Please create an account to participate in the Slashdot moderation system

 


Forgot your password?
Close
typodupeerror
×
Science

Online 'Sand Mouse' Tests Neurobiologists 86

Posted by timothy from the can-it-understand-"one-click"? dept.
The Metahacker writes: " A Princeton professor and his former student have created a 'mouse' (really, a neural net) that recognizes the word 'one' as spoken by a variety of speakers. The interesting part? They're challenging the neurobiology community to discover the mechanism it uses, using only the tools available to analyze live patients - observation and experimentation. You can upload your own sound files to test the mouse, and view experiments other scientists have performed. Cash prizes will be awarded to those who explain the mouse's behavior or can train the same number of neurons to perform a new task. You can read the New York Times article about it (free registration), or go directly to the site."
This discussion has been archived. No new comments can be posted.

Online 'Sand Mouse' Tests Neurobiologists More

Online 'Sand Mouse' Tests Neurobiologists

Comments Filter:
  • THAT's Informative? Sheesh, there goes the neighborhood.
  • The point of the neural net these people have set up is not to have a neural net that recognizes words well; nor is it to have a neural net to recognize words in noisy environments from different speakers. Rather, the point of the research is to study how biological neurons can process sensory information over time to produce a sensation of a "moment." Recognizing the word "one" is an example of such temporal sensory processing, and a relatively easy one to test, so that's why they chose that task for their research project.
  • I dread the day that computers, particularly handhelds, get truely useable voice recognition capability. Then, besides the ass sitting in the restaurant bellowing into his cell phone, I can look forward to legions of idiots doing the same with their Palm Pilots. Instead of the business dweeb clicking away at his laptop keys in the next airline seat, he'll be talking at the damned thing. Time to invent that portable HEMP generator.
  • "Science is like sex: sometimes something useful comes out, but that is not the reason we are doing it"

    Richard Feynman
  • The name of the link, "partners", seems pretty self-explanatory to me. It's for partners of the NYTimes, which I would be fairly confident Slashdot is not one. I assume this is so other news outlets or "partners" of the NYTimes can link to their articles without the registration info.

    Is it a stupid way of doing it? Yeah, since there's no authentication check to see where they are coming from. Does that alone make the theft of their service right? That's where I disagree.

    Just because a car is unlocked and has the keys in it doesn't give you the right to drive off with it.

    I don't know.. I see both sides of this, and it all comes down to morals, which are in serious decline in this society anyway. They provide a service, and in return for that service, they ask for some information. Not even $, just information. I don't see the big deal.

    As a webmaster myself, I would be ticked off if I offered services requiring registration, and people got around it, viewing my content for essentially nothing. But then again, I'd probably be smart enough to have some kind of check in there at the very least.

    Reminds me all of the Dilbert comic:

    "What do our customers want?"
    "High quality products for free."

    Such is the way of the Internet. :(

  • They don't ahve a right to your registration information... you can always choose to not read their information.
    --
  • Except for the running upstairs scared shitless part all of this happened to me too. Then I realized that if I didn't run around the room, thrashing madly while playing, not nearly as many 'interesting' things would happen.
  • Are you saying John Hopfield should be the authorized examiner of neuroscientists? Absolutely not. Who said anything about athorising anyone? His experiment can be peer reviewed just like any other work. Time spent working on this problem is time not spent working on the brain directly. No - but it might be time spent sharpening the skills required for brain research. Then again - maybe not. That's the nature of research.
    --
  • Beyond a certain point - after you've taken your last exam - academics are no longer accountable to anyone. Sure - you get peer reviewed. But what happens when you and your peers all belong to a clique that have a vested interest in promulgating a particular scientific dogma?

    I'm not saying that I don't agree with your post, however, at least in the physics community (which I know best, I assume other sciences are similar), as long as the journal you are publishing to covers a large readership, the peer review process works quite well. Science is VERY competitive, and people outside of a certain colaboration (or 'clique') are sure to make sure that everything is right with a paper, or they won't deem it worthy of publication. You don't want people you are in 'competition' with to make false claims, thats cheating. If your opponent (author) in a sport cheats, you (the reviewer) will be sure to tell the ref (publisher) that they are wrong. The peer review process is set up to make sure that the reviewers are anonymous, and un-affiliated with the authors.


    --Xandu
  • Warning: The below is a rant. You have been warned.

    I am sick and tired of this fucking stupid analogy being parroted by half of Slashdot and everyone in the rest of the world. Just as Jack Valenti's keys-to-the-department-store analogy is bullshit, so is this one. And now I'm going to explain exactly why it's crap.

    First of all, it is an analogy, a word which is based on the word analog which in this instance comes not from the opposite of digital but from the greek related. Note analog refers to related, not exactly the same. An analogy is a simpler, easier to understand (at least for the intended recipient) example of what the topic is like. Not what it is exactly, but what it's similar to. An analogy cannot hold water in all cases or it would not be an analogy, but a directly correlated example.

    What do I mean by that? Well, take the following analogy: CDs are analogous to vinyl. They play music. They're round. The CD spins around and a reader of some type reads the music and produces sound. If you'd never seen a CD before, had no idea what one was, but knew vinyl, it'd be a good analogy. But if you then tried to build a CD player out of that analogy, you'd fail. Why? Because there are differences--there must be, for it to be an analogy and not simply an example.

    Just the same with the car-as-analog-to-website or other internet service. Finding the keys in someone's car and driving it away without permission is an analogy to hijacking a net service without permission. But it fails after only a simple scrutiny. Why? Because the car is no longer available to the original owner. But the service is! In addition, a car is private property that has been paid for so that the owner can use it himself, and is unable to be used by multiple people at one time. A web page, on the other hand, is designed for use by multiple people and, if on the internet, is assumed to be available for public consumption. In addition, my use of it does not exclude your use of it.

    All of the other stupid analogies follow the same reasoning. The portscan-is-like-rattling-the-windows analogy is also bullshit. First of all, the internet is not the real world. People keep saying if you don't want people using your machine don't put it on the net, and others respond that that's bullshit, but it's not. The net is a public network of machines, designed for interoperability. Private houses are not. If we could put a house in an alternate dimension and only the owners and their guests could get there, we would. But we can't, because that dimension doesn't exist. The net, on the other hand, is specifically there to share information. That was the original DARPAnet design and intent, and that's still the design and intent. If you want a machine that is there for you to do work and not to be part of the global network, you don't have to connect it. If you want to connect it, you run the risk of it being out there. It's like if you put something out there on the sidewalk, which is public space.

    But that's an analogy, and not perfectly apt, because a computer is private property. Which means breaking in and destroying or changing data is illegal. But tapping it to see what ports are open is NOT. Why? Because it's a private computer in public space, which means that they may have put it up there to make ports available. You can't know until you tap on the ports to see. Now if you get in and do something that is obviously not allowed (and frequently that's why they state it in ftp login pages), you are violating privacy. If you use a public service, you are not. There's no perfect analogy because it's a totally new concept of half-private half-public, without a perfect analog in the real world. In the real world it's either in private space or it's in public space. The net is public space made out of private data. Thus copying someone else's data is illegal but allowed by netiquette (with proper citation). Why? Because it's a new concept, a new medium, without the same workings of the old.

    There are many more bad analogies out there, and I don't have the energy to debunk them all. Just remember, if you've bothered to read this far, that an analogy cannot be used to explain all the rules and details of the analog. They just aren't the same. An analogy can only be used to deal with stuff that's similar, and when the two diverge, the analogy fails.

    Jeff

  • This isn't an open-source idea for research at all.

    Why? They _already_ know the answer. They're just not telling us so it'll be a "Fun Challenge." It's actually quite insulting to neural biologists to imply that their methodology won't yield results, which seems to be their point.

    I suppose that it _may_ yield some interesting results if someone discovered "Oh my! We've been looking at this all wrong!" and invented a new scientific method, but I'm not counting on it.

    Many great discoveries come from inspiration rather than brute force experimentation, and this "challenge" would only be reinforcing that.

    The paper is correct about it being novel though. Sort of the chicken and egg paradox - bet you can't figure out how this thing works without knowing how it works.
  • "Just because a car is unlocked and has the keys in it doesn't give you the right to drive off with it. "

    That doesn't seem like a good analogy...more like...if i left my car unlocked with the keys in it, and someone came and looked at it...and then it was still there when i came back. No property is removed from someone else...That webpage is still there for everyone else.
  • Many neuroscientists think that this challenge is a diversion and a waste. Implicitly, Hopfield thinks that they need a challenge that he is the one to give it. Anyone can see why that attitude is potentially offensive. The quotes in the article show this attitude. It's questionable whether solving the toy problem of one (very accomplished, but still human) scientist helps with "sharpening the skills required for brain research."
  • Seems I have read about this before. A mouse guiding humans, after a long search, to the question to the answer of life, the universe and everything.
  • I had to read this in grade 10. I actually wasn't dissapointed by it. Wasn't it also titled Charly?
  • The registration is for the NY Times article. The "directly to the site" link has absolute nothing to do with getting around registering at NY Times.
  • Can it tell between 'one' and 'won'?

  • First, this is a step in a long series of steps in the right direction for true verbal computing. Having a computer be able to recognize a word regardless of whom spoke it, how it was spoken, and the amount of background noise is much much better than what we have now, which requires usually lots of 'training' sessions with the software just to have it recognize your voice in constrained patterns. Of course, the next steps would be to have the 'mouse' respond to two or more words, but differently, ability to piece together multi-syllabic words, and then grammer parse sentences. Each taks is more difficult than the former, but this is an interesting start.

    Second, what I think is cool about this is sort of an open-source idea for research; there is a large community of nerds that while they don't have the stamina for research, are interested in new cool technologies such as this, and would be willing to help if it's easy to do. One thing I'd like to see more of are distributed projects that can use our idle time to put out more cool ideas; I'd love to see the research that created the 'lifeforms' made by a computer put out into a distributed form as to help their research. Here, further studies of their 'mouse' could be done by simply asking for voice samples of various words by people across the globe as to maximum the sample size (they only used 8 in the paper for testing purposes, but the web site seems to be up to about 600). Certainly, there's important issues such as disclosure of invention, but I think projects like these that challenge the community show excellent progress in research.

  • Does this piss anyone else off?

    What makes you think they have a right to my registration info? You know full well that information is not intended for my benefit.

    Typical Slashdot.. I shouldn't be surprised, but everytime I see this it just strikes a nerve. If they want me to not read that information, they should be be smart enough not to give links to it. But far be it from anyone at NYTimes to do that..

    .. go ahead, mod it down now, that's what happened last time. :\
    --
  • > Does this piss anyone else off?

    Absolutely not. If it's not for public consumption, then it shouldn't be on the publicly accessible web. If some people care about their privacy, don't want to give any information about themselves to the NYT, and can find a way around the login, good for them.
  • 64 posts in 60 minutes, and all but about 2 are /. in-jokes. They're all very, very clever - but did anyone read the paper and have anything interesting to say about it?

    I read it, but I'm not sure I get it. I -think- they're saying that they set out to implement a neural net for voice recognition, came out with something that worked much better than expected.

    I think they're saying that the -reason- that it works so much better than expected is a fairly novel reason (meaning not derivative of common neural net principles), and the process of understanding why this novel method works is best understood by treating the whole problem from a biology, rathern than compsci perspective. e.g. as you would go about trying to figure out how some organism that does something in a novel way does what it does.

    It seems to me that they go to great lengths in the paper to be "cute" about using biology terms to describe the behavior of their computer program, because they want to emphasize how organism-like their program is. Do all AI researchers talk this way?

    It would be really great if someone who actually understands the paper would post a translation, so everyone can understand what they're really talking about.
  • If they don't want people using it they should protect it the simple fact is they put it out there where it is pretty easy to find and use. Also as much as we like to think we are big and many in terms of the number of views they get in a day we just plain are not. Therefore it takes nothing away from them and it reduces pain for us. The real question of course is why do they leave it wide open? Simple answer their target market can and will not find it. They know they will have some people using it for whom it was not put there for but from the total lack of security on it you must come to the conclusion that they don't care. So yes this is typical /. for the most part smart people doing smart things in a way that most don't think of and that is why we are chaning the world.

  • A new Turing test? (Score:3)

    by K8Fan ( 37875 ) on Wednesday October 04, 2000 @09:11AM (#731704) Journal

    The most amusing possibility is that someone outside the research community may come up with the answer. As this doesn't involve building apparatus, getting a grant, publishing a paper or anything other than thinking, it's very possible an undergrad or a total amatuer will come up with the answer.

    Dr. Sejnowski sounded like sour grapes when he called this an "advertising gimmick". Yeah, that's what Fermat must have been doing. Too often scientists confuse the stuff associated with the practice of science - grants, publishing, peer review, experimental proof - with science. Science is what happens in your brain when you're not doing all that other stuff...usually while taking a shower.

  • Imagine a beowulf cluster of these mice. Unsanitary yes, but cuter than Alphas. Might use the IMPS (Infinite Mouse Protocol Suite) Oh, wait...
  • i don't think ultimately comes down to morals. that partner's info is supposed to be hidden, so i think it comes down to the same thing that's up with decss, people found a way around it. i don't think it's necessary to act as though the l.a. times' partner-only info is sacred. you don't tiptoe around the issue that the security is lax.

    or, maybe it does come down to morals, and that mine are simply of the mindset that if you're not able or even willing to protect it, you will lose it. the internet is young and needs to evolve, and to play 'good christian' with this stuff is foolish.
  • For some time now neuro-biologists (and worse, cognitive psychologists) have been misappropriating and misusing terms and experiments from cognitive computing to justify their often assinine guesses about mental processes. It lets them dress what is essentially bar-room speculation in the clothes of science. Mostly so they can get research grants.

    This experiment is calling them out. If they actually get it right then they have some justification in the processes they use. Of course if they fail...

    StrutterX
  • You haven't been around long have you?
  • It was originally published as a short story entitled "Flowers for Algernon", and was later expanded into a novel entitled "Charly". The short story was unutterably tremendous. Absolutely the most heart-rending (and terrifying) thing I've ever read.

    The book should not have been written. Alone, it was pretty good. But compared to the shorter version, it appeared spread too thin. Sort of like how you think you get really blasted from the schwag you normally smoke, until you finally encounter some truly good stuff.. :)

    Actually, in the seventies they made a movie out it that I saw for the first time a few weeks ago. Nice, but again nowhere near the original.

    I have read way too much sci-fi over the years, but this story remains my absolute uncontested favorite. Oh, by the way.. this post is off-topic.

    --

  • Re:Seriously... (Score:3)

    by MrGrendel ( 119863 ) on Wednesday October 04, 2000 @11:26AM (#731710)
    I think they're saying that the -reason- that it works so much better than expected is a fairly novel reason (meaning not derivative of common neural net principles), and the process of understanding why this novel method works is best understood by treating the whole problem from a biology, rathern than compsci perspective. e.g. as you would go about trying to figure out how some organism that does something in a novel way does what it does.

    It looks like there are a couple of things that differentiate Hopfield's approach from the traditional neural net approach. All NNs are biologically inspired to some degree, but so far the really common implementations (like backprop) have been simplified too much to give an accurate reflection of what really goes in a biological network.

    The two big differences between this and traditional networks that I can see (based on a quick reading) is that it is using spiking neurons and neurons are given specific computational roles. Spiking neurons add up inputs over time and send out a spike to other neurons after the inputs have reached some threshold value. Inputs also decay over time, so a few inputs occuring within a couple of miliseconds of each other count for a lot more than hundreds of input spikes spread out over a number of seconds. Traditional nets add up all of the inputs at once, decide whether or not to fire, and then reset (sometimes there is a training step in there also). Since time dependence is built into spiking networks as a feature, they are very good at detecting temporal patterns.

    The second difference I noticed, computational roles, means that neurons in different parts of the network may be specialized to do certain kinds of computation. One type of neuron could be used to detect patterns in a small frequency range, while other neurons detect patterns relating to which frequency ranges are currently active (I don't know if this is a realistic example, but you get the point). Traditional neural nets treat all neurons the same -- they act more like complex switches than computational units.

    This kind of setup is much closer to what goes on in biological networks. Neuroscientists used to believe that neurons are much more simplistic than they have turned out to be. Individual neurons do all sorts of computations that at one time were thought to be fairly complex. Edge detection and motion detection in the visual system are examples of this. It was once thought that these tasks required collections of neurons, but it has been discovered that individual neurons can detect motion in a particular direction and pairs of neurons can detect edges.

    I think there is also something interesting going on with the geometry of the network here, but I haven't quite absorbed that yet. Maybe somebody else has noticed this also and can comment (or correct me).

  • I liked that book : the end a bit sad, but I never forgot it.

    Hint to (too young) moderators : like in "Grape of wrath", this book deals with "slower" people and mices... and more stuff... So mod this mouse up, and let's keep on reading.

    "Flowers for Algernon" by Daniel Keyes.
    Published by Harcourt Brace (1966) and by Bantam Books (1967). Reissued in the Harcourt Brace Modern Classics series (1995).

    But enough of that mod rants : I read their article, and really like the way they wants people to attack the problem. Too often in the past, neural networks have been treated as "black
    boxes, mainly because of lack of mathematical fundations". I specially liked the way they stated it : "the novel essential principles of operation can be deduced based on the experimental results presented here alone. "
    I'm planning to read everything tonight (@work now :-() Any ./ reader to use that forum to share ideas instead of trying to break the moderation system into pieces ???
  • Although physical theories are often hard to test experimentally there is still the standard of internal consistency to a theory. Many fields of inquiry have no such checks. Consider a new 'toy' physical system that a physicist might come up with. They can ask concrete questions like 'what are the energy levels?'. These questions often have definite answers and it's easy to judge if someone's work on this problem is competent. It's much harder to do that in other fields si I find it very interesting to see a problem posed like this. Physicists are very used to working with 'toy' systems. I've a feeling you may be wrong in assuming other sciences are like physics. Try reading some psychology papers...
    --
  • Scientists are hardly limited to single unit recording: in addition to single unit recording they can perform psychophysical experiments, do parallel recording, introspect, examine neuroanatomy, do genetics, perform functional imaging, perform in-situ staining, introduce various drugs, to name just a few. That doesn't make the problem of figuring out how brains work easy, but it certainly makes it a lot more tractable than single unit recording.
  • To use just TWO images for training your net is completely hopeless, and anyone who has any knowledge whatsovever in neural nets should realize this.

    You might train your net with a couple of hundred pictures - each classifying a tank or not. The use some additional hundred pictures (none of the same) to use for validation.

    Your point though, is valid. Neural nets have a tendency to overspecialize, adjusting the fitness landscape to close to the input-cases. So you have a separate validation set, to stop the training when the validation set matching drops...
  • This looks like a pretty cool project, but anyone else concerned that we'll now have folks training it to recognize all sorts of 4 letter words?

    I wonder if opening it up to the general public to do whatever they want was a wise idea.

    Anyone remember what happened to that "Post what you want" website that was posted to /.?

    - "I am the cheese!"
  • is this supposed to be like a good will hunting thing? the prof leaves the question on the board outside of class and we all fumble around with the answer?

    1. I LOVE YOU [mikegallay.com]

  • We've had that for ages (Score:5)

    by Anne Marie ( 239347 ) on Wednesday October 04, 2000 @08:10AM (#731718)
    A neurologically simple brain for determining whether the number "1" has been achieved? Sounds like a first-poster if I ever heard of one.

    My hypothesis: the mouse checks the cid# like the rest of us.

  • Perhaps they should name the mouse Algernon.




  • Wasn't intended to be karma whoring, and if you check my past postings, you'd see this.

    I wasn't going for moderation points, just doing the favor of posting the link.

    If it'll make you feel better, I'll even ask here that it NOT be moderated past 2.

    -LjM
  • Who will be the first to upload a sound file of him/herself saying "FIRST POST"?
  • 1) A door that opens for computer consultants (one with both hands full carrying huge sacks of cash) when he says "open sesame". (laugh, it's funny) 2) A dictation machine for people... oops, that's been done already... 3) A Natural language machine that gives detailed instructions to drivers trying to figure out how to go from "here" to "this location". (this has been done, to some extent.. being able to recognize anybody's voice would make this even easier on said driver)
  • You siad it your self. ..."requires usually lots of 'training' sessions with the software just to have it recognize your voice...". Now i'm sure you have met people with akk kinds of acceents, some are easily understood others should not even bother making the (noble) attempt at english (or any language). Is it really possible to pull this off. I love to read about breakthroughs like this. But for once I want to go to a trade show or something like that & see it.

  • As easy as some people think SDMI will be cracked... maybe this artificial net can detect the watermark in SDMI and remove it.

  • [re NYT articles] What makes you think you have a right to view that content for free?

    Because I can? Without breaking any laws, I might add. That's sufficient justification for me.

    You know full well that link is not intended for you,

    Why in the world should I care about the intentions of publishers? The book "Perl Cookbook", for example, is intended for Perl programmers. Does it mean that if I am not a Perl programmer I should not buy this book?

    If New York Times wants this information not to be world-readable, there are plenty of simple ways to achieve this.

    everytime I see this it just strikes a nerve.

    It does? You must be really in pain reading the CueCat stories, then. Isn't it also using something in a way that was not intended?

    Kaa
  • Yeah, I need to go back and continue reading the paper, myself. I only had a chance to skim it earlier.

  • Hopfield's last theorum... (Score:4)

    by ucblockhead ( 63650 ) on Wednesday October 04, 2000 @08:33AM (#731727) Homepage Journal
    I can see it now...Hopfield and his grad student are going to die in a horrible car accident and scientists are going to spend the next three hundred years trying to figure out what he meant. An obscure professor will finally produce the answer in 2391 in a 1300 page paper that uses quantum theory, the psychology of preadolescent children and a statistical analysis performed by a 300 Exohertz computer, but only five people will actually be able to understand it.

  • Re:Cool, but.. (Score:3)

    by Masem ( 1171 ) on Wednesday October 04, 2000 @08:34AM (#731728)
    The mouse was trained on only one vocal pattern and one instance of it saying 'one'. Everything else on the site is just simply ways to test what the trained mouse does when you pass it a sound sample. The fact that the training only had one input set is part of the challenge they are asking others to look at - exactly what is the network topology of the neural net, and what sort of 'objective function' did they use to train it... hopefully, they did not hard code 'one' into the neural net, so that this mouse could have been trained on any single syllable word and still gotten the same results.

  • And now I shall be able to build a voice fingerprint for each and every hacker in the /. community and Dextor's lab will be no more! Ha ha ha! ha ha ha! ha ha ha!

    --
  • Except most first-posters actually fail in the task, posting 17th, etc.
  • What makes you think you have a right to view that content for free?

    Uh, gee, the fact that they have a link that doesn't require a password?

    You know full well that link is not intended for you, or most likely anyone else reading this board.

    Why don't you tell us who it is for? If this is only for certain people, why is it completely accessible to everyone on the Internet, with no statement about who can or can't use it?

    I suppose now you're going to rant about how I'm stealing from all these web sites because I run a proxy server [guidescope.com] and block their ads and cookies? And sometimes I tape shows and skip over the commercials!

    OK, lock me up. I've seen the error of my ways.

  • I wonder if someone can train a 660 newral net rat to detect all possible "First Post" variation is Slashdot posting subjects and automatically moderate them down. We can then train it to detect trolls in noisy conditions. Making Slashdot better, one step at a time.
  • So, how long will it be before someone starts teaching this thing four letter words?
  • OH! The controversy! The age old question.. is it right for me to bypass the link? For the sake of not beating the topic dead, and it's almost one boot the teeth away from that point.. but.. It seems this registration info is for the purpose of entering it into a database. Which could help them determine who's reading and who the target audience should be. *DING* Happy sound... Maybe more articles targeted at our diverse slashdot demographic? Right. I'm not registering. Reason? Time issues, Spam issues. Plus it may be there but I didn't notice that they DIDN'T say they wouldn't give out the info to others, and all I need (as a female tech) is another "HOT XXX CHICKS" e-mail to pop up in my inbox while I'm at work. (Oh yeah sure you THINK the NY times wouldn't sell to the porn guys... ) But seriously. Looking at the page I noticed no "no spam" disclaimer, and they are asking if you want to join up with at least ONE known spamzilla. Plus, on final note. No money lost - no harm.

  • Re:Standard bypassing registration link... (Score:1)

    by Anonymous Coward
    Hey, question.
    If I go to Radio Shack, they always ask for my name, address etc. Now it's no secret that they do that so they can sell me to mailing lists. I would imagine that _Radio_Shack_ sees this as part of the transaction. Perhaps it even lowers the prices of the merchandise.

    Now, by your system of morals, is it therefore unfair for me to make up a fake address? On Radio Shack's side of the transaction, the collection of the marketing info is an imperative. On my side of the transaction, it isn't; I pay money I get stuff, and I'm under no obligation (legal or otherwise) to provide them with that information.

    To say that this is a matter of morals is going a bit far, since obviously there is a matter of disagreement here among morally capable individuals such as yourself and I. At the present time, the only thing we can be sure of are the society's written rules, which do not forbid using the partners link. As far as morals go, this would be much clearer if the NYT specified clearly what their access policy was.

    In fact, actually just now I checked, and I guess this clears it up. From http://www.nytimes.com/subscribe/help/copyright.ht ml :

    "However, you may download material from The New York Times on the Web (one machine readable copy and one print copy per page) for your personal, noncommercial use only."

    Incidentally this page does not require registration to read, so that offer is given to the general public. So the NYT gives a free offer indeed. If they would post "You may download pages only if you are a registered user" I would be more likely to agree with you.

  • They already have those. In fact, they've had those for over fifteen years. All you need is a MIDI-enabled pickup that's fast enough to process the input. I saw my first one in 1985. I believe you could stick a microphone into a sax and get MIDI data off it. Polytonal MIDI recognition is only a matter of processing speed.

    That's all much simpler than recognizing the word "one" spoken by different voices.
  • Are you saying John Hopfield should be the authorized examiner of neuroscientists? The challenge is taken as a stunt and an insult by some. Who are these two to tell everyone else how to work? I can see their point.

    Secondly, will solving Hopfield's network give us any insight into the brain? He is a leader, but this problem may not be so relevant. Perhaps it won't help. Time spent working on this problem is time not spent working on the brain directly.

  • But it's not perfect (Score:3)

    by Morgaine ( 4316 ) on Wednesday October 04, 2000 @01:13PM (#731738)
    The peer review process is set up to make sure that the reviewers are anonymous, and un-affiliated with the authors.

    Unfortunately, there's a little more to it than that. If you return to the publisher a negative review of a paper written by a respected figure in your scientific community, there is an element of "black mark" against your name in some quarters as a result of the conflict of interest that the publisher has through needing the famous name to appear in his or her journal rather than in a competing one. As a reviewer you're anonymous to the author, but not to the publisher!

    And I'm not even going to mention what happens when the journal's editorial board includes researchers interested in the same paradigm or method employed by the famous person, so that publication of that paper validates their own research area ...

    Peer review is a fairly good process on the whole, but I doubt that anyone who's been involved in it [I have] would suggest it approaches perfection. :-) A dollup of cynicism is always helpful, here as in so many other areas where humans err. Yes, even in hard science.
  • What is terrific about this research is not simply that it can recognize a word independently of the speaker, rate of utterance, and noise characteristics. The 'design patterns' used in this system can provide a building block for systems which recognize patterns in many different areas. Generalizing a system like this from recognizing one word to recognizing many seems to be simply a matter of determining which 'word-neuron' is the most excited. Recognizing phrases may be accomplished by using a sequence of 'word-neuron' firings as input and determining which 'phrase-neuron' is the most excited. Presumably, the system works by using some sort of time-delay neuron chain which ensures that all the monosyllables of the word 'one', when spoken in the correct sequence, generate signals which arrive at the 'aggregator' neuron at the same time, thus pushing it over the threshold. (I won't attempt to try to figure out how the -learning- system works right now, which is the really deep part of this :). I would really like to know which principles of neuron activity play a factor in building such a system- there is a relatively fixed physical model which is simulated in a piece of software; as well as the the interconnection patterns which realize the learning and pattern recognition model. This knowlege could form a great foundation for a general purpose piece of software that could be applied to many areas of pattern analysis. Mike
  • Anyone can see why that attitude is potentially offensive.

    And suggesting that the earth went around the sun was extremely offensive to the bulk of the scientific community of the day.
  • You could probably train it to recognize different chords and notes, and then you never again would need to buy sheet music or tab out songs by yourself.
  • Too long...

    -----------

    "You can't shake the Devil's hand and say you're only kidding."

  • Hopfield and Brody are challenging the neuroscience field (and others) to solve their puzzle and demonstrate their reasoning ability. The USC team is working on something that's just what it is, a machine to recognize speech.
  • There are astronomical data to support the modern understanding of the solar system.

    Imagine that Copernicus had not learned anything about the solar system. Instead, he made up a system that they claimed worked in a way similar to the solar system. They told everybody to stop looking at the sky and to look at their system because everybody is thinking incorrectly. Telling everybody that he/she is thinking badly and showing him/her a toy is not the same thing as having an answer.
  • I saw this mouse ages ago.

    It had a human ear grafted onto its back.

    So what's the news? It was an Iranian ear, but the mouse understands English?

  • You're probably right. I was going from what I heard in an AI course a year or so back, and it could be that there were many pictures taken, but the pictures were taken in batches, with the tank pictures taken in different light/exposure conditions than the pictures without tanks.
  • Well, let's wait until the mouse can respond to words like "Yeepah, yeepah!", "Andale!", and "Arriba!". Good thing it wasn't a dog, otherwise it would respond to "Taco Bell".

  • Hey,

    Maybe they could train it to recognise the phrase 'No, hold the anchovies'. If they could, it would be substantially more inteligent than the goods answering the phone at my local pizza delivery place.

    Michael

    ...another comment from Michael Tandy.

  • Are (Cue)CAT-Scans allowed on this mouse?

    Strong data typing is for those with weak minds.

  • >This looks like a pretty cool project, but anyone else concerned that we'll now have folks training it to recognize all sorts of 4 letter words?

    That would be silly. I don't think the mouse is actually trained by the soundfiles people send in. Otherwise it would be completely spoiled in a short time. The mouse can recognize one now, and it is this version of the mouse that people can try to figure out, not some random changed version.

    I think this is a very cool project. We know that the solution is pretty simple. Biologists can apply their normal methods to it. If they can't figure it out, maybe that would say something about the methods.

    I haven't actually had time to read the paper yet, so maybe I'm being more nonsensical than I think.
  • Our second and more important motivation for presenting the material in two parts is
    to open a discussion on the role of deductive thinking in neurobiology. As we have described in
    the introduction, we firmly believe that careful and rigorous deductive analysis based on
    incomplete knowledge may still lead to novel conclusions and clearly indicate what the most
    incisive next experiments are. Nevertheless, incomplete data all too often discourages deep
    deductive thinking in neurobiology.

    The whole point of doing science is to think. If, as they claim, there's too much data-collection and not enough synthesis, then this is a fun way to get people going.

    Good science starts in the lab, but it reaches it's zenith in the shower (or bathtub, if you're Archimedes). Time to pull out Popper and Kuhn and think about how and why science is done.
  • I like "Brain" or Topo Gigo (sp?) myself.

  • Neural Nets and Voice Recognition (Score:3)

    by _Splat ( 22170 ) on Wednesday October 04, 2000 @08:50AM (#731753)
    We've seen things like this before. One of the major problem with neural nets is their tendancy to specialize.. Building a system to recognize one word doesn't remotely compare to a system that can tell one from fun and done while also having the capacity to tell Bob from Rob from Cobb. The experiments posted on the site only show that the system can differentiate between 1 and the other numbers 2-9 and from various nonverbal tones. A neural net will very likely lock on to the specific differences between the sounds of these numbers. Example: A while back someone was creating a neural net to identify tanks on the ground in satellite photos. Two samples were used and the net learned to successfully differentiate them. When other samples were tried, however, the system was completely wrong. Eventually it was determined that the photo with tanks was brighter than the photo without, and that was what the system used to differentiate the photos.
  • That was just about the most worthless, uninformed post I've ever read here on Slashdot.

    -----------

    "You can't shake the Devil's hand and say you're only kidding."

  • C'mon, get over it!
    It's almost like these days people think that anything that isn't a troll is "karma whoring."
    The damn buzzword has been way overused. You're beating a dead horse. Who CARES if the post is just there to soak up karma of not? Are you jealous because your karma is in the negatives or something? If whoever's moderating appreciates the post, they mod it up. It was obviously a good post regardless of the reason it was posted, so quit it with the fruitless acusations!

  • Philosophically important (Score:5)

    by SIGFPE ( 97527 ) on Wednesday October 04, 2000 @08:56AM (#731756) Homepage
    I think this experiment could be very important for neurobiological research and maybe other typs of research. There are many fields of science where it is possible to go on forever publishing research without any checks. Obvious areas where this goes on are fields like so-called postmodern literary criticism. But it happens in the sciences too. In behavioural evolutionary biology you can make up just-so stories in paper after paper safe in the knowledge that nobody else can rerun evolution for you and demonstrate that you are wrong. In psychology you can repeatedly perform experiments measuring correlation between this variable and that. By chance one in 20 results are 95% significant and you publish those results as if they are something other than noise. Hopfield's experiment is going to be a sanity check against this kind of work - a kind of experimental control. Here's a situation where somebody does know the answer and work can be checked. A neural net involving only a few hundred neurons. If researchers are unable to reverse engineer this then should they really have jobs supposedly reverse engineering animal or even human brains? We need to see a few more tests like this in academia. Beyond a certain point - after you've taken your last exam - academics are no longer accountable to anyone. Sure - you get peer reviewed. But what happens when you and your peers all belong to a clique that have a vested interest in promulgating a particular scientific dogma? This experiment is a wonderful way to ensure that researchers still are being tested.
    --
  • While the purpose of this project is to make the mouse recognize things, I think the greater significance is shown by the doctor's following comment: "I had to use a way of thinking that felt very different from what I normally use," Dr. Brody said. I think that this is probably the most interesting thing in the whole NYT article. By making this into a contest he is not only bringing more interest into the field but also challenging people to really 'think'! This is how progress really happens. It will definately be interesting to see which different disciplines his methodology will affect, especially if his creation is as novel as it seems.
  • Hopfield's experiment is going to be a sanity check against this kind of work - a kind of experimental control. Here's a situation where somebody does know the answer and work can be checked. A neural net involving only a few hundred neurons. If researchers are unable to reverse engineer this then should they really have jobs supposedly reverse engineering animal or even human brains?

    Hopfield is a well respected theoretician. He is not in the business of reverse engineering animal or human brains. The fact is most of the physiologists that do reverse engineer sensory systems (in a manner of speaking) would not prioritize Hopfield's game very high.

    The likely result is that lots of theoreticians will engage in his game, and maybe there will be some fun in it for them. Most of the physiologists will be busy engaging in the game for a living. It may very well result in a science of reverse engineering neural systems evolving to describe what successful physiolgists do - in the same manner that Kuhn described how science progesses in his career.

    But don't kid yourself that this is a sanity check on people who do this for a living. That is not Hopfield's contribution to neuroscience.
  • Consider a new 'toy' physical system that a physicist might come up with. They can ask concrete questions like 'what are the energy levels?'.

    Sure, if there is a 'toy' physical system. But physicists only use those when they can (as they are easy to check, like you point out). What about "Is the universe flat, open, or closed?" or "Gravity at the quantum (really small) level work?". You can't 'check' these either. The educational problems have 'toy' systems, but on the forefront (where the research is), there are not.


    --Xandu
  • Sounds more probable :)
  • There are astronomical data to support the modern understanding of the solar system.

    That data wasn't available then.

    Telling everybody that he/she is thinking badly and showing him/her a toy is not the same thing as having an answer.

    But all the models of science are toys, without exception. That is the whole idea behind science, to produce mathematical toy models that hopefully might approximate the behaviour of reality as determined through empirical tests.

    Scientists have never had The Answer, and no competent scientist would ever profess to do so -- the relationship between reality and the models of science are in Science 101, after all. But they're getting damn good at creating models that accurately mimic a lot of reality's behaviours, despite the scientific method not having any ability to determine The Answer or The Truth or whatever.

    And that's why it is always good to point out the error of their ways to those "scientists" that waffle on interminably without producing hard testable models based on the hard thinking of logic and mathemetics. It takes more than just mimicking the forms to produce real scientific results. Reality checks may be painful, but they're important.
  • i don't think ultimately comes down to morals. that partner's info is supposed to be hidden

    I'm sorry, I must have missed the cloaking device. What makes you think this is "supposed to be hidden"? Or did you mean to say "they were probably hoping no one would find out about it"?

    With all due respect, the keys-in-the-car analogy is particularly useless here. Cars have keys because they are only supposed to be driven by the people who paid for them, or those given access by the owner. The fact that the keys were there does not negate the fact that you need keys to start a car.

    This is more like someone leaving printed material (better yet, a newspaper!) in a public place (park bench, on a seat on a bus/train), and then expecting no one to ever look at it (not TAKE or MOVE it, but even READ it) because they paid their 50 cents for it, not them.

    Or like a company mailing you hardware unsolicited, therefore with no restrictions on it, and then saying you shouldn't use it except in the way they want you to because it's not really yours. Oh wait, that really happened.

  • If the data were not available, how did Copernicus get his idea? Science relies on data. Copernicus had his ideas after looking at the available data. Heliocentrism was more than a lucky guess. I hope you haven't been reading Koyre.

    I am complaining that Hopfield's challenge is not a reality check. It's a fantasy check. Their toy is not a model of anything. It's a toy. He wants people to come look at his toy. Why? Because he thinks that people are not thinking correctly. How is his toy going to help? It beats me.

    I think you're misunderstanding Hopfield's challenge, and both of us are misunderstanding one another. Sciences use models to describe and predict events. We agree. Neurobiologists endeavor to describe and understand the brain through a wide variety of approaches. If Hopfield were offering a toy that models the brain, they would be excited. He's not offering one. He's offering a toy problem that isn't the brain as a test to the community. It doesn't approximate anything in the real world. It is, simply, what it is.

  • Does this piss anyone else off?

    What makes you think you have a right to view that content for free? You know full well that link is not intended for you, or most likely anyone else reading this board.

    Typical Slashdot.. I shouldn't be surprised, but everytime I see this it just strikes a nerve. If you want that information, be courteous enough to give them the registration info for it. But far be it from anyone on Slashdot to do that..

    .. go ahead, mod it down now, that's what happened last time. :\

  • Re:asdf (Score:1)

    by Eso ( 205333 )
    FRTFP!!!!! FRTFP!!!!! First reply to first post! I like cheese! I like goats! I am a l33t h4x0r! w00t! m0d m3 uP b33y4tch3s!!!!!!!!!
  • Since most offtopic/flamebait/troll posts are short (usually like "phirst pozt"), let's train this thing to filter them out!
  • I havent been following neural nets since the very early days, but i was under the impression that this [the abitity to distinguish one thing {even a complexd thing}] was state of the art.

    Is the point more in the competition, were by they can discover the relationship between the 'best guess' on the question model, and the actual workings and consider the implications of this to our [their?] assumptions on how to build a 'net, and indeed how HUman minds work.
    ...or is it just a publicity stunt?

    Also does anybody know of a good [non-too-tech] site on the present state of play in 'nets?

Slashdot Top Deals

I have hardly ever known a mathematician who was capable of reasoning. -- Plato

Close