Do I have to hate people who don't exist?

I had a disturbing thought earlier when Veeshir left a comment pointing out that a commenter (who calls himself "Steve") appeared to be soliciting links. I remembered that in an earlier post, the same commenter had solicited a link from another blogger, even though that blogger had not even left a comment.

So I did something I almost never do. I Googled the content of the comment in question (his blog name plus the words "link exchange" and I got over 47,000 hits. Now, that's a lot. It would take me many years to leave 47,000 comments, and the blog only seems to have been there since 2008.

While I have very low standards, I really don't see why I have to have my time (and other people's time) wasted by commenters who are not human beings. As I said to Veeshir, "Is it asking too much for commenters to at least be people?" I realize that people are calling this commenter a "link whore," but if the comments are generated by a bot, then he's not even a real link whore, he's a fake link whore. Do there really have to be such things?

But I'm worried about a bigger problem. Troubling as fake commenters and fake link whores are, I began to wonder whether we might be on the verge of fake blogs. Blogs which are not written by human beings, but which crank out increasingly intelligent, readable posts, right down to taking controversial positions on heated issues. Since anonymity and pseudonymity are with us to stay, how are we to ever know whether a blogger is in fact a human being? How are we to verify a blogger's personhood? Or are my concerns justified? Perhaps it doesn't matter. I mean, there could be hundreds of thousands of fake "conservative" blogs, all linking to each other, while attacking and deriding the hundreds of thousands of fake "liberal" blogs. Gigantic fake blog wars. No real bloggers would be needed.

As it is now, I often feel as if I am repeating myself. There are millions of words in this blog, and I see no reason why someone couldn't design a program that would just cut and paste stuff I've said into coherent dialogues on whatever the latest news is, using simple search terms to spot and flag items of concern for the fake discussion, of course. It would make my "job" easier.

There is something downright demoralizing about knowing how ultimately replaceable we all are.

At least I know I'm alive and writing this post.

Or do I?

MORE: Needless to say (as if anything would be needful to say), the program used to generate the typical fake blog posts in the typical fake blogs could be based on this famous but typical incendiary blog post template that so many people have linked that I don't know whom to credit. (Probably Gates of Vienna, via Instapundit, both of whom are known to still exist, even if the latter contemplates his cyber-replacement occasionally.)

I wouldn't mind it if the fake blogs pursued their fake blog war indefinitely, as long as they leave us real folks alone. What I don't like is the idea of political robot blogs being set up and made to look real as part of some sick "experiment."

Would that mean they're not true fakes?

posted by Eric on 02.23.10 at 07:48 PM










Comments

Are you Peter or are you Valentine?

Captain Ned   ·  February 23, 2010 8:35 PM

There are lots of fake-yet-readable computer generated blogs already. They are mostly used by internet marketers for search engine purposes, but I can see them being turned towards politics very easily.

Tim   ·  February 23, 2010 9:30 PM

Yeah, I have seen the fake spammer type blogs, but I'm seriously wondering aloud whether political blogs have already been faked. There is absolutely no reason to believe they wouldn't be. Most likely this would start as "course work" by computer science, psychology, or political science students.

"We created fifty fake right wing blogs, programmed them to talk a certain way, and soon found them linked to innumerable other conservative blogs." (Obviously, this means that conservatives are mindless authoritarian dupes....)

The only way I can see to stop this is to demand proof of personhood!

Eric Scheie   ·  February 23, 2010 10:25 PM

So, "Eric", tell me about these electric sheep you dream of.....

guy   ·  February 24, 2010 12:06 AM

At DPUD there are the occasional spam comments that copy what someone else wrote and then adds a link to watches or wigs or something.
It's funny when it copies an eddiebear rant or something else, the spambots there cuss.

If someone can make a bot that can fool people, well, I think that's cool. And if some lefty uses it to "prove" something about the right, I'm sure Tom Maguire will figure out a way to funnily show the left is "more" whatever they proved the right was.


I don't think we'll be replaced anytime soon though though so your "job" is probably safe.

I used to read Omni probably 20+ years ago, mostly for the very cool short stories. I read it when it was new and didn't keep it up, so it could have been as long ago as 1981 (according to wikipedia), certainly no later than 1988.
They had a bot that wrote a couple stories. It wasn't bad, it even pretty much made sense.
And that was long, long ago. Computers have advanced hugely since then so they should be more advanced.

Remember the early days of Instapundit when poeple joked about how he was a bot?
How do we know Glenn Reynolds isn't a bot?

Seriously, he would be the easiest to fake and he's so prolific it's hard to see how he does any work when he would have to be reading the intertubes about 28 hours a day to see all the blogs he sees.

All you need is a bot to link stuff and add "indeed", "Heh" or "read the whole thing".
I could probably write that script.

Veeshir   ·  February 24, 2010 8:59 AM

I did research into Human Intelligence Tasks (HITs) on Amazon's Mechanical Turk crowd-sourcing system for a customer, and ended up working on around 60 tasks. There were many hundreds, probably thousands, of tasks that could easily have been used to generate fake blogs. I performed several that involved re-writing a paragraph but retaining specific keywords. Others provided 2 or 3 keywords and asked for 20-30 words that formed complete sentences about the topic. These tasks paid mere pennies each, yet hundreds of "turkers" as they call themselves were performing them 24x7.

Craig   ·  February 24, 2010 2:13 PM

Great post. I too often feel as if I'm repeating. You might be interested in my blog on intelligent and personhood:
#ERROR EMPTY LINK

AFFA   ·  February 24, 2010 9:02 PM

Post a comment


April 2011
Sun Mon Tue Wed Thu Fri Sat
          1 2
3 4 5 6 7 8 9
10 11 12 13 14 15 16
17 18 19 20 21 22 23
24 25 26 27 28 29 30

ANCIENT (AND MODERN)
WORLD-WIDE CALENDAR


Search the Site


E-mail



Classics To Go

Classical Values PDA Link



Archives



Recent Entries



Links



Site Credits