Recent Posts

Pages: [1] 2 3 ... 10
Tech Talk / Re: Self-Driving Cars
« Last post by 2397 on Today at 05:04:03 AM »
what is more you can design an industry wide neural network style AI so that every time one car drives over ice, that experience is translated to every other car.

Now I'm worried about 4chan doing to cars what they do to chatbots.
TV & Movies / Re: Best running joke in a TV series?
« Last post by Muriel on Today at 04:36:05 AM »
Ugly Naked Guy on "Friends" had some good moments :)
Podcast Episodes / Re: Episode #602
« Last post by Gravity Allen on Today at 02:34:08 AM »
In last week's Who's That Noisy, Jay et al. spent some time wondering out loud about whistled languages, and hoped someone might write in to answer some of their questions. And someone did! Alas, my e-mail didn't make it onto the show. So my response won't have been entirely in vain, I thought I'd post the body of it here, for anyone who might have been interested in learning more. ;)

First, it looks like they aren't exactly languages, per se, so much as they are registers -- versions of a language spoken in specific contexts, like the formal and informal varieties of English that speakers use in different social settings, maybe involving different vocabulary choices. What the Silbo Gomero -- "the Gomeran whistle" -- does is encode and transmit (via whistle) certain important aspects of the spoken language, meaning it doesn't really have an existence independent of ordinary speech.

To understand the advantage of whistling over speaking, you need to know a little something about what makes speech comprehensible in the first place. When you bring together your vocal folds, which are the membranes stretched across the opening inside your larynx (or voice box), and force the air out of your lungs, you generate an oscillation in them via the Bernoulli effect. As they come together and move apart, you can either tighten or loosen them, modulating the frequency of your voice. This frequency corresponds to the pitch of your voice, and plays many roles, including relative emphasis (a stressed syllable has a higher pitch than an unstressed one), communicating age or gender (on average, a young woman will speak with a higher pitch than an older man), and asking questions (think about ending a sentence with a rising intonation).

But what actually makes speech intelligible, and lets us convey complex ideas, is the way we filter the sound generated by our voice box. When our vocal folds vibrate, they do so not only at some fundamental frequency, but at whole-number multiples of that frequency. So, if I produced a sound at 100 Hz, I would also be producing sound at 200 Hz, 300 Hz, et cetera. These are the harmonics of human speech. And depending on how you configure your tongue (and your teeth, and your lips), you either amplify or dampen certain harmonics, in ways that are characteristic to each configuration. In other words, the sound that finally comes out of your mouth is marked with he signature of the shape your mouth was in when it was produced. To grossly oversimplify things, an /i/ sound ("eeeee") amplifies higher harmonics, while an /u/ sound ("ooooo") amplifies lower ones. We tell apart vowels by how our mouths effect the harmonics (consonants are a little more complicated, relating to harmonic transitions).

And like you guys said, lower frequencies travel farther than higher ones, since higher frequencies lose energy more quickly. Moreover, harmonics are always quieter than the fundamental frequency. So, the qualities that make speech speech are represented in high frequencies at a low volume. This kind of information can't carry very far at all, because it starts off quiet, and drops off fast as you move away from the source. I don't know how reliable Guinness World Records is generally considered, but they claim that the normal range of a man's voice outdoors is about 600 feet. In contrast, whistling can be heard for miles; I've heard numbers like 1-3 miles at the low end, and 7-10 miles at the high end. You could scream, but you'll strain your voice, and you're still starting off with a handicap: all the important stuff is still at a quieter, higher frequency, and so it just won't travel that far.

Ultimately, the reason whistled languages carry farther than speech is because, in encoding the sounds of a language in this way, you pour all your vocal energy into transmitting the information that actually matters. Recalling that it's what we do with the harmonics of speech that really makes the difference, more than the original fundamental frequency, the absolute frequency of the whistle corresponds directly to the harmonic frequencies that would normally allow you to tell apart an /i/ from an /u/. You can see how this works for yourself by placing your tongue where you'd normally put it to make an /i/ sound, and then whistling; you'll get a high pitched note. If you place your tongue where you would to make an /u/ sound, and then whistle, you'll get a low pitched note. So, speakers train themselves to speak and understand whistled speech by interpreting the changing frequencies of the whistle as the different speech sounds (both vowels and consonants). And with almost all your acoustic energy being dumped into the whistle, the key info ends up making it that much farther.

I'd be happy to expand on these points in more detail, or share more of what I learned researching whistled languages, if there's interest. :)
Tech Talk / Re: Self-Driving Cars
« Last post by Enkidu Shamesh on Today at 12:46:34 AM »
Average historical temperature for today at my location is below freezing.  We have a high of 60 degrees in the forecast.  This whole planet is going to burn, if the fires aren't put out by the torrents of blood spilled in the inevitable conflicts that will rise.

you bastard. return your warmth immediately to New Zealand where it belongs today.

include a self-addressed stamped envelope and I'll happily send your cold rain to you in return
You're new here, so maybe you're unfamiliar with the tendency of creationists and other cranks to post hour-long videos that they are unable to paraphrase in their own words. If you can't be bothered to give any indication whatsoever of your understanding of what the lectures say about event horizons, then I can'tbe bothered to watch them.

Spacetime is locally perfectly smooth across the event horizon of a Schwarzschild black hole, according to General Relativity.

(And I'm not so sure it was me whom brilligtove was accusing of gish galloping.)
Where's my summer?

(click to show/hide)

Oh sorry we're borrowing it.  60 degrees today (F) in Chicago.  Average historical temperature for this day is below freezing.

If the  road gets cleared I might go skiing tomorrow.  Of course we were sweltering in 30ยบ+ a couple of days ago.
Podcast Episodes / Re: Episode #602
« Last post by lonely moa on Today at 12:23:56 AM »
I sort of couldn't believe that the Rogues didn't know that Svante Arrhenius published his thesis in 1896.  I thought every thinking person knows this.  Guess not.
I find that most discussions of P-zombies are plagued by conflicting definitions.  Can we agree before continuing what kind of P-zombie we are talking about?
  • One which is just behaviorally indistinguishable but which derives its behaviors without imitating the function of the human nervous system.
  • One which is behaviorally indistinguishable and functionally equivalent but not physically identical; that is, it derives its behaviors by imitating the function of the human nervous system without using the same biological means as a human brain.
  • One which is behaviorally indistinguishable, functionally equivalent, and physically identical to a human being; that is, one which derives its behaviors from the function of a nervous system physically identical to that of a human being, but which does so without consciousness.

It seems to me that the first two examples are merely simulations of humans exhibiting human-like behavior. I would have no reason to assume that they are actually conscious in the same way as real humans.

But I have a question about type 3. If said organism is really behaviorally indistinguishable, functionally equivalent, and physically identical to a human being (including a physically identical nervous system), then how can we possibly know that it does not possess actual consciousness?

I'd expect that the tricky part about dealing with all 3 types would be knowing how to distinguish a real person from the simulated one.

If the very definition of a "p-zombie" specifies outright that they don't possess real consciousness, then what's the use of this thought experiment in questions about the nature of consciousness?  Are they usually invoked for other purposes, like questions of ethics perhaps?
General Discussion / Re: The Twitter Thread--A Thread of Great Consequence
« Last post by Hanes on January 21, 2017, 11:40:58 PM »
Where's my summer?

(click to show/hide)

Oh sorry we're borrowing it.  60 degrees today (F) in Chicago.  Average historical temperature for this day is below freezing.
Pages: [1] 2 3 ... 10