No Dynabook Yet: An Interview with Computing Pioneer Alan Kay

April 11, 2013 at 1:51 am 12 comments

Nice interview with Alan Kay, with nice links to Engelbart and Knowledge Navigator videos.  Here’s the segment where Alan describes why the iPad is not a Dynabook.

The interesting thing about this question is that it is quite clear from the several early papers that it was an ancillary point for the Dynabook to be able to simulate all existing media in an editable/authorable form in a highly portable networked (including wireless) form. The main point was for it to be able to qualitatively extend the notions of “reading, writing, sharing, publishing, etc. of ideas” literacy to include the “computer reading, writing, sharing, publishing of ideas” that is the computer’s special province.

For all media, the original intent was “symmetric authoring and consuming”.

Isn’t it crystal clear that this last and most important service is quite lacking in today’s computing for the general public? Apple with the iPad and iPhone goes even further and does not allow children to download an Etoy made by another child somewhere in the world. This could not be farther from the original intentions of the entire ARPA-IPTO/PARC community in the ’60s and ’70s.

via An Interview with Computing Pioneer Alan Kay |

Entry filed under: Uncategorized. Tags: , , .

Congratulations to Eric Roberts on the 2012 ACM Karl V. Karlstrom Outstanding Educator Award! Neil Fraser on CS in Vietnam and (unfortunately) in US

12 Comments Add your own

  • 1. Jeff Rick  |  April 11, 2013 at 10:03 am

    While I appreciate Alan’s vision of “symmetric authoring and consuming,” I think it is unrealistic for many media. There is usually a significant gulf between understanding and creation. For a symmetric medium (or knowledge building environment) to be successful, its users must bridge that gap. My standard example is third grade science textbooks. Can third graders understand these textbooks? Yes, hopefully. Can they create them? No. Hence, third grade science textbooks are not a good medium for collaborative knowledge building. It doesn’t mean that they are not valuable; they just are not valuable for this symmetric goal. In contrast, as an academic, I am able to read journal papers and write journal papers. In my community, journals are a good medium for collaborative knowledge building. There are other examples of this working: Wikipedia, Facebook, etc. The challenge is creating media that make it possible to bridge the gulf. That’s not easy. Understanding comes before creation. To sing well, you have to understand what it means to sing well. Sometimes, understanding is enough to be valuable. I can enjoy the Harry Potter series without the impetus to write my own novel. I can enjoy The King’s Speech without any competence in creating something similar. Roger Ebert, who recently passed away, did an amazing job of understanding movies. Instead of trying to create movies about movies (i.e., symmetric authoring and consuming), he just wrote about them. Sometimes, that’s enough.

    Asking whether a piece of hardware, such as the iPad, is the dynabook is the wrong question. It all depends on the ecosystem and the software. As Alan points out, the ecosystem is a bit restrictive and the software puts too much emphasis on consumption. In other ways, the iPad is there. It is portable, connected, flexible, and reasonably priced. I think we can make the dynabook (or rather create and distribute media that more closely align with the dynabook vision) using the iPad. That’s up to us as developers. Whether others will use the device thusly or just use it to consume depends on many factors but availability of that kind of application is the next step we should work towards.

  • 2. alanone1  |  April 12, 2013 at 8:18 am

    Hi Jeff

    There is a bit of confusion between “media” and “content”. The idea was to make all media on computers symmetrically authorable as well as consumable. And not to confuse this with the level of content that can be reached in any medium by experts.

    One of the projects from one of the best teachers I ever had (4th grade) was to have children write and make books. The content of these books was very 4th grade, but the process was the real deal, both from the standpoint of doing the writing, and also learning how to make the bindings and covers. I remember her apologizing that she didn’t have a little printing press, but we did do “woodcuts” (using linoleum) for “printing” some of the illustrations. This enfranchised us into an important part of the adult world of thought and creation.

    A brilliant early example of symmetric authoring and consumption was Hypercard — the level of content and appearance had a wide range but it really was a dynamic authoring medium for pretty much everyone.

    And of course you can read your Harry Potter novel without having to write something similar. But please don’t tell me that you shouldn’t learn how to write, and that children don’t need good authoring systems for this. You are confusing “being a musician” with “being a great musician”.

    And you don’t have to make the “King’s Speech” to make movies. The point is to have authoring systems to make movies that can be used by all if movies are media on computers.

    The iPad could be useable if Apple were to simply try to understand the original ethos. However, besides the many restrictions on content and how content can be made and shared, consider this: the iPad is not shipped with a stylus, nor is there any place to store one if you buy one in a store. A simple image of what they’ve done is to go back to what babies can do (a good idea to start) but not allow the baby to grow up and learn to do more (a really bad idea). Seymour Papert used to say “low threshold, no ceiling”. The iPad is “low threshold, very low ceiling” (too low to stand up).

    • 3. Jeff Rick  |  April 30, 2013 at 10:47 am

      Hi Alan,

      you’ve misinterpreted my post a bit. My guess is that it comes down to what we mean by media. I view media quite broadly (a la McLuhan) and would include things like genre in definition of a medium. Novels and twitter posts are both examples of writing; however, I would argue that they are pretty distinct media: different structures, different strictures, different practices, etc. Twitter enables symmetric authoring (i.e., most people who read tweets also author tweets). In contrast, most people do not respond to a novel with another novel. That medium isn’t normally symmetric. My main point is that’s okay. Most people don’t need to author a novel. We can have media, such as Hollywood movies, that are not symmetric.

      Now, that doesn’t imply that there shouldn’t be authoring tools. Ideally, people should be able to write novels or create full length movies with their iPads. We’re definitely not there yet. That said, we are starting to see more and more authoring tools on iPads. I have three programming languages on mine. None of them are “no ceiling” programming languages (e.g., they can’t be used to program other iPad applications) but we’re making progress. Video editing through iMovie is actually pretty cool and intuitive. Scrivener is about to unleash their writing program on the iPad, so novels will be a possibility. Progress is being made every year.

      In your example, the tools that the teacher used for fourth graders to author books were not the tools that people who write books for fourth graders use. They don’t have to be. There’s no need for symmetry there. It’s okay for there to be good authoring tools for adults writing for fourth graders that are completely separate from tools that fourth graders use to compose books.

      There is this concept of a knowledge building environment which relies on the concept of symmetric media, where the environment allows consumers to produce and share with the community. I agree with you that those are worth pursuing. Twitter, wikis, and academic journals all succeed here. The Hypercard system you mentioned also seems to have this property. So far, the iPad has too few of those, though there are notable exceptions (e.g., Instagram).

      Our main disagreement is that I think the hardware is not the problem. Rather, I think the software, developer toolkits, and usage culture is not taking great advantage of it yet. Currently, there is too much remediation (making the new technology behave like an older one rather than finding a good fit between what the technology allows and what the user wants to do). If you were given the challenge of how to enter text into an iPad starting from scratch you would never come up with a QWERTY keyboard. Yet, that’s what currently exists. On Android, there are some novel text entry techniques that are starting to take off. I think they will ultimately lead to tablet text entry that is faster than keyboard text entry. It may take a while to do that, but a significant time also elapsed between the invention of the mouse and the WIMP interface. We need to give it time to develop. Perhaps there will be advances in hardware technologies (e.g., some Android tablets do support pens) but, to me, the challenge lies in software. Having programmed multi-touch interfaces for over five years now, I can say that it has tremendous potential; however, there’s still a long way to go. Symmetric authoring is definitely a good goal; it just doesn’t have to be the only goal.

      • 4. alanone1  |  April 30, 2013 at 1:36 pm

        Hi Rick

        Please explain what you think McLuhan meant when he said “The Medium is the Message”, and then I’ll have more clues as to how to reply to the above.

        • 5. Jeff Rick  |  May 13, 2013 at 10:50 am

          In general, McLuhan meant that the systems by which we inscribe and transmit information (a medium) have a strong effect on what gets conveyed (the message). In Understanding Media, he builds on Innis’s The Bias of Information to talk about the societal effects of adopting a medium (i.e., that the medium that a society uses does more to shape society than what people try to do with that medium). In The Medium is the Massage, he concentrates more on the individual level (e.g., that this exchange is fundamentally different than if we were doing it face-to-face). McLuhan was also quite expansive in his definition of media, including clothing, housing, money, bicycles and weapons in his analysis of the societal effects of various media. For example, he considers both the typewriter and the written word as media, though one is an authoring tool for the other. He separates television and movies, though the former can be used to broadcast the latter. I’m not sure whether that helps.

          One media notion to consider when analyzing the iPad is information bandwidth. The multi-touch display takes in more bits of information than a mouse and keyboard. Of course, if all the interface allows is button presses and vertical scrolls, the input bandwidth doesn’t translate into a high interaction bandwidth. Multi-touch is still new enough that we haven’t figured out how best to map input to interaction. Currently, we still have it do very mouse-like and keyboard-like interaction. We can and should do better. As we make progress on that, authoring (requiring more sophisticated input from the user) will be a natural consequence.

          • 6. alanone1  |  May 13, 2013 at 12:22 pm

            To Jeff Rick

            Just so we don’t get into a fruitless argument about what McLuhan really meant … I asked you for your interpretation to see what you thought McLuhan meant, not to find out for myself. This was to help me formulate as short a reply as I can.

            Some years ago I was invited to deliver the McLuhan Lecture — sponsored by the Canadian Government and the McLuhan family. At the pleasant lunch before the talk I got a chance to chat in depth with Eric McLuhan (Marshall’s son and the director of the McLuhan Institute). This was not quite as wonderful as Woody Allen producing Marshall himself in “Annie Hall” to settle an argument, but close enough! And he confirmed that the interpretations I’d formed were accurate from my studies of McLuhan many years before.

            In a nutshell, if McLuhan had wanted to convey the idea that the medium that carries a message modifies/filters the message, he would have said so (he did think this, but this wasn’t the big idea). Another thing he *did* say is that “We become what we behold”. To cut to the chase, when he said “The Medium is the Message” he was equating the two terms as much as can be done in a metaphor.

            What he meant was: the real message (impact) of a medium is what a human being has to become in order to use it (and especially to communicate through and with it).

            In hindsight this should be completely obvious because we can only deal with things our brain/mind has models of (even if, as is often the case, they are bad models). The process of learning all the way to fluency produces many skills that manifest below the conscious levels (think of the transitions when one learns to drive a car), and it is these invisible mechanisms that do much of the interpretations of signals that we inaccurately call “normal” and “reality”.

            But, even though “obvious”, the invisibility of “we become what we behold” renders this idea as quite unobvious.

            In any case, when I finally grokked this I thought McLuhan’s idea was one of the great insights of the 20th century.

            And, this is why Anthropologists discovered that a literate culture thinks qualitatively differently than an oral culture, and most interestingly, than an oral culture that has recently acquired a writing system.

            This means: when we design a new tool, and especially a communications medium, and especially one that will be “wall to wall” in some way, I think there is a real ethic that says “we have to try to understand what people’s minds will become if they get fluent, especially “centrally fluent”.

            This is what I meant when I said years ago that “Television should be the last mass medium allowed to be distributed without a Surgeon General’s warning”. I meant that the new meta-medium of the computer would be much more far reaching, and potentially much more dangerous to “civilized thought” than TV — just as we inventors thought that it could also go far beyond the printing press in a positive direction if carefully designed.

            All of these ideas were well understood in the late 60s and through the 70s by many of the researchers who invented the technologies we have today.

            And, because both the mouse and a very good stylus/tablet were invented in 1964, we got a chance to look at both “on-stage” practicalities and affordances, and also at the “back-stage” influences. Because a good multi-touch system was invented in the 70s, we had a chance to understand what was great vs what had to be done additionally to avoid a system just for babies.

            And, because many “pretty good” authoring systems got invented in the 70s through the end of the 80s, we had a place from which to criticize the distressingly retro hacks in the web browser, and now in consumer devices such as the iPad.

            Defenders of both say “we are getting there” (as you did about the iPad) but the manifest point is that we were already at the first few plateaus with regard to authoring, but now there are millions of dumbed-down devices resetting “normal” and “reality” for millions of people, especially children. “Getting there” in a consumer pop culture is very often asymptotic
            and thus only represents the illusion of progress.

            In any case, if you read the article, I was simply asked whether the iPad is a Dynabook, and I said it isn’t.

            Adding a stylus would be good, but it misses the main point: that by Apple policy, children using Etoys and Scratch (and other authoring systems in which content contains code) are forbidden by Apple to share their creations with others over the Internet.

            That is not just anti-Dynabook, but anti-ARPA, anti-Internet, anti-PARC, etc. There are defenders who say “well, they could email their content”. But do you really think this is remotely the same “medium” as e.g. the nice Scratch website that is organized a bit like YouTube that allows millions of children to share and show off?

            Now, it happens that 45 years later I don’t think my original “service model” for the Dynabook was quite strong enough either. One would hope that the aspirations and visions of today would be much richer than the ones we had back when there were not a lot of examples to look at. Today we have lots of good examples of what things could and should be like, but the pop and consumer culture is disinclined to look.

            But that is a very different discussion from this one.

          • 7. Jeff Rick  |  May 15, 2013 at 11:44 am

            Hi Alan,

            I very much appreciate your latest reply. We are in agreement on McLuhan. I 100% agree with what you wrote and I like your formulation (and look forward to using it in the future). I did not understand what you wanted to know from me with that question. I thought the disagreement was about the definition of the medium, not on the effect / mechanism that you describe. I also agree that this was a critical insight and it has inspired my own work (e.g., medium-based design).

            As far as the limitations of the iPad, you are right. It does stink that you can’t (really) share code. It does stink that it is mainly used for delivering consumables. It is not so much that I disagreed but that I had two follow-up points to make.

            First, that the software situation is improving. Initially, Apple wouldn’t even allow you to run a system with an interpreter on the iPad. Their policy changed. Initially, there were no programming environments allowed on the device. Now, there are several. Initially, Apple marketed products mainly about consuming media (books, music, websites). Recently, they added products about authoring media (iMovie, GarageBand, Numbers). So, progress is being made. Is it currently the Dynabook? No. Would it be better if it were more Dynabook like? Yes. Is there significant progress in that direction? I think so; you seem more negative there. That inspired my post. While I appreciate the criticism, I think a call for action (e.g., developers can and need to fix it) is helpful.

            Second, that the hardware is not the limiting factor. I agree that many current multi-touch interfaces are subpar; however, based on my own experience (I’ve created nine innovative applications for multi-touch in the last five years), this is not based on an inherent deficiency in the technology, but rather an immaturity in knowing how to appropriate that technology. Developers / researchers should put more time into inventing that future. I am doing that myself. Unfortunately, the EU funding bodies disagree on the value of authoring systems for multi-touch. Both of my grant proposals on that topic were rejected and I’m having to work on this relatively ambitious project in my limited spare time. If I can achieve a proof of concept (it is novel enough that I don’t know whether it will actually work), I’ll send it along. It is very much inspired by your insights (authoring, the importance of message passing, media changing us, etc.).

            So, in conclusion, there were some misunderstandings in this discussion. My guess is that we would have avoided them, or resolved them better, face-to-face. Your work / insight continues to inspire my own work and I appreciate you taking the time to engage in this discussion.

  • […] Go read the whole interview, it’s worth it. (via Computing Education Blog) […]

    • 9. alanone1  |  May 15, 2013 at 11:55 am

      To Jeff Rick

      But the first progress you mention (allowing interpreters) came only because I personally spent a long time with Steve Jobs using my long term friendship with him to have him “intervene as Monarch”.

      And he did not go further than that to fix the sharing mistake (he did have serious health problems even then, but he might not have anyway).

      I think that Apple is now a company that will only do the right thing if there is a competitive reason for it (and that’s not progress).

      This is the distinction between marketeers catering to wants and educators trying to deal with needs. Most things people really need, they don’t want. So you can’t use market forces to make progress here.

      I like the idea of resolving through writing — both the pace and the form really help thinking (a point that I’m now sure we both agree on!)



      • 10. Jeff Rick  |  June 3, 2013 at 12:36 pm

        Interesting and, unfortunately, depressing. I had always figured that the interpreters rule was to combat Flash. There was a real chance that the iPad would be just a Flash vehicle and that the opportunities to create something revolutionary would be hampered by Flash’s limitations. In that sense, I thought it was somewhat reasonable. My guess is that the current rule about not sharing code is based on monetization. They want their 10% profit of all Apps created for iOS. If you could install arbitrary code, then you could circumvent that revenue stream. While that may be a theoretical possibility, it is practically silly. The average person will download things on the official AppStore. Only geeks would go outside that and they are likely to jailbreak the device anyway. So, there is no real benefit of the rule to Apple that I can see (lawyers might disagree). Perhaps you can befriend Tim Cook 🙂

        • 11. alanone1  |  June 3, 2013 at 1:32 pm

          I would settle for them just getting Javascript in the browser to be as fast as it can be on the iPad.

          We haven’t done this test recently, but a few years ago we looked at speed ratios for app code between the MacPro and the iPad and found that that the normalized ratio for Javascript benchmarks was a factor of 16(!) slower. This said to us that they just don’t want anything like a decent app being delivered in the browser and have made Javascript on the iPad much much slower than it could be.

          We brought this to Apple’s attention also and they professed “surprise”. It would be interesting to see what the ratios are today.

        • 12. Mark Miller  |  June 3, 2013 at 10:42 pm

          My guess is that the current rule about not sharing code is based on monetization. They want their 10% profit of all Apps created for iOS. If you could install arbitrary code, then you could circumvent that revenue stream.

          Ew…Kind of the GNU license in reverse… I hadn’t thought of it that way before.

          My sense of what Apple has done is they want a predictable IT environment, which they think makes their devices more attractive to consumers. No worries about viruses. No need to refresh the system in case of infection. To make an analogy, this is the way many school district IT systems are (unfortunately) organized. About 7 years ago, now, the school district in my area went through this transition. They were mostly using PCs, but a significant number of school classes had been using Macs. There was a central district IT dept., which offered maintenance services, but some of the administration of the computers was done within each school, with technically skilled teachers, and volunteers. The school district decided to do away with that, and replaced all the Macs, and existing PC systems, with locked down Windows PCs that could be centrally administered by the school district’s IT dept. No software could be added to the machines (though a couple of the high schools, which taught programming classes–and still do, were probably an exception). Every computer had the same suite of software on it, or one of a set of standard configurations. If any machines got infected via. a malicious website, they could wipe the hard drive and refresh from a DVD image of the OS and installed applications, or perhaps do this over a network. This effectively turned every computer in the district into more of a “terminal,” though there was still some “computing” going on, in the sense of each computer running its own “office” software, and whatnot. What they wanted to achieve was administrative efficiency using an insecure platform. That, in my mind, is what Apple has tried to achieve as well with its App Store, though they’ve added in performance considerations, along with their proprietary interests.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.

Trackback this post  |  Subscribe to the comments via RSS Feed

Enter your email address to follow this blog and receive notifications of new posts by email.

Join 6,164 other followers


Recent Posts

Blog Stats

  • 1,621,404 hits
April 2013
« Mar   May »

CS Teaching Tips

%d bloggers like this: