Artificial intelligence vs natural stupidity: K.C. Verma
What is this chat jee-bee thing?” asked the little woman, without any preamble. Bleary eyed as I usually am early in the morning, I was in no mood to be quizzed. But the little woman has this endearing habit of reading the newspapers while sipping her morning tea and sharing tidbits about miscellaneous disasters. She believes that after this early battle inoculation, one can easily shrug off whatever slings and arrows outrageous fortune hurls at us later in the day.
“Eh?” said I. “Chat is what you get in Bengali Market, New Delhi. If you want really good papri-chaat, I recommend the lanes of Chandni Chowk.”
“Hello stupid! Wake up! I said Chat gee-pee-tea, not Chaat!”
“Oh, that is an AI thing. You won’t understand it,” I said dismissively, pouring a cup of tea for myself. After a few sips, I felt quite fortified to handle any slings and arrows that might be hurled by the missus. Okay, maybe I was not fortified; but I certainly felt twentified enough.
“Is it a thing? Or a person or what? Some girls at the kitty party said that it is the new demon that has been let loose on humanity. One girl said that Modi ji had gone to Paris to tame this beast.”
I was surprised to hear that my wife’s kitty group discusses topics like AI. I was even more surprised to hear that the babushkas refer to themselves as girls. Well, so long as they don’t fancy themselves to be nubile nymphs, I said to myself. Aloud I asked, “Oh! And what else did the girls say?”
“There was some talk of a new Chinese dragon. I suspect it was a joke because they said it was named Deepsh*t. Now, you’re ever so clever, so tell me about this AI thing.”
Even though I had yet not had breakfast, I was sweet-talked into educating the missus about artificial intelligence.
“You mean that when my phone screws up my messages by changing spellings—that’s artificial intelligence?”
“Well, in a basic way. But there are other applications. Indeed, AI will soon be everywhere. We already have many examples of the technology of tomorrow. Are you aware that your washing machine can send emails to the manufacturer for troubleshooting?”
“What?” said the missus, saucer eyed. “That’s so creepy! That machine sends emails behind my back to people I don’t know! Instead of washing, it’s showing our dirty linen to strangers?”
“Well, you better get used to such things. Your fridge would know when you were running short of butter and place an order for more. Your air conditioners would automatically switch on before you reach home. If you hold a doorknob, it will measure your blood pressure.”
“All of a sudden, why do you sound like a vendor of AI?”
“Oh no, I’m not!” I protested. “However, the age of AGI is here and soon computers will be able to make more nuanced decisions. AGI will result in better governance, a better society and improved quality of life. But the potential for misuse is tremendous. Just imagine, a computer might order a nuclear strike if it learns to think maliciously. Even at a mundane level, there are unknown dangers posed by polymorphic malware.”
I confess I don’t have the foggiest idea what polymorphic whatever means, but one must impress one’s wife, no? Anyway, I rounded off my lecture by declaring pompously, “With great power comes great responsibility. We need to use AI capabilities in a judicious and ethical manner.”
“Don’t you think you are being a bit patronising? If AI is so bloody intelligent, it should become self-aware and be able to itself overcome all ethical dilemmas. It should also develop a conscience and then, before you know it, it will have likes and dislikes and petulances and quirky behaviour. Then it will itself decide whether the saintly X will use it or the unethical Y.”
That left me speechless. I looked at the old girl with new respect. She had given the ethics debate a unique twist! I wondered which quality in my wife had deceived me earlier—her natural stupidity, her artificial stupidity or her artful stupidity? Or have I been underestimating the educating power of kitty parties?
But I still needed to make my point. “You mark my words; AI will certainly prove to be a Frankenstein some day!”
“You mean Frankenstein’s monster, don’t you?”
“Yes, I mean Frankenstein. The monster.”
My wife tut-tutted. “Victor Frankenstein was the name of the guy who created the monster. The monster’s name was not Frankenstein.”
“No, Frankenstein was the name of the monster,” I insisted.
“You’re so wrong! Why don’t we Google it?”
“Let’s do better than a simple search. Let’s use DeepSh*t for a more nuanced answer.” And I typed in my query, ‘Tell me something about the monster Frankenstein’.
Within a couple of seconds, the computer spewed out—‘Frankenstein was not the name of the monster. Frankenstein was the name of the creator. The problem with you humans is that you so often forget who the creator is and who the creation!’
Now that was a patronising attitude! “Stupid DeepSh*t,” I said and banged the laptop shut.
K.C. Verma is former chief of R&AW. kcverma345@gmail.com
Columns