Monday, September 22, 2014

iPhone 6, Siri & Her Cousins

Saturday, September 20, 2014 7:21:47 AM

iPhone 6, Siri & Her Cousins


i have noticed that in all The New Blurbs for The iPhone 6, there is never any mention of Siri ( ? )

You may recall that after Siri was introduced, there was a flood of Television advertisements for this feature, that included many Celebrities hawking this amazing new technology.

And then they Disappeared.


Because these Ads were so Amazingly Disingenuous that Apple was effectively committing Fraud.

First of all; The Actual use of Siri reveals that Siri is Astounding Shy & Uncertain about her own ability to Understand what you’re saying. After you utter anything, Siri would bring up a text rendering of what it thought you said, upon which you’d confirm it, correct it or hit a button confirming it. This entire process would take about as long as typing in a request, especially if you’re fluent at typing with both thumbs.

Siri is Not Apple’s first attempt to allow you to speak to their computers, nor is Apple The only Company offering this service. It has been offered time & again since The 1980s or before.

Even before there were proper computers, there were attempts to create mechanical or electronic devices that would respond to spoken commands.

Thomas Edison, as a child, was supposed to have created a mechanical device that Spoke Intelligible Words & Short Phrases. ( ! ) ?


But they’ve never Worked.

Not Really.


Not in Actual Usage Situations. Not so that anyone that actually wanted to use them, could depend on them.

Quite a few Companies use something like this as a Cybernetic Telephone Receptionist, that allows you to answer simple questions with precise key phrases.

Have you experienced one of these Tele-Operators ?

Do they Work ?


No, they do Not.

Their functionality does in a back-handed way work exactly The way The Company or Service want it to work however, And that is to Frustrate you sufficiently, So that you’ll hang up & quit bothering them. This is especially True for Social Services by Any given Government.

The curious aspect of that; Is that The Machinery & Protocols for these services seems to be keyed to scare away those people that are in most need of these services, while swindlers & fraudsters have The Sanity & fortitude to endure all these layers upon layers of trials & redundant scrutiny to finally get their ill-gotten rewards.




This all seems so very odd to me, because i can easy imagine a few simple conventions that if applied; would allow these voice recognition systems to work.


a) Put all of this Voice Recognition Hard & Soft-Ware onto a Dedicated Chip or Black-Box whose ONLY Function is to Analyze Sounds to Pick out Spoken Commands or Other Sounds that The Host Device will be programmed to act upon, such as distinguishing between a balloon popping & a gun shot. It is completely crazy to run all of this processing through The Main CPU that is required to process all of The Device’s Other Functions. ( It is my prediction that at some time in The ( near ) future, Computer Engineers will eventually realize that instead of creating bigger & faster CPUs, They will distribute all The Various Functions to Entirely Autonomous Processing ‘Agents’ that will cooperate to serve The User. ( ? ) )


b) These Speech Recognition Devices need to Introduce themselves when they are first opened, To agree upon a name for The Device, & then ask The User to read a series of sentences which have been tailored to include all The Given ( English ) Language’s Phonemes in various combinations, So that The Device will know how you, As a unique & Autonomous Individual, pronounce various words & phrases.


c) The User will Then be able to Train ( Directly or Tacitly ) The Device to Anticipate Questions or The Class of Answers that The User is Desirous of knowing. This simply means that The Device knows what The User is Interested in, & discard any possible solutions or Answers to Inquiries that are outside of these parameters, If an inclusive solution is available.


d) If The Device is Confused by a Question’s structure, pronunciation of words or The Ambiguity of a Word or Phrase, Rather than ‘Guess’ what is meant, It will ask a pointed & direct question to resolve The Uncertainty. ( it drives me crazy that commercial search engines return millions of ‘hits’ that are completely irrelevant to what i’m looking for, when The Search Engine should have taken a few additional moments to hone in on what i’d intended to ask in The first place. ) !

: - - - - - - - - - - - - - - - : o


e) i don’t know how computers are currently being trained to understand speech, but considering that there isn’t a reliable electronic Sentience yet; it may be that this suggested approach will be what they’re looking for ( ! )


i have this idea that The way to resolve ambiguity in ordinary sentence construction or poems that intensionally integrate double entendres into their construction, may be easily resolved with -Lists-.

This assumes that ( which may be entirely unfounded ! ) these Speech Recognition ( or typed sentence questions & requests ) Devices, can understand ( parse ) Ordinary Sentence Constructions.

While i spent The Entire 8th Grade in 1968 diagraming Sentences, Training a Computer to do this, seems to be one of those incomprehensibly ‘Hard Problems’ ( ? )


But if this step has been achieved


Your computer has some special IC’s ( Integrated Circuits ( Little CPUs ( Central Processing Units ) ) that take all The words in The Sentence, Phrase, Paragraph, Book or Poem, & Make a List for each one that looks up its definition & all thesaurus references, & then compares those lists to The other Word’s Lists & Looks for ‘Patterns’ or Repetitions in ‘Themes’ So that if The Theme of both ‘Golf’ & ‘Sex’ come up again & again; Then The Computer will assume The Speaker is attempting to be Clever & is using a Double Entendre. Otherwise; A Single Consistent Theme will remove all of The Ambiguity from any Word or Phrase Usages.

Another Innovation that i’d like to see is that The Speech Recognition ‘Black-Box’ would include it’s own Encyclopedic Dictionary in A Computer Language that all ( English ) Sentences would be translated into before being acted on or Translated into a Second Language.

This ‘Computer Language’ would contain all ‘Ideas’ that every Humyn Language incorporates into their regional Languages, So that The Computer Language has a full Vocabulary of all Humyn Concepts. Each of these Ideas would be correlated with ONE Word, & Each Word would have ONE Concept associated with it, Thus Completely Eliminating any Ambiguity within this Computer Language.

After The Humyn’s Sentence, Phrase or Request is Translated into The Computer Language, The Device Resolves this Query & Translates its response into The Least Ambiguous ( English ) Sentence or Phrase.

( Unless The Device has been programmed to be Snarky ! )

: - - - - - - - - - - - - - - - : o


So why don’t these companies produce a functional Speech Recognition Device or Service that Actually Works ( ? )


One of The Services that i envision for such a device or service is a ‘Talk-Line’ that anyone can call into, & it would cost something like 2 a minute to use. The Service would Listen Attentively to everything you said, Develop a Customer Recognition File for Each Caller, So that when you called again, The Sentient Agent would take on The Roll of a Confidant & Close Friend.

It would then be able to Respond Insightfully to your concerns, questions & comments & provide The Caller with any Information that’s available on The Interweb.


Think about that for a moment.


You could have any question that you have about anything, quickly & easily Answered.

If you were confused by some point or fine detail; You could stay on The line for hours & hours until you finally had a full & comprehensive understanding of The most subtle ideas that were previously accessible only to Doctorates that spent decades in Universities struggling to master.


Do Big MultiInternational Corporations or Governments want you have access to that kind of information ( ? )


The Freedom of Information ( Movement ) is based on The Implicit assumption that most information is effectively inaccessible, even if its available in some esoteric theoretical sense.


If that ‘Theoretical Sense’ were removed from any inquiry

Many things would have to be made explicitly secret.

Lots of Things.


Lots & Lots of Things.

Things that Terrorists may be curious about of course,

But Industrial Secrets as well.

Propitiatory Trade or Guild Secrets would be added to The List of Confidential Particulars.

Personal Information that many people are bashful about being revealed would be withheld from these Information Scouring Pads.

Anything that The Librarians of Enlightenment deem Sufficiently Esoteric to Require Inclusive Interpretations or Explanations for, Would Not want these Shards of their Edicts & Precepts so easily available to anyone without A Proper Jurisdiction to Clarify them.


Any Legal Issue that provide Lawyers with their Cheese & Crackers,

As well as Medical Information that Doctors currently have a monopoly on, would be available to everyone.

In Fact; Nearly Every Job that doesn’t require A Kinesthetic Mastery to Perform, would be suddenly accessible to Everyone.


If this service were to suddenly become available with A New Cell-Phone, The Economy would collapse within The First Week.

Maybe this Service would help us out of this disaster,

Or maybe it would find it intractable,

& then shut down as basic services, Such as cell-phone networks collapsed under The Weight of A Buckling Fiscal Bureaucracy.


Maybe that’s why Siri doesn’t work.

No comments: