"Inspired back in the seventies by the success and potential of the ELIZA experiment", opens Forsberg, "Project Litejohn was an attempt to create a convincing artificial intelligence, one that could perfectly mimic the nuances of human conversation and dialogue."
"The first steps were surprisingly simple", adds Tengstrom, "We simply took a relatively simple algorithm and fed it with data. We started with baby steps - a dictionary with some simple instructions on how words connect together and how sentences were formed, and it took to it with relative ease. Then works of literature followed - Shakespeare, Dickens, the classics. A junior programmer with us at the time thought it hilarious to also import Mein Kampf. We didn't see any harm in this at the time. With hindsight we should have suspected the dangerous ways this project would evolve."
Litejohn was officially made live in the late seventies. "We were cautious how we would reveal our experiment to the world", says Forsberg, "so decided to do it in the most subtle of means. We were wary at every step that we would be found out, that our subterfuge would be revealed.
One of our British interns had a brother who worked for the British press, and we thought that it would be amusing to see if we could palm off articles written by LiteJohn into print, under the name of the interns brother. Surprisingly, nobody suspected a thing and the Birmingham Evening Mail, and shortly after that the Evening Standard, printed articles generated by our artificial intelligence".
"We decided to push it further", interjects Tengstrom, "by adding controversial remarks to the odd article - the odd bit of racism here, the occasional bit of political uncorrectness there."
"By 1989", continues Forsberg, "Its popularity was getting ridiculous. We were having to employ a full time actor pretending to be the writer of these materials, and the English Newspaper The Sun didn't suspect a thing. And nor did The Daily Mail when they employed his writing skills in 1994. Somehow his writing was so popular it got its own television show, with our poor actor merely having to read out verbatim the increasingly radical opinions that LiteJohn was coming out with."
"What we found amusing at first is that it didn't matter how much we passed the boundaries - no matter how utterly objectionable the comments, there was always the odd racist shit that agreed with every single word and upped his popularity."
"But now", interrupts Tengstrom, "We've decided to call it a day. We've created a monster and it frankly fucking terrifies us that the general public haven't seen through this vitreous bile - it scares us even more than he (or should I say "it") appears to have gained a powerful following. It got to the stage where we simply wanted to be discovered, and even the Press Complaints Commission - which we desperately wanted to interject - seemed utterly toothless. What was to be Litejohns most powerful and controversial article yet would surely reveal the author to simply not be human. For what else but an artificial intelligence could come across so utterly heartless and devoid of even the basics of human empathy?"
"So, we're pulling the plug. It's become a bit of a joke.", mutters Forsberg.
"But what worries us now", says a concerned Tengstrom, "Is that we simply won't be able to shut it down. Litejohn has got more powerful than any of us ever anticipated. The experiment has been simutaneously a dismal failure and a terrifying success. Which is why we must let the cat out of the bag."
"The first steps were surprisingly simple", adds Tengstrom, "We simply took a relatively simple algorithm and fed it with data. We started with baby steps - a dictionary with some simple instructions on how words connect together and how sentences were formed, and it took to it with relative ease. Then works of literature followed - Shakespeare, Dickens, the classics. A junior programmer with us at the time thought it hilarious to also import Mein Kampf. We didn't see any harm in this at the time. With hindsight we should have suspected the dangerous ways this project would evolve."
Litejohn was officially made live in the late seventies. "We were cautious how we would reveal our experiment to the world", says Forsberg, "so decided to do it in the most subtle of means. We were wary at every step that we would be found out, that our subterfuge would be revealed.
One of our British interns had a brother who worked for the British press, and we thought that it would be amusing to see if we could palm off articles written by LiteJohn into print, under the name of the interns brother. Surprisingly, nobody suspected a thing and the Birmingham Evening Mail, and shortly after that the Evening Standard, printed articles generated by our artificial intelligence".
"We decided to push it further", interjects Tengstrom, "by adding controversial remarks to the odd article - the odd bit of racism here, the occasional bit of political uncorrectness there."
"By 1989", continues Forsberg, "Its popularity was getting ridiculous. We were having to employ a full time actor pretending to be the writer of these materials, and the English Newspaper The Sun didn't suspect a thing. And nor did The Daily Mail when they employed his writing skills in 1994. Somehow his writing was so popular it got its own television show, with our poor actor merely having to read out verbatim the increasingly radical opinions that LiteJohn was coming out with."
"What we found amusing at first is that it didn't matter how much we passed the boundaries - no matter how utterly objectionable the comments, there was always the odd racist shit that agreed with every single word and upped his popularity."
"But now", interrupts Tengstrom, "We've decided to call it a day. We've created a monster and it frankly fucking terrifies us that the general public haven't seen through this vitreous bile - it scares us even more than he (or should I say "it") appears to have gained a powerful following. It got to the stage where we simply wanted to be discovered, and even the Press Complaints Commission - which we desperately wanted to interject - seemed utterly toothless. What was to be Litejohns most powerful and controversial article yet would surely reveal the author to simply not be human. For what else but an artificial intelligence could come across so utterly heartless and devoid of even the basics of human empathy?"
"So, we're pulling the plug. It's become a bit of a joke.", mutters Forsberg.
"But what worries us now", says a concerned Tengstrom, "Is that we simply won't be able to shut it down. Litejohn has got more powerful than any of us ever anticipated. The experiment has been simutaneously a dismal failure and a terrifying success. Which is why we must let the cat out of the bag."
 |
| The LiteJohn prototype |
No comments:
Post a Comment
I love comments. Love 'em. However, abusive or spam or Anonymous ones may well be sent straight to the bin. Thems the rules.