Saturday, July 18, 2009

BunkerSofa's evaluation of smart.fm from a memetic standpoint




I have followed the progress of smart.fm since it was originally called "iKnow.jp".

Even if smart.fm appears and is mentioned by some members of the japanese IT community as one of the hottest web start-ups in Japan,

I would like to argue here, from a logical and bunkersofist standpoint, that there is nothing exciting in the current version of smart.fm and that it will predictably fail to be a standard for its lack of usefulness, simplicity and coolness.


My only purpose here is to evaluate the meme constituted by "smart.fm" from a personal perspective. It is not at all to bump off the diligent team behind it.
Ultimately if this post can contribute to the improvement of 
smart.fm, that would be great


After a quick review, I managed to extract several patterns underlying the product's features and functions. These are the following:



1) On the notion of Sharing:


One of the greatest properties of the Internet is that, once something is on the web, it is potentially for everybody.

With smart.fm, this "something" is supposed to be a list of items (chunks of knowledge?).


But, is a list of items (made for learning purposes) something you want to spontaneously share?

Is it shareable?

→ I do not think that sharing a list of items is a powerful meme because, when you have bothered building a learning list of items just for your own learning and training purposes, you want to hold it only for yourself.



2) On the notion of Learning:


Can Learning be reduced to memorizing a list of items?

→ I do not think so.


If smart.fm can do something, it is only that it might help you memorize a list of items.

So instead of the current catch phrase, I find that "smart.fm, the place - you go to memorize?" would be more appropriate, if any.



3) On the principle that smart.fm is supposed to be based on.


Apparently, smart.fm needs a 4:45min-long video presentation to explain what it does, which suggests that there is nothing simple in smart.fm

Below are some key-words as well as key-phrases that I extracted from this video (http://smart.fm/tour/) and that are used to describe the service:


"proven learning principles combined with web technology to improve the way we learn"

→ But exactly which principles?


"neuroscience", "cognitive science"

→ Sounds like narrative fallacy...


"you can learn more quickly and remember longer", "spaced rehearsal over increased period of time, "optimal moment of review exists"

→ Is Learning only about remembering,repeating and reviewing?


"Ebbinghaus curve"

→ What's that?


"optimal moment before you are likely to forget it"

→ Is that the core principle?


"evaluating what we know and what we don't know is notoriously difficult"

→ What does this sentence mean?


"people don't spent the right amount of time on the things they need most to review"

→ I agree on that.


"where you learn something, you want to really know it"

→ The meaning is not clear.


"smart.fm, your adaptive learning platform"

→ Is that why it is called "smart"?



I believe that Reality is by essence complex.

From that standpoint, something is useful if and only if it has managed to extract simple underlying patterns from that complexity;


My question is: what is the simple underlying (but powerful) principle that smart.fm discovered and is based on?

→ I think that in fact, there is none.



4) About the culture brought by smart.fm


- Why using a ".fm" domain name?

This has nothing to do with a radio.

I hypothesize that smart.fm has wanted to be naively associated with the success of "last.fm"


- Equally, I think that smart.fm is imitating a little twitter in his design and wants to take advantage of the successful memeplex constituted by twitter:

* current layout of smart.fm's homepage looks like twitter's one with a large rectangle

* use of a bird as a mascot

* use of blue as main color

* use of the first letter of the product's name as favicon.



- There are too many "tools" and functions

What to choose between Iknow, brainspeed, dictation, items, sentences, journals, goodies, partner series, search lists, showcase lists, SNS functions etc...?


I am confused and lost because smart.fm seems to mix many features of other services without having its own identity.

Again, what is smart.fm?

→ I don't know. That is too complex for my understanding.


Remember: Less is More and More is Less.



5) About "money"


I heard that in total, the project has got way more than $10M in funding...!

That seems to me relatively enormous, huge and way way way too disproportionate...


Also, I heard that the business model would be on "Ads" because smart.fm is "free" for its users...


From a logical standpoint, I do not think that rough and noisy "ads" is the proper way to monetize for this category of learning websites.

→ When you learn, you may not well want to click on ads when you are studying seriously.




As a conclusion, I am very doubtful about the viability of the meme "smart.fm".

As a matter of fact, in order to be successful, a product does not have to be better, but just a powerful meme.


From the memetics point of view, smart.fm has the following scores:

- heredity: LOW

What smart.fm does is unclear; thus, the human brain is confused.

It does not know what to copy and as a result, retains nothing substantial from smart.fm

- fecundity: VERY LOW

The service has no simple and powerful principle; thus the service cannot be a fertile substrate on which you can build something richer and from which technology can evolve.



- longevity: LOW

I predict the service will be made obsolete within 1 year to 2 years.


For example, just adding a function of subtitling (with text selectable) on YouTube would allow viewers to learn foreign languages with TV shows they like or personalities they love.

I believe that this alone blows away already the current platform of smart.fm .


As a matter of fact, (and that is a simple statement but very powerful) LEARNING is all about IMITATION. Imitating from other people. And this process is all the more efficient that you imitate people you like, e.g. singers.

To convince yourself, consider how many american girls learn Japanese imitating idols like "ARASHI" :) 



So currently, smart.fm is arguably a very weak meme because:

*it has too many flaws.

*it is based on a wrong assumption, namely "LEARNING is MEMORIZING chunks of knowledge"


As a result, I really doubt that I can get smarter at smart.fm.

In fact, I might go a little more numb than I might possibly already be^^

18 comments:

  1. Very interesting analysis, julianamonik

    I agree with most of your remarks. And I would like to make the following comments:
    - sharing a list of items is not a problem assuming that people are altruistic and generally speaking they are.
    - from my perspective, the underlying principle is displaying the words (chunk of knowledge) that you forget easily more frequently than the others
    - instead of calling it a place for learning, I definitively agree that it should be called a place for memorizing and I would say memorizing single items. It is not a good thing to pretend to do something that you do not provide...it s a marketing fallacy, I believe.
    - I totally agree on the fact that learning is imitating not remembering.
    - I predict that smart.fm will be obsolete within one year or so if they do not change the way it it...

    And I found the system very annoying. You have to adapt to the pace of the algorithm. And it gets mechanical. It should be the other way around. The machine adapts to the human brain...
    So "adaptive" yes, but unfortunately in the wrong direction, at least from my perspective...

    ReplyDelete
  2. Just a suggestion: how to blow away smart.fm☆

    - Take this video (http://www.youtube.com/watch?v=9Mg-HjqDcAM)

    and

    -Superpose its lyrics (http://www.uta-net.com/user/phplib/Link.php?ID=78442)

    in real-time to the video,

    -Make each word [clickable] to edit the [meaning] and you got something that blows away smart.fm

    ReplyDelete
  3. potentially blow away; have to formalize that a lot though;
    that's an intuition;
    Paskal, that's your job!

    ReplyDelete
  4. yes, i will be glad to formalize that.

    ReplyDelete
  5. Nice article. I am agree that the only underlaying principle i understood, that is "leaning = memorizing", is questionable for us foreigner. But this is the japanese way of teaching or learning. And according to my past experience in the e-learning field, all the specialists in learning/teaching science were agree on the fact that the learning = memorizing principle is an old and non efficient principle. however they were less agree about learning = imitating or understanding ? In my point of view, it's both; you start imititating, like a child mimic is parent and later on, when you are familiar with a concept, you understood it, you start developing your own style. And that would have been great achievement as learning plateform principal/ base algorithm: Be able to define automatically when the assimilation phase (imitation) is finished and when start the validation, reinforcement phase (understanding).

    I would also say that i way more impressed by lang-8 that is technologically not advanced at all but the base principle is great; exchanging knowledge.

    ReplyDelete
  6. I agree with paskalamonik on the notion of sharing. Come in, do you know the open source community or web 2.0?

    if some people can be altruist enough to share a complet OS kernel or their playlists, there is no doubt some will share a hundred words long list.


    On the notion of learning, there are areas where remembering a list of item is a huge part of the learning process. To my mind, to learn a language you have to spend a lot of time just trying to remember words (or damned kanjis !!). however it's true that is can be simplistic to see learning as memorizing.

    on the following topics, it's amazing that they managed to get funding for 10M$ when doing a copycat of hype things for UI, and not being able to explain the vision of the service simply.

    Yep I agree, the valuation is way to high !

    Alex

    PS : found a post that may be of interest to you and to ascott project members :
    http://unqualified-reservations.blogspot.com/2009/07/wolfram-alpha-and-hubristic-user.html

    ReplyDelete
  7. Pretty much agree with you on this although I think remembering can be learning. The way they do it as proposed by superMemo is as well a proven way to mangage a huge flash card list. Remembering words is one of several things needed when learning a language.

    BUT 10M$ is far beyond the value I think they should need... but I heard their burn-rate is pretty high (not very surprising since their office is in a super expensive location).

    This tells us though that there are people out there, at this moment, virtually giving money to start-ups. That gives us hope.

    ReplyDelete
  8. An excellent piece of analysis. We need far more from you. Nobody else seems to be even attempting to analyze "new" products in this level of detail. Many, many thanks.

    ReplyDelete
  9. To paskalamonik:

    Still in the framework of memetics:
    Yes, I think you are right that people are altruistic enough to share their work and "want" to spread their memes in general;
    So a more exact way to put it is:
    Is an [list of items] a good meme, that can easily spread form one brain to another?

    A way much powerful meme: a video with clickable subtitles

    Remember: there are way much more memes that there are brains out there; only the most powerful memes will spread and survive.

    ReplyDelete
  10. To v3n3

    imitation+knowing when to reinforce = maybe it is way to explore though I don't get intuitively why it should work...

    It does not prove anything but a simple example to get intuitively the idea: consider one moment how many chinese children manage to learn naturally Chinese language without memorizing artificially anything and without knowing even how to write/read.

    I think we have to exploit that the brain is a big meme machine, whose speciality is to imitate

    ReplyDelete
  11. To Alex,

    indeed the link you provided is interesting;
    However, the article is relatively very difficult to decipher;
    Mencus Moldbug is quite... confused in his ideas
    Some good insights appear though:

    1) The notion of Control interface:

    "A screwdriver is a control interface because if I am screwing in a screw and I turn the handle clockwise, I expect the screw to want to go in. If the screw is reverse threaded, it will want to come out instead, confusing me dreadfully. Fortunately, this mapping is not random; it is predictable."

    2) Control interface must not be intelligent

    "Briefly, intelligent user interfaces should be limited to applications in which the user does not expect to control the behavior of the product. If the product is used as a tool, its interface should be as unintelligent as possible. Stupid is predictable; predictable is learnable; learnable is usable"

    3) The Graffiti effect

    "If the user has an affordance by which to tell the algorithm the purpose or category of her search, the whole problem of guessing which application to direct the query to disappears and is solved perfectly. A whole class of category errors ceases to exist."


    4) Prediction done by the article, that encourages the Ascot Project

    "My guess is that if there is any "next thing" in search interfaces, it will come not from smarter UIs, but from dumber ones in which the user does more work - the Graffiti effect. If a small quantity of user effort can produce a substantial improvement in user experience (which is a big if), the user will accept the bargain. Hey, it made Jeff Hawkins rich."

    ReplyDelete
  12. To AJ

    Intuitively, remembering is of course part of the process of learning since you have to do save data somewhere in your brain to "learn"

    I just have some serious doubt on the necessity of forcing artificially and mechanically the task of memorizing...

    I once heard that the budget was around , actully, $30M....

    Exactly, as you suggest, there is plenty of money out there...

    ReplyDelete
  13. To Anonymous,

    Thanks for the encouragement^^
    We will for sure keep on criticizing memes and produce plenty other analysis!

    ReplyDelete
  14. What I think was exactly advocating is achieved by www.englishcentral.com a webservice i just came across randomly:

    My opinion:
    "For example, just adding a function of subtitling (with text selectable) on YouTube would allow viewers to learn foreign languages with TV shows they like or personalities they love.
    I believe that this alone blows away already the current platform of smart.fm ."

    Now look at:

    http://www.englishcentral.com/ja/player/10240/Steve+Jobs%3A+What+is+Truly+Important%3F

    I think it is really cool and needs no any more comment, to know its value compared to smart.fm

    ReplyDelete
  15. terrific, julianamonik.
    you have been telling me that such a service (putting subtitles on videos (songs, speech, movies scene)) will blow away smart.fm, at least from a bunkersofa perspective, for at least 6 months and now englishcentral has done it.
    congrats julianamonik for your prediction based on memetics, singularity and computation.

    To the smart.fm team.
    The game is not over yet. You can use the opportunity to question your service and improve drastically.
    or you can buy out the service :)

    No seriously, it must be a strong stimulation for you guys.

    gambare.

    ReplyDelete
  16. Also, I really think there is a big difference between the current "smart.fm" and "englishcentral"

    In smart.fm, there are videos of steve jobs or Madonna. But you cannot hide English or Japanese, you cannot check the dictionary, and most importantly there is no indexation according to the part you wanna study.
    I think that if you wanna learn English, you d better use englishcentral because the starting point is what "do you want to watch?". It is not "what do you want to memorize as words?"...
    I think that 2) is appropriate for electronic computer, at least for current ones. But it s not for electrochemical computer, at least for current ones. :)

    The hypothesis underlying smart.fm that says that 2) is the right question is incorrect.

    And this is what happens when you do not do primary mental simulations, almost using a mathematical or logical approach...

    ReplyDelete
  17. Really, I always get some nausea whenever I try smart.fm.
    Not because of the people behind the system.
    I have nothing wrong against them. We even have a strong common point with some of the team since we re going to the singularity summit in oct (NYC).
    No.The nausea is induced by the fact that my brain cannot tolerate the following statement:

    "mechanical memorization = learning"

    I don't know the role of memorization in learning but my brain feels that s a way too much in smart.fm case.

    Have a look at the Japanese prefectures learning session: http://smart.fm/lists/118955....

    ReplyDelete
  18. it's now official. cerego raised more than 23M USD since 2000 for smart.fm.

    but cerego is different from smart.fm

    smart.fm has raised more than 3M USD from NTT docomo:

    http://www.techcrunch.com/2009/09/10/cerego-raises-3-4-million-for-smart-fm-launches-facebook-friend-quiz/

    ReplyDelete