Can an algorithm know more about you, your choices and intentions than you do yourself?
It was only by the time he turned 21 that Yuval Noah Harari understood he was gay. When questioned about a new algorithm that can decipher whether someone is gay or straight, based on a few facial images, the celebrated thought leader acknowledges that increasingly algorithms know more about us than we ourselves can possibly do.
Using his own example, he says hypothetically if an algorithm available with Coca Cola had deciphered that he was gay even before he knew it, they could have served up ads with photos of a shirtless man, while Pepsi (if unaware of his sexuality), may have sent him ads with women in bikinis. As a result, he would automatically veer towards Coke. This is how a corporate can manipulate our choices insidiously and sell us anything – product or politician — without our even being aware of it!
Scary thought, isn’t it? The recent documentary ‘The Social Dilemma’ has brought to the fore fears of privacy violation, polarization, and spread of fake news and communal tensions through internet companies. This is driven by their profit model, which thrives on extremely focused target advertising based on personal information gleaned from our internet usage.
Be it your shopping, music, food, reading or movie choices, by collecting all the right biometric data, companies can hit you with the right emotional message at the right time.
The greater danger is that most of this works at a subliminal level, such as the hypothetical Coke example above, where we are influenced at a subconscious level without even realizing it. When you know the danger of being influenced, you can work against that, but what happens when you do not even realise you are being manipulated? Subliminal messages work at the level of the subconscious – advertisers flash a message for a period shorter than the conscious mind can catch, or one too subtle for it!
But the same algorithms that pick up our biometric details can also be used to benefit us. Yuval talks of a health algorithm that could constantly monitor the body without our being aware of it. It could help us deal with health issues well in time. Or, an algorithm that predicts markets in time for you to make profitable decisions. Or, one that understands your sadness, and plays a song that will uplift your mood!
The subconscious is privy to information that we know nothing of. That is why sometimes we just know something without being able to explain why or how, and we call it instinct. Ultimately however, whatever AI uses will come from us… we are indeed being played by machines!
All these years we have been told to tap into that vast unknown within us and unleash our power. While humanity is still engaged in an attempt to do so, what if now technology is doing that to benefit others and to insidiously manipulate us and our choices with secret sales pitches? As with all else, hopefully humanity will ultimately tread the fine line between the advantages and disadvantages. Do we need an AI code of morality?
A double-edged sword indeed!