Knowledge Navigator Revisited

Posted:
in Future Apple Hardware edited January 2014
These videos were produced in the late 1980's -1990's by apple computer.



Given the state of the art today can this be achieved?



by any cost?

by a reasonable cost?



http://www.bu.edu/jlengel/kn65kfs.mov



http://www.billzarchy.com/clips/clips_apple_nav.htm

Comments

  • Reply 1 of 13
    Is ti just me or does that look like Bill Nye?
  • Reply 2 of 13
    tednditedndi Posts: 1,921member
    Quote:

    Originally posted by psgamer0921

    Is ti just me or does that look like Bill Nye?



    sorry, who is bill nye?



    and also sorry for the treble thread that ended up getting posted please delete the copies.



    I am a Spaz today.
  • Reply 3 of 13
    jeff79mjeff79m Posts: 37member
    Bill Nye, Bill Nye The Science Guy...Bill, Bill, Bill Nye.......



    He used to have a kids science show on PBS it was great.





    As far as the idea I really am not ready to talk to my computer.
  • Reply 4 of 13
    tednditedndi Posts: 1,921member
    Quote:

    Originally posted by jeff79m

    Bill Nye, Bill Nye The Science Guy...Bill, Bill, Bill Nye.......



    He used to have a kids science show on PBS it was great.





    As far as the idea I really am not ready to talk to my computer.






    oh.





    I would talk to my compute and have it talk to me. I have a nice female voice on it now and she tells me the time and things that need my attention. If she looked really good and spoke to me in avatar like fashion I could really dig it!!



    how good is voice recognition these days? I can talk to the fedex computer system by phone now.



    They could then sell agent looks on itms voices and appearances for customization.



  • Reply 5 of 13
    dhagan4755dhagan4755 Posts: 2,151member
    Geez, we're mighty close with that tech now! A lot of the things demonstrated in the Tiger demo are pretty close to the foundation of what was shown in that video. And in 6 or 7 years that video might be reality!



    The video gave a pretty good example of how the spoken interface would integrate with what Tiger is incorporating...automater, spotlight, iChat AV, iCal, Address Book, and on and on. Tiger could be seen as the foundation of what was shown in that video...
  • Reply 6 of 13
    tednditedndi Posts: 1,921member
    the video was produced in 1988



    before html and the internet (although hypertext was there with hypercard)



    google for searching didn't exist now, finding the video on google was like the guy in the video's query "apple computer quicktime knowledge navigator"





    also, Steve Jobs hired John Scully away from pepsi cola by asking him how many years of your life do you want to spend making colored water when you can change the world.



    ARM processors from the early newton are in the ipaq's running pocket pc



  • Reply 7 of 13
    rolandgrolandg Posts: 632member
    Funny to see that Knowledge Navigator again.



    I am studying inforation systems (don't know the exact equivalent; it's a mixture of econimics, computer science and information systems) and one day a tutor showed us that exact same video clip to show us what ideal we were supposed to follow - making an executive's life easier and more efficient.



    I think that what keeps this vision from being put into a real product is the fact that AI is not advanced enough. How should the device know which information is important to you and which is not? Structured problems can be solved fairly well be computers by now but unstructured ones not.



    The product would not work as disired right out of the box. It would need to learn and understand how you think, what your preferences are, your opinions. It would need to predict these things accurately in order to be a useful tool.
  • Reply 8 of 13
    tednditedndi Posts: 1,921member
    Quote:

    Originally posted by RolandG

    Funny to see that Knowledge Navigator again.



    I am studying inforation systems (don't know the exact equivalent; it's a mixture of econimics, computer science and information systems) and one day a tutor showed us that exact same video clip to show us what ideal we were supposed to follow - making an executive's life easier and more efficient.



    I think that what keeps this vision from being put into a real product is the fact that AI is not advanced enough. How should the device know which information is important to you and which is not? Structured problems can be solved fairly well be computers by now but unstructured ones not.



    The product would not work as disired right out of the box. It would need to learn and understand how you think, what your preferences are, your opinions. It would need to predict these things accurately in order to be a useful tool.




    As you use the interface it should be able to learn your preferences. When upgrading to another unit the old unit with all of your "knowledge" would then transfer like your settings and mail accounts.
  • Reply 9 of 13
    rolandgrolandg Posts: 632member
    Quote:

    Originally posted by TednDi

    As you use the interface it should be able to learn your preferences. When upgrading to another unit the old unit with all of your "knowledge" would then transfer like your settings and mail accounts.



    True, but people expect things to "just work".



    A little adaption time might be tolerated: Dictation systems are pretty advanced today but not very commonly used. I guess this is to the "learning time" it takes to adapt to the user's voice.



    Further more, they make still make to many mistakes. The texts need to be revised. It might be the human nature, but every word the system does not understand properly is annoying because you would definitely be faster using a keyboard.



    And this is just one task. And a "simple" one at that. Simple in the sense that your computer does not need to interpret what you said. I wonder how far along Apple's upcoming Voice Command system is.



    The more tasks, the more mistakes. The more mistakes, the more likly it is the system will not be accepted.



    It would be possible to train a system the way you educate a child. At first, you would do a lot of things manually, showing the system what objects are used to accomplish a certain task and what the connections are between them.



    A child will adapt what it learned to similar situations by itself, and I am not too sure AI is advanced enough to this.



    Sure, you could tell your system to apply certain acquired techniques to a given problem.



    The system would have to think exactly like you do. It would have to concentrate on certain things due to its (still) limited capabilities as you have to, too. But what happens when an issue from the vast field of what you had to abstract from gains relevance for you? Something completely unexpected. How can you teach your system to account for something like this in advance?
  • Reply 10 of 13
    tednditedndi Posts: 1,921member
    yes but it could tell me my mom is on the phone.



    the learning curve could be long and flat going from one os to another culling prefs along the way. by os 15 or so perhaps it could be a reality.



    or I could ask it to find a user in my address book and start ichat.



    not too close to the level of the knowledge navigator yet.



    the video is in ichat and qt 7 is getting closer tho.
  • Reply 11 of 13
    rolandgrolandg Posts: 632member
    Quote:

    Originally posted by TednDi

    yes but it could tell me my mom is on the phone.



    the learning curve could be long and flat going from one os to another culling prefs along the way. by os 15 or so perhaps it could be a reality.



    or I could ask it to find a user in my address book and start ichat.



    not too close to the level of the knowledge navigator yet.



    the video is in ichat and qt 7 is getting closer tho.




    All of what you say could be done with todays technology, indeed, and most of it is allready implemented.



    But from what I recall from the Knowledge Navigator (and that is the essence of the system, in my opinion) is that it digs through hugh piles of data to find the information that is relevant for you and puts these informations into context to form knowledge. That is then at your disposal. For example you coming to the kitchen table for breakfast having your KN report to you what happened while you slept and what consequences should be taken on this basis.
  • Reply 12 of 13
    curiousuburbcuriousuburb Posts: 3,325member
    Apple's concept video for the intelligent assistant in Knowledge Navigator shows one direction.

    Decades later, both the vision, and the conceptual piece itself are still hailed as ideals.



    Microsoft took the idea of assistants in another direction.

    Bill's then-fiancee Melissa was project leader for the patronizing atrocity that was Bob.



    Bob lives in zombie form that everybody now knows and universally despises as Clippy.



    Oh please, let Apple's vision be the one that gets followed.
  • Reply 13 of 13
    tednditedndi Posts: 1,921member
    bob is absolutely not sexy, Apple could produce some sexy assistants I am sure!
Sign In or Register to comment.