AI and Us: Decision Making Should Not Be Outsourced to AI

AI literacy should become as important as digital literacy. An informed user is ultimately the first and most effective line of defence against blindly accepting what a machine tells us.  

Image
AI and Us: Decision Making Should Not Be Outsourced to AI

When I started trading - not investing - in the stock market in 2016, there were a plethora of YouTubers, mostly self-appointed experts, advising all and sundry on what to trade each day. For a while, I followed a couple of them and found that I was incurring a net loss.

I then spent time learning the nuances of trading and began technically studying the market. I also homed in on some free online software that helped me with the analysis. Subsequently, I found that I was consistently tallying more winswhen the decisions were mine.

The experience taught me something that goes well beyond stock-market trading.

Seeking Advice vs Outsourcing Decisions

We make decisions every day, from small choices such as what to have for breakfast to complex decisions that affect businesses, teams and, sometimes, entire societies. While simple choices come naturally, difficult decisions require information, analysis, assessment of alternatives and, ultimately, judgement.

There is, however, an important distinction between seeking advice and outsourcing a decision.

Experts, advisers, algorithms and, increasingly, AI can provide information, identify alternatives and even recommend a course of action. They can make our decision-making better informed. But the responsibility for the final decision remains with us.

Ancient Lesson From the Gita

This is not a new dilemma. It finds a remarkably powerful expression in the Bhagavad Gita.

In Chapter 2, Verse 7, Arjuna, overwhelmed by his moral dilemma, tells Krishna that he is confused about his duty and asks Krishna to tell him what is best. He places himself in the position of a pupil and seeks guidance from his teacher.

Krishna then proceeds to explain the nature of the self, duty, action, knowledge and dharma. His discourse extends through the eighteen chapters of the Gita.

Yet, after all that teaching, Krishna does not make the decision for Arjuna. 

In Chapter 18, Verse 63, Krishna concludes his instruction by telling Arjuna, in essence:

I have explained this knowledge to you. Reflect upon it fully, and then do as you wish.

That final instruction is perhaps as important as everything that precedes it.

Krishna provides knowledge, perspective and guidance. He does not take away Arjuna's agency. The final decision remains with Arjuna, along with the responsibility for that decision.

The AI Within Us

This ancient lesson assumes a new significance in the age of AI.

Human beings have always had a tendency to seek easy answers. Social media has amplified this tendency. People readily consume stories that are emotionally satisfying and allow them to shape and channel their thoughts. Information that has been forwarded many times can acquire an aura of truth simply through repetition.

Yet clarity is not the same as truth.

Complex systems rarely obey simple narratives. Economies are influenced by thousands of variables. Political conflicts often emerge from centuries of history. Scientific discoveries require years of careful experimentation. Reality is messy, uncertain and full of contradictions.

But the human brain does not always enjoy complexity. It prefers stories that are coherent, simple and emotionally powerful.

AI takes this tendency a step further by delivering "knowledge" to order. It can search, process, compare, explain and formulate conclusions within seconds. That convenience is enormously valuable. But it raises an important question: How much of the thinking remains with the individual, and how much is gradually being left to AI?

AI as a Power Tool

There are two very different ways of using AI.

At one end, AI becomes a power tool. We formulate the question, bring our experience and context, examine the response, challenge it, exercise our judgement and take the decision. AI does much of the searching, processing, comparison and drafting.

At the other end, we delegate the entire cognitive chain: AI formulates the question, provides the answer, we accept it and act on it. The difference is fundamental.

The first approach augments human intelligence. The second risks atrophying it.

This frailty in human nature can often be our undoing. Reading, for example, requires the brain to employ imagination, correlation, assessment, retention and many other cognitive faculties. When "knowledge" is delivered to order, much like fast food, the effort involved in acquiring and assimilating it is reduced. If this becomes habitual, the brain itself risks becoming rusty.

The paradox is therefore striking:

The better AI becomes, the greater the temptation for us to think less. The long-term question is not simply how intelligent AI becomes. It is how much of our own intellectual activity we continue to exercise.

A Team Member, not the Decision-Maker

Perhaps the real challenge before humanity is not to resist AI, but to use it intelligently. We can seek advice. We can seek expertise. We can use technology. We can include AI in our team. We can even ask AI to help us think through a problem. But the decision and the responsibility for itmust remain ours.

The greater lesson for humanity today is simple: Use AI as a member of the team, not as the decision-maker.

Regulation and AI Literacy

This also raises a larger societal responsibility. AI cannot be left entirely to individual discretion.

Regulation must begin at the source, with developers and providers building appropriate safeguards, transparency and accountability into AI systems rather than attempting to correct every problem after deployment. At the same time, guardrails are necessary at the point of use, particularly where AI can influence consequential decisions.

But regulation alone will not suffice.

AI literacy should become as important as digital literacy. An informed user is ultimately the first and most effective line of defence against blindly accepting what a machine tells us.

Ultimately, the objective should not be to keep AI out of our lives, but to ensure that AI remains a powerful tool in human hands without becoming a substitute for human judgement.

(The author is an Indian Army veteran and a contemporary affairs commentator. The views are personal. He can be reached at  kl.viswanathan@gmail.com )

Post a Comment

The content of this field is kept private and will not be shown publicly.