows. Initially, our idea was to solve the problem of very small screens - overlapping windows on them were a great solution. “Huge amounts of information” - this cannot be dealt with with the help of a simple enlargement of screens (in the end, information objects must “fit” the scale of what the human brain can simultaneously recognize and understand). And one more thing - Big data is not so important, as Big sense.

    Question 15


    original
    15) What do you personally use as your personal computer work environment and what are your primary working tools?

    My primary working tools are between my ears and my hands, especially with drawing. (I specified the Dynabook in 1968 to have a stylus and really great sensing and inking, and have been very disappointed at how long it has taken industry to even hand wave at this - and even on tablet computers.) The drawing provided is not aimed at people who can really draw.

    The next level is books and reading in general. Again, industry has not provided something as readable as paper. The readability provided is not aimed at people who can really read.

    High up in the list of tools for me is a nice couch to stretch out and read and write on.

    I need to play music every day.

    - What tools form your personal workspace on a computer?

    - My main working tools are everything between my ears and my hands, especially when it comes to drawing. (back when I was developing the Dynabook concept, I said that a tablet should have a pen and it should support drawing, and I was deeply disappointed at how much time it took the industry to do at least that, and at least on tablet PCs) . And the level of drawing support that is in tablets, it does not reach the level of those who really know how to draw.

    At the next level, we have books and reading in general. And again, the industry has not yet produced anything as readable as plain paper. What the industry has managed to do is not very convenient for those who like to read.

    Even higher in the list of my tools for work is a comfortable sofa - you can stretch out, read and write on it.

    I need to play a musical instrument every day.

    Question 16


    original
    16) Aren't you disappointed with what computers turned into: games, porn, twitter, etc.?

    I'm disappointed with the combination of consumer businesses and the very poor education system. Computers are “instruments whose music is ideas”, and it's possible to create horrible noise with the best violins, etc. or the most beautiful music. So we have to point the fingers at the humans involved.

    And see the answer to 15.

    - You are not disappointed with what computers have turned into? Toys, porn, twitter, etc.?

    - I am disappointed with consumer businesses and a very, very poor educational system. Computers can be compared to musical instruments; their music is ideas; as you know, you can manage to get terrible sounds even when using the best violin in the world, or you can play the most unusual piece of music. So, you should poke a finger at the users themselves.

    See the answer to question No. 15.

    Question 17


    original
    17) Doug Engelbart wanted computers to augment human intellect including group intellect. Why was he ultimately denied a chance to implement his vision of augmentation?

    I don't think Doug was denied a chance. There were several problems including the answer to 16. Most people didn't even get what personal computing should be about let alone the great ideas Doug helped create.

    “Douglas Engelbart dreamed of using computers to enhance human intelligence, including collective intelligence. Why was he not given a chance to realize his concept of amplification?

    “I don’t think Dag was forbidden.” There were several problems, including those mentioned in the answer to the 16th question. Most people did not understand why personal computers were invented, not to mention the great ideas that Doug helped bring to life.

    Question 18


    original
    18) Did OLPC achieve all goals that you envisioned for DynaBook in terms of pedagogy?

    Not even close.

    - Did the OLPC project achieve all the pedagogical tasks that you set in the days of DynaBook?

    - Not even lying around.

    Question 19


    original
    19) What are the next steps for the OLPC project?

    It is worthwhile doing a project like this every 10-15 years. The next few years could be a good time to start another one.

    - What awaits OLPC in the future?

    - For me, it would be a great idea to repeat this kind of project every 10-15 years. In a few years it will be the most to start a new such project.

    Question 20


    original
    20) What did you manage to do, where you failed?

    This question I don't quite understand. But the main idea is to “keep on keeping on”

    - What did you manage to achieve in the OLPC project, and where did you fail?

    - Honestly, I did not quite understand this question. But the basic idea is simple - you need to continue to "continue."

    Anatoliy Levenchuk

    (scientific advisor, System Management School)

    Question 21


    original
    21) There were major changes in connectivity modeling. Do you find it the spark of the finally coming next computer revolution, or do you still await for the classic computer revolution (which is still in the limbo state)? Do you find it bad to mix up symbol and connectivity-based computing?

    - I don't think I understand the first question. As to the second, I think it's important for any kind of “correlative modeling” (such as with neural nets, so-called “deep learning” etc) to be coupled with “cognitive modeling” which uses reflective models and simulations of knowledge not just reactions. (Minsky's “Society of Mind” book is still a good guide to all of these different paradigms.)

    To me, the classic computer revolution is one that is supposed to provide a combination of intellectual amplification not just outside the human mind, but to help better form the human mind. In other words, it's what happens “between the ears” that counts - and the pioneers wanted this to be “much better” than what was happening with most human development. (I'd say that today, for most people who use computers as conveniences and social communication, what's happening between the ears is retrograde.)

    - Over the past five years, connectivist modeling has taken a huge step forward. Does Alan Kay think that this is part of the next computer revolution that has finally begun, or is he still waiting for the classical computer revolution (which will not start in any way) and would not mix symbolist and connectionist computing?

    - I think I did not quite understand the first question. As for the second, then, as I think, for any kind of so-called “Correlation modeling” (such as neural networks, “deep learning”, etc.) it is very important to be paired with “cognitive modeling”, which uses reflective models and simulations of knowledge, and not just reactions. (Minsky's Society of Mind is still an excellent introduction to all paradigms of this kind).

    For me, the classic computer revolution lies in the idea of ​​not only giving a person a set of tools that strengthen the human mind, but also helping to better shape the human mind. In other words, what happens “between the ears” plays a very important role - and the so-called the pioneers wanted to improve these processes compared to all that humanity has had for most of its history. (I would say that today what happens “between the ears” of people who for the most part use computers for entertainment and social communications can be called “retrograde”).

    Question 22


    original
    22) What's your take on Julia programming language as the next gen scientific computing programming language, coming after Fortran-Matlab-Python?

    - I don't know how to criticize a language that has lots of features without trying to do a major project in it (and I haven't tried Julia). I think scientific models eventually require simulations - and these eventually imply loosely coupled objects - and it's hard to see in the descriptions of Julia whether these are part of what's optimized as a center of the design.

    - How do you feel about the Julia language, which is now slowly entering the role of the leading language of scientific computing after fortran-matlab-python?

    - Hmm, I do not think that I have ideas on how to criticize a programming language, and even with such rich functionality, without trying to write at least one large project on it (and I have not tried to write programs in Julia). In my opinion, scientific models ultimately need simulations - and they, in turn, need loosely coupled objects - and judging by the documentation for the programming language Julia, it is not very clear how these loosely coupled objects are the center whole language, or are not.

    Question 23


    original
    23) What's the biggest achievement of VPRI in the last year / five years?

    Take a look at the Viewpoints “Writings” page and see for yourself.

    - What do you consider the main achievement of VPRI over the past year / 5 years?

    “Look at the Writings section on my Viewpoints Research Institute’s website , and judge for yourself.

    Pavel luksha

    (Educator, systemic social innovation catalyst, futurologist, founder of Global Education Futures Forum (www.edu2035.org))

    Question 24


    original
    24) We are already deep in the information age, and impact of new technologies has become evident even for the mass public. People voice various significant risks of emergent technologies such as autonomous robotics and machine learning, from massive global destruction of jobs to civilizational existential risks. Yes, these threats have been with us for a while (similar discussions occurred even in early 1970s), but now they are evident and accumulating. In this context, suppose we talk of "positive" future scenarios in which humans and technosphere do not compete with each other, but coexist and cocreate in positive and productive ways. What are your top picks for "skills for complex society" (individual and maybe collective) that need to developed massively, in order to increase the probability of this scenario,

    I think an approach to this question might best start with the amount of new clean energy (both with regard to emissions and to heat created) that can be made available via science and technology. This will establish goals for dealing with all of the basics in equitable ways (including population).

    Under such circumstances, the biggest skills needed will be “systems awareness” along many dimensions (including not just cooperation, but understanding that systems shouldn't be “gamed”), and skills that deal with “growth”, avoiding boredom, and finding excitement without the need for greed, war, power, etc.

    This will be a very tough collection of new things for human beings to learn fluently. And as you said, these issues have been around for a long time. There has been some excellent writing about them eg many of Bertrand Russell's books starting about 100 years ago dealt with many of these issues.)

    - We are in the phase of active development of a “digital civilization”. It became clear that emerging technologies, such as autonomous robotics and machine learning, pose significant risks for humanity - from the mass disappearance of jobs to the threat of the destruction of civilization per se. Of course, these threats are not new (similar discussions took place back in the 1970s), but now they are becoming more real, and the risk is growing due to the acceleration of technological progress and the mutual influence of technologies on each other. In this context, let's look at a positive scenario for the future where people and the technosphere do not compete with each other, but co-exist productively and co-create. What key competencies (individual and possibly collective) do people need to massively develop in order to increase the likelihood of this particular scenario - and in the long run,

    - I suppose, in order to answer this question, it is worth starting to understand what approximate volumes of clean energy (taking into account emissions and heat generated) we can produce thanks to science and technology. This will help us determine further goals, and with them, indicators (including the size of the population, etc.).

    Under such conditions, one of the most important qualities or abilities will be the ability to "understand the system, take it into account, and its feedback" in many dimensions at the same time (and this is not only about cooperation, but also about the fact that you need to understand that you should not try " beat "system). Other major abilities will be those related to “growth”, the fight against boredom, and the ability to enjoy, admire life without the need for war, envy, power, etc.

    In my opinion, there are many things ahead that people will have to learn very quickly. As you said, all these problems have existed for many years. There were some very remarkable thoughts about these problems, for example, in the books of Bertrand Russell, about a hundred years ago, where he tried to figure out how to solve these difficulties.

    Question 25


    original
    25) The project of creating tools for "augmented intelligence", including input and output interfaces, that you at PARC and Doug Engelbart at SRI have been implementing since 1960s, has only found partial way into massive ICT applications. What would be the main avenues for future development of interfaces (and maybe complementary skills of children and adults?) That would help complete this unfinished revolution? As part of the question: what do you think of the potential of neural interfaces? Do you envisage them as one of the indispensable components of human-machine interfaces as the complexity of social life and technosphere increases further into 21st century?

    I don't view neural interfaces as indispensable (first we need to get human brain-minds working better, etc.) The current day shows us what improved communications technologies wind up being used for if humans aren't helped to improve their brain- minds!

    Both Doug and I (and others) thought then that enormous changes to general education would be needed to make use of the new tools and media (what would a person use a violin for if they didn't learn how to play it?)

    And , given that, it's clear that what his and my groups did 40-50 years ago was just a bare start at what's needed. It's definitely time for another pass at these problems. (Bret Victor is a really great young thinker about these issues)

    - The project of creating “augmented intelligence” tools, including the creation of new “human-computer” interfaces, which you have been implementing at PARC since the late 1960s, and Douglas Engelbart at SRI, has largely become the basis of modern ICTs - but a lot of those ideas and approaches are still not implemented in mass solutions. What, in your opinion, should be the main directions of the development of interfaces (and, possibly, the skills of children and adults) that can move further the "unfinished revolution"?
    And as part of the question: what do you think of the potential of neural interfaces? Do you think that they will become necessary components of human-machine interfaces as our society and technosphere become more complex (say, by the middle of the 21st century)?


    - For me, neural interfaces are not as indispensable as they might seem (and in general, for starters, it is necessary that human minds learn to work more efficiently by themselves). Today it is clear that even the most effective communication technologies used by people do not at all help their minds become more efficient!

    And Doug, and I, and our colleagues, we all believed that for the new media tools to bring maximum benefit, a tremendous transformation of general education is necessary (in the end, what can a person with a violin do if he does not know how to play it ?)

    Thus, what our group and I did 40-50 years ago is only a fraction of what needs to be done. And now is a great time for a new attempt to solve these problems (Bret Victor is an amazing young thinker in this field).

    Question 26


    original
    26) So far, technologies have been able to make only limited transformative impact on education. And, as you have yourself stated in one of your interviews, «music is not in the piano», it is not the matter of technologies but of the way in which they are applied. If you were to nominate top three necessary changes in the *social* context of new technologies to make impact on the next generation of learners, what would these be?

    Yes, “the music is not in the piano”, but because of the piano we can wind up with great music that wouldn’t have happened otherwise. The great music happened in the minds of artists who contemplated some of the musical possibilities of the new instruments. (And some of the instruments happened because of some of the new possibilities for music in the minds of the composers!) It’s hard to think about any of these areas without considering how co-evolution of ideas, works, tools, and media all intertwine.

    Many of the inventions that contribute to “civilization” are “anti-genetic” in several senses (a) they are not found “naturally” in traditional cultures, and (b) they often oppose a genetic tendency that is thought to be natural (for example: revenge, inequality, etc.)

    If we look for “natural things” that get us into trouble, we can find many: for example: stories, desire for status and power, envy, desires for revenge, etc. Take a look at Anthropology’s list of several hundred “human universals” (Donald Brown etc.) and pick your favorites to be opposed by better forms of education.

    — Не смотря на обещанную “цифровую революцию” в образовании, новые технологии всё ещё оказывают довольно незначительное влияние на модели массового образования. Как вы сами сказали в одном из интервью, “музыка находится не в пианино”, вопрос не в технологиях, а в том, каким образом они применяются. Можете ли вы назвать три ключевых изменения в социальном контексте применения технологий, необходимые для того, чтобы следующие поколения начали действительно учиться по-другому?

    - Yes, as they say, “music is not in the piano,” but it is thanks to the piano that we can play the greatest music in our history. And this greatest music was born in the minds of its creators due to the fact that the new musical instruments of their time gave them new opportunities to play this music. (and some of these instruments came about because these musicians came up with new types of music that required these new instruments!). I would say that it is impossible to think about new areas of development without thinking about how ideas, works, tools and media develop, evolve together.

    If you think about it, it will be seen that most of the inventions that developed our civilization are not the product of "genetics" in the sense that they were not born "naturally" in traditional cultures, "and they very often go against the so-called" genetic "tendencies that seem natural (for example, revenge, inequality, etc.).

    If you look for those things that are" natural "for a person and that cause us troubles, then you can find many such things, for example:" stories "," The pursuit of status and power "," envy "," the desire to revenge ”, etc. Look at the list of several hundred human universals (“ human universals ”see the book of Donald Brown, etc.), and choose the training methods for working on those that interest you.

    Question 27


    original
    27) Strategically — especially given your role in the constructivist «movement» as one of its key «ideologists» — what is in your opinion a preferred way of human beings to learn into the 21st century? Should they, for instance, learn more from self organized technology intense learning environments, or from and within groups of their peers organized as «communities of practice», or within more traditional institutional set-ups such as schools and universities «augmented» with certain technological and social innovations, or?

    — I’m not much of an ideologist or any kind of “ist”. The real questions here have to do with (a) what needs to happen to the many parts of the brain-mind to achieve fluency in some pursuit? And (b) what are the best ways to get those changes to happen?

    An important societal case is © how to deal with teaching and learning the subjects that the society has decided that everyone shall learn, whether they want to or not?

    I’ve talked and written about some of this elsewhere. But for here, the simplest place to start might be to read Montessori’s books (they are all good). The one she did towards the end of her life — The Absorbent Mind — is particularly good.

    “Given your role as one of the ideologists of the“ constructivist ”movement in education, what do you think will be the preferred way to educate people in the 21st century? For example, will they mainly study in self-organizing digital environments under the control of artificial intelligence, or all kinds of “communities of practice” and project teams where they can learn from their peers and peers - or they will continue to study in more traditional formats, at schools and universities (taking into account the fact that they will update with keeping and formats of curricula), or something else?


    - I’m not really an ideologist (and generally a “log”). Here’s what you really should think about:

    • a) What should happen to the many components of the human brain-mind, so that he becomes well versed in a particular matter?
    • b) What needs to be done to achieve such a change?
    • c) A very important point in the context of society as a whole. What should we do with the training and study of subjects that society considers necessary to study, regardless of whether the students want to study it or not?

    I talked a lot about this and wrote ... But to begin with, it is best to start by reading Montessori books (they are all very good). The book she wrote near the end of her life - The Absorbent Mind - is especially good.

    Petr levich

    (Head of Science, Technology and Society Department at Moscow Technological Institute, Founder at Future Foundation)

    Question 30


    original
    30) What’s your opinion on the following possible approach:
    What if we’d build a hackathon-like environment where any kind of technical experiments are allowed, w/o requirement to apply found solutions, or find a business model; where people would live and seek for the new meanings in the constantly changing and full of complexity world world? Would you be interested to participate in the creation of such an environment-space?

    This is partly what Engelbart wanted to do.

    One thing to ponder is that “improvisation” is quite different from “composing”, especially as you enter more and more developed areas. The pop culture likes to improvise and to consume improvisations, but is much less inclined towards the planning etc needed for more elaborate compositions. “Hackathons” are much too much about “doing” and not enough about “thinking”.

    - Do you think it is advisable to create a space with permission to experiment in the technical field, such a constant hackathon regime, but without the obligatoriness of applicability and the presence of a business model where people would live and look for new meanings in a rapidly changing complex world, prototyping new technologies on themselves? Would you be interested in participating in the creation of such a project?

    “In part, this is exactly what Engelbart wanted.” An important nuance here is that “improvisation” and “composition” (creation) are far from the same thing. It is common for mass culture to improvise and consume improvisation, but it is not at all characteristic of the planning and other activities necessary for creation, or composition. In this context, “hackathons” are about “doing” rather than “thinking”.

    Ilya Sobol

    (Inventor, entrepreneur, process/experience designer, making evangelist and maker)

    Вопрос 31


    original
    31) What is you creative process? Do you use special methods to create ideas? Or its just periodical insights and matter of luck? Or may be you have a task to design and in process of problem solving this task you usually find unexpected solution? Or may be you have your own creative method?

    “Luck” is a complex idea, but it does at least stand for “things happening you can’t control”, and so it’s a big factor in most creative processes in many ways. I answered this a little in question 15.

    - How do you come up with new ideas? Is it enlightenment, luck and chance, or do you still formulate for yourself some tasks, solving which you come up with? Or do you have your own way?

    “Luck is a complicated idea. It partially stands behind “things that happen out of our control”, and it is an important factor in most creative processes. I wrote a little more about this in my answer to question No. 15.

    Question 32


    original
    32) If your creative process based on problem solving approach, what was the problems you tried to solve when you found dynabook concept?

    — My process is much more based on “problem *finding*” than problem solving (there’s more in other parts of this document). The Dynabook was an extremely simple “vision” that matched up with what I thought was wonderful about what Seymour Papert was doing. In other words it was a “fit” into the qualitative universe of how to learn difficult to learn things.

    - If new ideas are the result of solving problems, then what tasks or questions did you set for yourself when you came up with a laptop / OOP / etc?


    - My process is much more focused on the “search” of problems than on their solution (I wrote about this in other answers to questions in this interview). So, Dynabook is a very, very simple idea, coinciding with the great work that Seymour Papert was working on. In other words, it was a way to “cling” to that steep universe, where they learn to study things that are difficult to learn .

    Dmitriy Sudakov

    (Education consultant, leader of Atlas of emerging jobs (atlas100.ru/en))

    Question 36


    original
    36) What do you think could be the next big step in the personal computing? We had PCs and laptops, then tablets and phones. What's next?

    How about really learning to use any of these?

    - What do you see as the next “big step” in the development of personal devices? We switched from computers to laptops, then tablets and smartphones. What's next?

    - How about learning to use at least these tools effectively?

    Igor Trapeznikov

    (Neuroscientist, cyborg and lifelogger. CEO AIcoda Co. LTD)

    Question 38


    original
    38) What made you so curious about life, and things, in your childhood?

    Most children start off somewhat curious, but are willing to accept stories as answers. I think I might have been a bit more curious than most, and wasn’t willing to accept stories as answers. I think a very big factor was that I more or less accidentally learned how to read fluently at a very early age, and reading lots of books really opens up one’s world view, and this really boosts curiosity.

    - What in childhood served the development of your curiosity?

    - Most children are born full of curiosity, but they are quite satisfied with the “stories” as answers to their questions. It seems to me that I was much more curious, and simple "stories" did not suit me. Perhaps a big role was played by the fact that I more or less accidentally learned to read fluently at a very young age, and reading a large number of books, as you know, significantly expands the human horizon, and only fuels curiosity.

    Question 39


    original
    39) What do you think people lack in terms of knowledge and technologies?

    I’ll come back to this one. The simplest answer is that “we can’t learn to see until we realize we are blind”, and most people don’t realize they are blind. This is a kind a meta-knowledge that — like context — makes all the difference in how one approaches learning and knowledge.

    - What do you think is missing humanity in terms of knowledge and technology?

    - I will return to this issue. The simplest answer is “we cannot learn to look until we realize that we are blind,” and most people are not aware that they are blind. This is a kind of meta-knowledge — and if used, it completely changes the approach to learning and to knowledge.

    Shapelez is a good person. He adhered very well, and therefore we have a block of corporate issues. Do as Shapelez .

    Qrator Labs (DDoS mitigation and network availability)



    Alexander Lyamin

    (founder and the CEO of Qrator Labs)

    Question 40


    original
    40) In the interview of 2008, Donald Knuth said: «it looks more or less like the hardware designers have run out of ideas, and that they’re trying to pass the blame for the future demise of Moore’s Law to the software writers by giving us machines that work faster only on a few key benchmarks! I won’t be surprised at all if the whole multithreading idea turns out to be a flop, worse than the [Itanium] approach that was supposed to be so terrific — until it turned out that the wished-for compilers were basically impossible to write.»

    This year even consumer processors being sold are 8-cores, top mainstream processors are dozens-core. So the processor industry took a wrong turn, according to Knuth. How do you evaluate the current state of the industry and what do you think is going to happen in the nearest decade?

    Interesting question. There are certainly many good ideas for hardware that have been around starting in the early 60s that haven’t been used, so *some* hardware people didn’t run out of ideas. However, the chip manufacturers (e.g. Intel and Motorola) were never interested in ideas (or really even in software).

    The next set of answers here are too long for this forum — but they are along the lines of *not* trying to compile bad languages into multiprocessors, but to create good languages and let the hardware designs optimize those good ideas.

    I certainly agree that the industry took a wrong turn — but the first one goes all the way back into the 60s before CPUs could be done on an integrated circuit.


    — Дональд Кнут в интервью от 2008 года сказал: «Для меня всё выглядит похожим на то, что разработчики аппаратного обеспечения исчерпали все идеи и пытаются свалить вину за близкую кончину Закона Мура на авторов ПО, предоставив машины работающие быстрее только в определённых ситуациях и с конкретными бенчмарками. Я нисколько не удивлюсь если вся многопоточная идея окажется пустышкой, хуже, чем подход Itanium который должен был быть «таким потерянным», пока не выяснилось, что желаемый компилятор невозможно написать в принципе.» В этом году даже обычные потребительские процессоры стали 8-ми ядерными, топовые mainstream процессоры исчисляют свои ядра десятками. Т.е. индустрия процессоростроителей «took a wrong turn», according to Knuth.

    And how do you assess what is happening and what do you think should and will happen with processors in the next decade?


    - Good question. Many good iron ideas that began in the early 60s were never used or implemented, so some engineers still have some ideas. However, chip makers (such as Intel and Motorola) were never interested in ideas (in fact, even in software).

    The following set of answers is too long for this format - but it is between the lines and concerns the unnecessary compilation of bad languages ​​into multiprocessors. The right way is to create good languages ​​and provide the hardware and design with the opportunity to optimize these good ideas.

    Of course, I agree that the (processor) industry has turned in the wrong direction - and the first one was committed back in the 60s before the CPU began to be produced as an integrated system on a chip.



    Alexander Azimov

    (network architect Qrator Labs)

    Question 40


    original
    41) We are moving from the world where we were using technology to the world where technology is using us. And context advertisement is most evident and harmless example. Do you see ongoing risks in this field?

    In the 19th century, the American writer Thoreau wrote that “we become the tools of our tools”. This has been studied quite a bit by a few people since the 50s - such as Marshall McLuhan - and it is a big problem which incurs deep risks.

    - From the world in which we use these or those technologies we move towards the world in which technologies use us. Contextual advertising is the most obvious and harmless example. What risks do you see in this field?

    “In the 19th century, the American writer Thoreau said that“ we become tools of our own tools. ” This phenomenon has been studied by some people since the early 1950s, such as Marshall McLuhan, and I see this trend as a big problem, entailing potentially high risks.

    Andrew Leskin

    (software engineer Qrator Labs)

    Question 44


    original
    44) Where to, from your point of view, the whole OOP development is going? Do you agree with the statement that OOP is becoming something in the middle between the functional and object-oriented programming? Is there a chance for evolving some new paradigm?

    In the way I thought about OOP (and what I meant by it) there is no conflict at all with functional programming. Take a look at the idea of “fluents” by John McCarthy, and “pseudotime” by a variety of researchers from Strachey onwards. It’s just that most people want to do imperative programming, and forced what they decided to call OOP into something that is essentially Abstract Data Types, not OOP.

    - Where do you think OOP development is going? Do you agree with the statement that OOP degenerates into something between object-oriented and functional programming? Is it possible to move to some other, new paradigm?

    - In the way I thought about OOP (and what I meant by this) there is absolutely no conflict with functional programming. Look at the idea of ​​“fluents” en.wikipedia.org/wiki/Fluent_ (artificial_intelligence) by John McCarthy and the “pseudo-time” formulated by several researchers following Streychi. The point is simply that people want to program imperatively and, as a result, are forced to call object-oriented programming what is actually abstract data types, not OOP.

    Anton Orlov

    (software engineer Qrator Labs)

    Question 45


    original
    45) Thinking of the latest machine learning progress could we expect any change in the way computer programs are written in the nearest future? Is there a chance computers would help a man in writing those programs? How could we move this way?

    Yes, I’ve given a few talks along this lines. They are on YouTube but I don’t post the talks and I don’t keep track of them. If I have time I’ll try to find one that says a few words about this.

    - Given the progress in machine learning, is it possible to expect that the way of writing computer programs will change in the near future? That is, can computers themselves help a person significantly in writing programs? How to move in this direction?

    - Yes, I have repeatedly spoken out on this subject - there are records on YouTube, but I do not publish my own speeches and do not even follow them. If I have time, I will try to find the most relevant one containing a couple of words on the topic of the question you asked.

    Meklon

    (Linux-admin, medicine researcher)

    Question 56


    original
    56) What do you think about computers in early childhood (3-7 years)? Many parents forbid their children to use computers and other gadgets.

    What do you think about books in early childhood? What do you think about TinkerToys in early childhood. These are all good ideas if the books are good and the tools are good.

    - How do you feel about immersing children in working with computers from a very young age? Many very parents very strictly limit this to at least 7 years.

    - What about reading in early childhood? Or a TinkerToys constructor ? These are great ideas if the books are good and the tools are good.

    Ivan Ponomarev

    (Java programmer)

    Question 58


    original
    58) As one of the key authors of Smalltalk language and OOP paradigm, what do you think about the fact that all the languages that are recognized nowadays are not purely object-oriented, but tend to be multi-paradigm? Is it because the power of pure OOP was never fully realized, or is it just the way programming languages evolve? (There are people e. g. in Java language community who are fighting for 'pure' OOP principles for the sake of code cleanliness, sort of 'back to the roots', what do you think about this?)

    Mulitparadigm is generally good. “Bad-paradigm” is almost always bad. Imperative programming with data structures (and using “setters” in “OOP”) is a very bad way to program. But combining good ways to program is good. Religions tend to be bad even when they are trying to be good, because they have a hard time tolerating multiple perspectives — and there are also religions based around ideas that are not good. So getting religion out of computing would be a very good thing. The real questions are about: how to make systems where “simple things are simple, and complex things are possible” and that includes being able to change one’s mind after the fact and having the resulting changes be a reasonable amount of effort.

    — Как один из авторов языка Smalltalk и парадигмы ООП, как вы относитесь к тому факту, что получившие распространение языки программирования не придерживаются чистого ООП, являясь по сути мультипарадигменными? Это результат того, что возможности чистого ООП оказались недооценены, или всё так и надо?

    - Multiparadigmness in a programming language is, in general, good. Bad - when a bad paradigm is used in a language. Imperative programming using structures as well as “setters” in the “so-called OOP” are very poor ways to program. A religious cult becomes evil even if it tries to be good, because it encounters difficulties in relation to various changes in the world - and it also happens that the cult arises with respect to incorrect ideas. Therefore, to remove any "religious cult" from the computer sphere would be a great blessing. The real question is: how to create systems in which "simple things would be simple, and complex things would be possible." This, in particular, implies the possibility of changing one’s point of view on things over time, as well as the ability to make changes,

    Artiom Zheltov

    (futurologist)

    Вопрос 60


    original
    60) How did Xerox PARC developers get tasks to do? How did they identify/foresee problems that they've solved with the tech they've developed at PARC?

    The researcher organization at Parc was “flat”, no one was told to do anything. Because Xerox “counted heads” when budgeting (a researcher counted the same as a technician) we had only a few technicians (we used the head count for researchers). We all came from the ARPA-IPTO research projects, and carried Licklider’s and other’s visions with us. There were “admired personalities”. There was a weekly meeting of all to sort things out. Interests were offered, pledges could be made. For recreation many of us would spend 20-30% of our time as “hands” for someone else (as a programmer or soldering, etc.) This was very soothing.

    In practice there were always a number of projects going on — they tended to orbit around the “admired personalities” and “exciting projects” (e.g. Butler Lampson, Chuck Thacker, the Alto, Ethernet, etc) — and these led to “grouplets” that persisted for varying lengths of time. My group — the Learning Research Group — persisted for most of 8 years.


    Вопрос 61


    original
    61) There's a hypothesis that the stuff you've invented/developed wasn't really needed at that time. They were nice to have, awesome futuristic things. But 40 years later they became the foundational element of people's lives. If so, what kind of nice to have, awesome futuristic things do you see as something that solves problems that look irrelevant or not yet important today?

    I would say that what we invented *was* needed at the time, it’s just that almost no one realized this. Try thinking about what is needed right now but no one can see it

    - There is a hypothesis that those things that you created or invented were not in demand at the time. They were something to enjoy, amazing futuristic stuff. But 40 years later, they became the cornerstone in people's lives. In this situation, what do you think, what awesome futuristic things that are irrelevant today can solve some problems in the future?

    - I would say that what we invented was relevant at the time, almost no one realized it. Try to imagine what is needed right now, but doesn’t occur to anyone?

    Thanks.

    Publications in Russian:


    «American business is completely fucked up because it is all about competition. Но наш мир создан для взаимовыгодного сотрудничества, и это именно то, чему надо обучать людей. Лучшая аналогия — это командные виды спорта.»



    P.S.


    I think Engelbard was glad that Alan Kay answered questions in the joint editing mode of the document.



    PPS


    What question would you ask Alan Kay?

    Also popular now: