Can AI dominate human society? A cliché question without answers. The definition of 'AI' itself is also vague, as the boundaries of 'robot' and individual thinking systems are ambiguous. But the absence of definition cannot be the insurance for it to stay as the subordinates. Why am I pondering upon this endless question? It all started the day I failed the CAPTCHA test for the fifth time.
It was that day, where a sudden CAPTCHA test demanded me to prove myself as a human being during login. The simple identification required me to click on all blocks containing signs. Abrupt indeed, but challenge accepted. The immediate dilemma was to whether include the block with the tiny fraction of sign or not. My mouse hovered over the block. Faltering for a moment or two, I absentmindedly clicked on the block, submitted it.
Try again.
Dumbfound, I clicked again. I thought I had finished clicking on all blocks with bus when the notice told me, 'Try again'.
So that is how I was rejected from logging into my account, accused of being a robot.
"Failed to prove human. Reason: Lack of humanity" Scene from a Korean cartoon <퓨리스틱> https://comic.naver.com/webtoon/detail.nhn?titleId=733771&no=3&weekday=sat |
The irony is that CAPTCHA is also a robot. It is a maze designed by men to trap the non-men using a non-man method. Is it effective? Recently a 'lying robot' had slyly ticked off the CAPTCHA box without any oppression. In its first trial. This again left me dumbfound, because I read this article after my series of unsuccessful login trials.
CAPTCHA is not the only trap designed for humans. Many other tools 'made to aid human activities' are in reality acting as oppressions. Nobody trusted a spelling checker or grammar editor like Grammarly in the past. Now students are using them to perfect their final drafts. One of my friends actually admitted that his scores are better when he writes his paper in his mother tongue, uses a translator, and uses a grammar editor. Whatever information is already in the robots, or has been incluated, now has become the wires of the cage in defining what is 'right' in human activity.
Let us talk a little more about my friend. How could he get better grades by the assistant of the artificial assistants? I asked my other friends, who told me that his teacher always preferred the Grammarly-edited version of writing. Some suspected that it was actually used in the procedure of grading - another opinion was that his teacher may have been 'tamed' into the Grammarly-refined form essay, and since the machine helped keeping the consistency, the essay turned out to be much more 'static'. This was not just happening in his classes. Newspaper clubs would use such checkers instead of cross-proofreading. BonPatron and Duolingo, both developed by data science, have been my best friend when studying French. Google Translate was widely used to convert the English textbook into Korean version, which would have never been purchasable.
Before I realized, the whole campus was dominated by artificial intelligence.
Yes, machine-dependency may sound like trustworthy insurance, but maybe it is time to take a step back and think again, aren't we relying on it too hard?