Feature 
 

Humanity

AI Apocalypse

Artificial Intelligence (AI) is simply showing the mirror to humans in which they can see their ugly selves, and one example would be the rise of racist robots, as rightly identified by the World Economic Forum.

By Atif Ilyas | April 2024


In his book ‘A Suitable Boy,’ Vikram Seth says, ‘God save us from people who mean well’. If you cast a glance at the governments of powerful Western forces, you would see their trenches filled up with people who supposedly ‘mean’ well. From dropping nuclear bombs on Japan to war crimes in Afghanistan to heinous crimes being committed in Gaza, everything is being done under the garb of ‘doing it for the greater good’ by the people who ‘mean well.’ Therefore, amid all these, the discussions around artificial intelligence’s ethical and moral conundrums and its impact on human life seem hypocritical at best.

UNESCO’s first-ever global standard on AI ethics – the ‘Recommendation on the Ethics of Artificial Intelligence’ – was adopted by all member countries in November 2021. It states that ‘the protection of human rights and dignity is the cornerstone of the Recommendation, based on the advancement of fundamental principles such as transparency and fairness, always remembering the importance of human oversight of AI systems.’ The last phrase, ‘human oversight of AI systems’ and its latent hypocrisy, can make one smirk with disgust because the marginalized part of this world is already reeling with the so-called human oversight of AI systems. Therefore, it makes one wonder what kind of moral and ethical quandaries AI poses and how they differ from the ones already footed by humans who were supposed to be the ones with a moral compass.

The World Economic Forum (WEF) has published nine ethical issues on artificial intelligence; interestingly, some are not exclusive to AI. For example, the WEF highlights that AI will condense the distribution of wealth as AI-driven companies will possess most of the resources. However, it fails to identify that the world is already marred with deplorable inequality where the wealthiest 1% own almost half of the world’s wealth, making it a human problem as well and not just the AI one.

Another ethical issue highlighted by the WEF is the inherent artificial stupidity in AI systems, which induces them to commit mistakes humans are biologically incapable of making. However, human error is also something that is stated as a universal truth, so it should be considered an ethical issue as well. We can further theorize that artificial stupidity in AI is just a consequence of human error as a creation can only be as stupid as its creator, albeit differently.


The truth is that AI is simply showing the mirror to humans in which they can see their ugly selves, and one example would be the rise of racist robots that are identified by WEF as well. For example, AI pioneers like Google faced a problem with its facial recognition programme that labelled the faces of several Black people as gorillas. Leading tech companies are spending millions of dollars to create racism-free algorithms where AI can sift through tonnes of data and draw meaningful connections and patterns that can make humans more productive. However, the problem is not with AI but with its creators, as AI cannot learn to be racist on its own. It is the skewed data being fed to the AI systems that is making them biased towards the marginalized segments of our population. Therefore, the debate should not be around ensuring ‘human oversight of AI.’ Still, it should be around how human intelligence has soiled something as elegant and extraordinary as artificial intelligence and what can be done to prevent it.

The problem is not with AI but with its creators, as AI cannot learn to be racist on its own. It is the skewed data being fed to the AI systems that is making them biased towards the marginalized segments of our population.

Much energy is being uselessly spent on making AI mimic humans, while the debate should revolve around how we can make AI systems lose the human touch. We do not need AI to possess human qualities because we already have them aplenty. Instead, we need AI to perform tasks for us that are unimaginable to a human mind, but before that, a boundary must be established. For example, the usage of advanced AI should be the greatest in industries, sectors and tasks where the need for objectivity is paramount. In contrast, its usage should be limited in industries that require subjectivity. Limiting AI influence and establishing boundaries will provide some semblance of control and help smother fears and contemplations around the consequences of AI that usually translate to apocalyptic scenarios in Hollywood movies.

We need to understand that human intelligence has created something extraordinary, not only because it has the potential to do gigantic tasks but because we finally have a first-world organism that is incapable of doing something out of ‘malice.’ It is biologically programmed to be selfless, and now all we need is to give it unbiased context and information, and this world can progress to a better place. But can we do it? Can we leave ‘humanity’ and mimic AI just for once?

Read More