VUCA World, Social Media and AI: Auguring a New Age (Part 1)

This article was written on 4^th December 2022 / 9 Jamadilawal 1444H. 

The VUCA World Today

The recent heart-thumping plot twists and cliffhangers dished out during Malaysia’s 15^th General Election drama is not a uniquely Malaysian phenomena. Neither will it be a passing quirk in history. What it has demonstrated is that the VUCA world has become super-VUCA. Volatility, Uncertainty, Complexity and Ambiguity (VUCA) reign in our world today.

The transmission of minute non-living organisms had brought the whole wide world on its knees within months of its discovery in 2020 and its effects are still felt today. The COVID-19 pandemic spread and killed millions and wrecked havoc more quickly than any other transmissible virus, partly riding on the same reasons that had driven prosperity across the modern world: systems and technology (trade, connectivity and transportation). It is an understatement to say that the pandemic had changed the world.

Our behaviors and other enablers that have shaped disruptive trends prior to the pandemic had gained further foothold. These have become “megachanges” (to use Daniel Franklin’s term in Megachange 2050, The Economist) and brought with them profound changes to our political, economic, and social landscapes.In light of these megachanges, some of our current thoughts and perspectives are now obsolete. There is no sensing of the future, as Dan Gardner shared in his fascinating book Future Babble. The ‘conventional wisdom’ of using the lenses of past scholars and changemakers who had created impact and change with their thoughts and actions based on the analysis of their moments in history may not be very useful, if we are not able to perform a contrasting frame of the different contexts.

But this is the trap that we fall into. Mamas, we forget that every moment of our existence as a mother is ibadah, too. 

Our Responsibility

While we are absolved of any responsibility to read crystal balls, leaders and activists should still be held accountable to create solutions to challenges and problems faced by our communities and nations.

أَلاَ كُلُّكُمْ رَاعٍ، وَكُلُّكُمْ مَسْئُولٌ عَنْ رَعِيَّتِهِ، 

“Surely! Everyone of you is a guardian and is responsible for his charges.”

Sahih al-Bukhari (7138)

There are several current and emerging trends we should be aware of. We should also take proactive measures to keep abreast with these developments to remain guided when we are affected by their implications and outcomes.

The confluence of the rise of Artificial Intelligence (AI), the ability to create content on the fly and the pervasiveness of social media have created an environment of misinformation at the speed of light, which is characteristic of a time where Al-Masih Ad-Dajjal’s (or The Deceiver) arrival is imminent. Failure to discern will lead to extreme confusion and further instability. Even though data and information are more readily available today than ever before, the nature of analytics itself is changing. AI engines in the back-end guide, package and massage the data stream that we consume.

The Rise of Misinformation & Confusion

Many of us may recall Cambridge Analytica’s use of data extracted from Facebook to influence the US presidential elections and the Brexit agenda. They have since closed down but to date, Facebook, Instagram and TikTok still operate engines which channel and curate their adverts and posts to profiles selectively (and sometimes, insidiously).

These algorithms of aggregating the “likes” while seemingly innocent, and sometimes also a good service to shield us from things we do not prefer, has its serious downsides.

The world, on one hand, is hyper-connected and becoming “smaller” and yet, it has become even more widely and deeply polarised across and between nations, communities and even families. Take the case of Brexit, where the United Kingdom left the European Union after decades of hard work bringing the European nations together, the animosities of the Hindu nationalists toward the Muslims in India leading to bloody attacks, and closer to home the brutal exchanges, mudslinging and political backstabs between leaders that are all Malays, Muslims and Qu’ran-quoting – which divides the Malaysian Malay-Muslim polity. 

We cannot play down the role of social media’s feed algorithms in reinforcing the human tendency towards selective exposure, leading to network activation and cascade of aggravating existing divides. These algorithms had inadvertently created huge online echo chambers, bouncing and reinforcing bias and misinformation which further entrenched misunderstandings and ill-feelings towards those outside these invisible chambers. These echo chambers are a mirage of transparency and openness. It has limited constructive exchanges that might have otherwise been able to allow readers or users to access diverse perspectives, unify, clarify or, at least, harmonise irreconcilable differences.

Beyond social media, the use of AI has been successful in many fields, performing specialised intelligent tasks without specific programming but by learning how to do it. The chatbots deployed by banks, commercial and government websites herald the successful open-ended customer engagement without human intervention. In Singapore, it is the familiar ‘Ask Jamie’ chatbot, which even gives the Salaam on the MUIS website. Your latest bank loan application may well have been assessed and approved by an AI capable of credit assessment, credit scoring and fraud detection.   

All these though are still in the category of narrow or limited AI serving specialised areas. They are far from the General AI like HAL 9000 from the sci-fi Space Odyssey or Genisys of the Terminator movie series which analysed and decided that the existence of humans served the biggest threat thus pursuing annihilation of humans.

The AI Controversy & Dilemma

While experts predict that such systems will not come into existence at least for the next 50 years on the average, practical and ethical issues have been raised even with current narrow AI deployments.

Take your Google maps direction feature, for example. It will guide you based on your preferences of best (most probably fastest) or least walking directions, or some other parameters, and you still may find that the routes suggested do not match your preferences or is even the shortest because it would suggest the use of expressways to improve speed instead of reducing mileage or particular sights that you would enjoy during your travels. Thus, you may well end up deciding on your own alternative route.

What is more worrying is that suggestions by AI systems gain very high confidence, acceptance and almost blind adherence. Grammarly is an AI use case that has gained much confidence in correcting our grammar for our reports and writing. It is supposed to make our writing more grammatically correct but yet, it may cause the loss of language and emotive nuances.   

Another example is the credit approval systems learnt from datasets of previous successful loan approvals and rejections. Unfortunately, this has been shown to incorporate biases against segments of the society, such as women or certain racial groups. This may deprive them from much needed credit to sustain and grow their businesses or help them tide through life’s hard patches.

We should not strip AI from all human biases. If we do, the solutions that AI generates will appear inhuman or detached. Pragmatism without love and compassion has been attributed to the basis of Nazism in Germany and the movement’s subsequent brutal genocide. AI needs to be taught some human biases while nurtured with higher order values to make or suggest humane decisions and actions.

Some of the current practices in guiding deep machine learning should incorporate ethical considerations. This includes being unbiased towards particular racial groups and women. While this is a promising start, the question is: which moral values and standards should we be basing the learning on?

The site chat.openai.com, for example, disclaims that “the system may occasionally generate incorrect or misleading information” and may even “produce offensive or biased content”. It even goes on to warn the user that “It is not intended to give advice”. For simple questions, like ‘How to scale a fish’, it is easy to get a straightforward answer but the same cannot be said for more philosophical and emotive questions or statements.

The issue here is that with a large set of “good” responses, confidence in its responses goes across the board. Do the masses that go online to seek an answer to their life challenges have access and opportunity to read alternative views and solutions that are more morally representative? Or will they fall into a collection of echo chambers driven by AI engines that promote certain values or lifestyles that contaminate our established moral values and systems?

Disclaimer: The views and opinions expressed in the articles on The Ocean’s Ink are the authors’ own, written in their personal capacity. They may not reflect the view of The Ocean’s Ink or IMSGP as an organisation.