Monday, November 30, 2009

Is it good writing?

by Olga LaPlante

At some point in my life, I subscribed to the Current Events by Izzit. The site offers other classroom ideas as well, and they often have a somewhat controversial content, which is great - whether you actually agree with the writings and clippings from press being of high quality - for generating discussions, because they often contradict the mainstream news and accounts, and I wouldn't bet my money to say that they are unbiased. Again, the value is in generating a discussion, plus the lessons have questions, and vocabulary to learn.

Anyway, this lesson was published just recently - the lessons are free and they stay online for a couple of weeks. This will make you chuckle at least, but also make you think. I would love to see comments on this one.

PS. Mind the beautiful English language, with words like "mark" (Br) and "sit" for "grade" and "take" (Am) or references to A Level English Exams.

6 comments:

Pam Kenney said...

The notion of computers rating writing strengths and weaknesses offends me on many levels. The salient inherent difficulty, however, is that a computer cannot incorporate the writer's purpose or audience into its assessment.

If I'm writing an article to exhort my readers to action, my semantics and syntax will be very different than if I'm writing one about a new tax law. I will probably repeat phrases for emphasis, overuse metaphors, and sprinkle some colloquialisms here and there. I'm not a lousy writer; I'm trying to catch and kept my readers' attention, to spur them to act.

Computers can do a lot, but they're not able to think about and analyze what I'm trying to say, and until they can do that, I'd prefer human feedback.

Jim Burke said...

"Computers can do a lot, but they're not able to think about and analyze what I'm trying to say, and until they can do that, I'd prefer human feedback."

Here's my question, Pam, Olga, and anyone else: Do you think that computers will ever be able to analyze, question, and think at a human level?

See http://www.youtube.com/watch?v=9PWXrnsSrf0

and http://en.wikipedia.org/wiki/Technological_singularity

Jim

Pam Kenney said...

I listened to Ray Kurzweil, and think I understand his Law of Accelerating Returns, although my limited knowledge of computers obfuscates all but the basics. Perhaps one day computers will be able to think, but I'd like to believe that day is so far in the future that I can tuck its emergence way back in the recesses of my brain.

Jim Burke said...

I hear you, Pam, but those singularity people think that incredible machine intelligence is right around the corner . . . perhaps in 25 year. This is all based on math extrapolation. Interestingly enough, in my Telstar group this afternoon, a Spanish teacher brought up how poor online translators are and that she felt they would never become accurate enough to be very useful. I think Kurzweil would disagree. ;)

Suppose we'll be around in 25 years to see what actually happens? ;)

Jim

Pam Kenney said...

Oh, probably, but I may not be compos mentis enough to raise a fuss either way.

OLAPLANTE said...

"a Spanish teacher brought up how poor online translators are and that she felt they would never become accurate enough to be very useful." While I strongly believe that online translators are lousy, they are still helpful to get at least some idea - sometimes incorrect - about the text.
What is remarkable though is that professional translators use translating software, except they create that database themselves. Say they start with one document, and translate it; the bilingual reference is stored in the database. Next time you do another document, obviously, similar in content, the software will offer the corresponding translation of similar phrases in a friendly way, so you can go back and clean it up, should it be different enough. Also, the more people add to the database, the more accurate it becomes. Also, when you ask Google to translate a page, you can go in and offer your own translation if the one displayed isn't good. The only thing is I don't know what Google does with your version.
No, I don't think computers will be able to think, or have high touch - although it's quite likely they will be able to simulate it at various levels.