Yesterday I prepared a whole post about my vote, which the website swallowed and destroyed when I hit submit. :( Ah well.

Victor, there is a difference between changing a convention (whether double negatives cancel themselves out—as in English—or add emphasis—as in German) and changing a fact (2002 vs 2012). As long as a convention is self-consistant, consistant with other rules in the language, and commonly known, then there is nothing wrong with adding it to the grammar organically rather than formally. The key thing is that the new grammar maps to the same facts (concepts, whatever you want to call them) in your head.

C R Hunt - Jan 24, 2012:

Yesterday I prepared a whole post about my vote, which the website swallowed and destroyed when I hit submit. :( Ah well.

I got a really good chuckle out of that.  Not laughing at what happened, but the way you describe it is hilarious   It has happened to me more than once… usually, for a very long-winded reply, I just crank open notepad, type my stuff and copy and paste.

Don’t laugh..its a good habit, so many websites have destroyed my submissions!!!!!

C R Hunt - Jan 24, 2012:

As long as a convention is self-consistant, consistant with other rules in the language, and commonly known, then there is nothing wrong with adding it to the grammar organically rather than formally.

So as long as we don’t have group of people ‘A’ thinking that “don’t do nothing” means “do NOT do anything” and group B thinking (as it really is), “DO something”.

The 2002/2012 thing was just a silly extreme case of the ‘don’t-do-nothing’/‘don’t-do-anything’ example.  ‘Don’t do anything’ can be considered a fact.  The speaker, in fact, does not want you to do anything.  NOT doing nothing means doing something, so the fact is incorrect.

I think I’ll have my bot insult someone for saying ‘don’t do nothing’.  Just kidding. 

Actually, I will have an option checked or not check, whether you want it to teach you.

On another note, interesting to see “3, 1, 2” gaining popularity.  Personally, I think 3 is only possible after 1 is done.

Victor Shulist - Jan 24, 2012:

Example, teenagers have a language all their own.  They are currently using the word “sick” to mean something, I **think** to mean “exciting” or “exceptional” ??!!?!

Recent Skynet-AI dialog:

USER:The truth can be unsettling and scary to some…not to me though
BOT: I am never frightened, I have nerves of steel.
USER:cool
BOT: Thanks, being cool is quite important to a supercomputer.

Language evolves much the same way that species evolve. Originally, different languages arose because of geographical obstacles limiting the spread of any given dialect. Yet many features are still common to entire groups of languages, and other features are lost or reacquired over time through subsequent interaction.

Most language constructs are invented and seem to start out “regular”—that is consistent with other similar constructs. They can become irregular over time through common usage. Those phrases that are used frequently get shortened and blurred together until they may become unrecognisable.

Other times, things change owing to fashion or ignorance. For example, “ain’t” is a valid word, it is the contraction of “am not”, but for whatever reason, common folk began to use it as a contraction for “had not” etc. This caused it’s use to become associated with ignorance, and it fell out of favour, even when used properly. Another example is when comparing two things. We used to say “different from” but so many people over compensate in their efforts to be correct that nowadays most people think that “different to” is the correct way to compare things. Another example of that sort of thing happening is words like “orient” that are changing into words like “orientate”.

There are plenty of examples of the same words meaning two different things in different places too. Just about everyone here will have had the experience of talking to someone who speaks a different dialect of English (e.g. Australian, North American etc) leading to awkward misunderstanding (e.g. “fag” used to mean “cigarette” in Australia).

Andrew Smith - Jan 24, 2012:

There are plenty of examples of the same words meaning two different things in different places too. Just about everyone here will have had the experience of talking to someone who speaks a different dialect of English (e.g. Australian, North American etc) leading to awkward misunderstanding (e.g. “fag” used to mean “cigarette” in Australia).

Ha! I was just about to mention the same thing.

I still remember the first time an English person asked me, “Are you alright?” She made it sound like I’d lost a leg or something. But that’s just English for “How’s it going?” Figured that out from her strange looks after bumbling about for a bit justifying why I might look out of sorts but was actually just fine… Ah, well. Crazy Brits and their foreign language.

CR, that particular example isn’t limited to Brits. That same phrase is used in the same way in many locations, particularly in the “south”. I ran into this while living in Tennessee, many years ago.

Or maybe America’s “Redneck Roots” originate in the Empire? Who knows?

Most language constructs are invented and seem to start out “regular”—that is consistent with other similar constructs. They can become irregular over time through common usage.

That’s odd, I thought it was the other way round: over time irregular verbs are pressured into the regular form (lower usage = easier change, so ‘be’, ‘have’, ‘do’, probably wont change very soon:
http://notexactlyrocketscience.wordpress.com/2007/10/10/the-evolution-of-the-past-tense-–-how-verbs-change-over-time/

I voted 1,3,2 because I regard 1 and 3 as the same - 3 is just an extra difficult version of 1 which may rely more on interpreting statements using context and the rest of the conversation (but this is also helpful/required for 1). 

I don’t rate 2 as highly because I think much of the knowledge is “out there” and doesn’t need to be stored or pre-processed (though that may help speed things up).  The problem is, most of it is encoded as natural language, so then we’re back to problem number 1 or a variant of it that’s appropriate to the source.

Good points Oliver.  Also, someone, I don’t feel like scrolling through right now, I think either Andrew or Jan, pointed out that probably a 4th component is necessary to have in the list, and I highly agree—- knowledge acquisition routines.

Yes, 3 is a subset of 1.  In fact, dealing with bad grammar is just adding layers of permutations—just more pre-processing, that will take more time and energy, CPU cycles.  Thus, to have a more efficent, and faster response time from an AI, try to minimize your spelling and grammar errors.  And of course, other ends of the extreme will probably require a time-out, example, if someone enters a 200 word sentence where the grammar makes no sense and every single word is misspelled…. imagine the astronomical number of permutations (200 ^ (similar-words-each-could-be) !!!!! yes, another crazy extreme example… but to illustrate a point—how will you know when your engine should say ‘the hell with it’ ....  Perhaps just a user-defined time-out for processing.

Of course, as we tweak our algorithms and Intel/AMD make their processors even faster, we can care a bit less about having to have these extra spelling-check, grammar-check wrappers of permutations and preprocessing. 

It brings up a good question—what is the level of tolerance of your bot?  I haven’t really defined mine, but perhaps, 5% of the words misspelled is being tolerant enough?  (after that, a dialog box appears…. “your input, after <x> minutes now of processing, isn’t making any sense to me , continue trying spell check permutations???”

For a ‘real life’ production system that would be useful, but for a Turing Test, it would be bad ... instead ‘bust’..... perhaps after a certain timeout, it gives up, and goes into ‘Eliza-mode’, and picks any word it does know, and responds with some mapping.