Adam Smith And The AI-Powered Invisible Hand

Reading Time: 3 minutes

June marks the tricentenary of Adam Smith – a Scottish economist, philosopher, and considered by many as the father of modern economics. Modern society is built around his ideas of prosperity powered by human effort, division of labor, and the concept of the invisible hand guiding market economies. 

Writing at a time when the industrial revolution brought new ways of thinking about work, human labor, and societal transformation – Smith’s Wealth of Nations has left a lasting impact in the field. One of the book’s central thesis is that lifting productivity, specialization in tasks, or better division of labor can directly impact living situations and yield better conditions for growth. 

But it’s been close to 250 years since the book was published. We have moved from working models of steam engines to AI-driven engines. Does Smith’s way of seeing the world still hold at a time when artificial intelligence seems to be animating the invisible hand itself? How can the Wealth of Nations lessons be translated to businesses at a similar inflection point – where new, growing technology is transforming commerce as we know it and redefining the nature of labor?

50,000 Feet & 50 Feet

Unsurprisingly, world productivity levels have dramatically shifted in the last centuries. The past few decades, however, have seen a steady and significant slowdown. The world bank released a study that pointed out that global productivity growth has been on a downward trend for the last decade, directly impacting living standards. 

The corollary is that studies have also that shown AI solutions have improved productivity by as much as 14%. A Goldman Sachs study shows that advances in NLP and machine learning could drive a 7% increase in global GDP and lift productivity growth by 1.5 percentage points over ten years.

Global GDP is the 50,000 feet angle, however! 

The 50 feet outlook is seeing how AI impacts productivity and shifts perceptions of what it means to be efficient for businesses. The answer is intricately tied to how businesses deploy AI. And once adopted, what can be considered as wins.

Automating The Mundane

Businesses today are increasingly looking at three things when considering a new technology – does it improve some metric within the company, does it add value to the end-user experience, and how effectively can the technology align a global workforce that is as connected as it is remote? 

Decision-making for any business is never wholly fluid or linear, especially when the end-user is fickle, and the nature of the business itself is capricious. Everything from managing inventory to understanding inventory to deciding what products need to be sold, when, how, and in what ways requires the ability to parse, filter, and build rules from all the data. Studies have shown that even when people are given identical pieces of information, how they process it, understand the urgency of that decision-making, and use the input differs widely from person to person. 

AI-based tools and solutions need to be seen as the answer. They reduce the variability and inconsistencies resulting from the diversity of decision-making. AI ensures productivity as a metric is not about the number of hours put into a task, but the critical decisions it allows people to make, when the mundane is automated. AI also is getting better at consuming massive volumes of data, learning from it, and creating rules that positively impact the end user. 

Deploying AI to facilitate the decision is only half the solution, however. Oversight for a rule created by the AI, signaling the confidence level in the decision, and understanding where subverting the AI is necessary remains broadly within the human realm. AI lowers the cost of wild predictions and allows people to base their decision-making on parsed data. It lowers the opportunity cost for businesses. 

All AI is Equal – Some More Than Others

There is a catch, though – All AI are not made the same! For automation and machine learning to be effective, they cannot be seen as yet another technology that is good to have. AI is a must-have in as many avenues and as many processes within the business as possible. Only then can it add value and move metrics. It is not enough to have personalization on a e-commerce site if all the AI does is help recommend products that commit the user to a specific journey and take them down the same chute every time. 

A good AI model is contextual, prioritizes the complete data over piecemeal information. It helps get a 360-degree view of the product, user, and site every time.

AI has always been seen as the harbinger of increasing task productivity and bettering labor’s value. Research shows that the share of jobs requiring decision-making increased from 6% in 1960 to 34 percent in 2018, with nearly half of the increase occurring just since 2007. 

Adam Smith’s work revolved around three things – productivity, labour experience and how effectively information could be transferred though society. Retailers today revolve their industry around three things – internal metric, user experience, and the ability to enable teams to move faster and more intelligently. The hand that will power every one of these today is AI.