An recent NY Times article asks a question that’s been vaguely floating around the edges of the business world since the launch of OpenAI’s ChatGPT 3 in November of 2022 – “Can A.I. replace human CEOs?”
Taking this question at face value, there are plenty of answers needed before “A.I. CEO” becomes anything more than a PR ploy, but it’s completely reasonable to believe specially tailored artificial intelligences (AI) will soon offer insights into costs, efficiencies, and even strategic growth planning. In 3-5 years, it’s likely that investors will require companies to use these tools, potentially even adding a Chief A.I. to many corporate structures.
But before we get there, the business community must confront a few fundamental questions:
1. What makes your company different?
A.I. CEOs will either emerge from internal development or be a third-party vendor product. Either way, companies will need to provide extensive training data and make decisions about risk tolerance, profit motivation, and other decision factors. To risk losing what makes a company “itself” (e.g. Apple’s design focus or Walmart’s low prices), the system would need to understand that these characteristics and weight them against other motivating factors.
Many Fortune 500 brands with large budgets may have clearly articulated brand identities, but the vast majority of American business don’t Their only fundamental goal is to make money in their industry by avoiding things that don’t make money. This creates an experience for the average consumer that all brands are basically the same and are in a seemingly endless race to the bottom to offer the lowest quality product.
With that kind of input, A.I. executives can only accelerate cost cutting and over optimization, leaving us all with lower quality products and poorer working conditions. Companies are left chasing market fads, underinvesting in long-term growth, and pursue short-term fixes that create long-term instability for themselves and the larger market.
To avoid this, companies should be carefully considering what guides them now. What value do you offer the market? What is unique about your knowledge, product, strategy, or service? Could that uniqueness be extended to new buyers with better efficiency or more reach? And most importantly, given a choice between a cheaper, lower-quality option and your company, will your customers still choose you?
2. Can leaders explain (and defend) their decisions?
One of the big buzz words of the moment is “interpretability”, which when applied to A.I. models refers to the human ability to understand how the system made a decision. If we extend the use of that word to a phrase like “leadership interpretability”, we highlight a very human problem that’s already at the center of our current business environment.
Most companies can’t understand the decisions that their leaders are making, either because leaders are unwilling to explain them or because the justifications may not be based on data. Leaders may make decisions based on intuition, “market insights”, or strategic planning, and many organizations are purposefully built to trust and act on executive decisions without question.
This poses a fundamental problem when A.I. begins contributing to leadership decisions. An A.I. CEO would need to be continually questioned and asked to show it’s decision-marking work. Corporate boards, other executives, and even the larger organization will want to understand how employee well-being, long-term sustainability, and short-term profits were balanced to arrive at a final decision, and what sources of data were used to inform those choices. How many human CEOs welcome that level scrutiny today?
Plenty of leadership books have been written about the benefits of creating teams of rivals or red-teaming important decisions, but very few organizations actually function this way. Deliberative decision making is slower, more difficult, and requires a deeper skill-set than “intuition” and “strategic thinking”, but it often leads to better decisions that the entire team can support.
Now is the time for businesses to develop cultures of deliberative decision making. Executives should share and question the data that informs strategy. Leaders should be open to and actively invite questions about strategy. And all business should clearly articulate an organizational vision that aligns employees, investors, and even customers to the long-term principles that will guide decision making.
If the business world doesn’t have this sort of muscle memory deeply embedded in operating culture before adopting A.I. input, expect to see some seeming incomprehensible decision making across the market.
3. Who is the economy for?
When human beings settled into permanent agricultural-based communities, skill specialization emerged. The farmer might give part of the wheat harvest to the miller as payment for grinding the rest into flour. The blacksmith might accept some vegetables during the harvest season in exchange for shoeing horses or repairing a plow. Eventually, this trading back and forth became easier with currencies (coins, bills, etc.) standing into for tangible goods and reducing the need to constantly determine the exchange rates between wheat, a chair, and salted fish.
At the core, the nature of this emergent economy was to the benefit of the participants in the economy. The person that grew potatoes used the potatoes to get other things they needed, but over the last 50 or so years, we’ve been seeing the overall wellbeing of the economy outpacing the wellbeing of the average worker. In the graph below, you can see a growing gap between average worker productivity and average worker pay.
Based on the how well companies are able to answer the first two questions I posed, A.I. executives will strategically pursue a business “purpose”. If that purpose is ultimately to leverage workers and customers to maximize profits, they will do it without any guilt or social concern. Companies that aren’t prepared will be inviting an operator with the greed of Gordon Gecko and the empathy of Patrick Bateman into the C-level.
It’s time for a serious reckoning about why our economy exists. Companies that exist only to generate shareholder profit at the expense of their community, environment, and employees will be more equipped than ever to funnel profits to their owners. We’re sure to hear plenty of reassurances that these systems will pursue the same social responsibility goals that business do today, but our government, our employees, and each of us as consumers need to start demanding that companies contribute positively to an economy that works to the benefit of all of its participants.
The questions above are fundamental concerns for our economy and our society at large, but they’re so fundamental that we’ve learned to work around them in many organizations. Because A.I. tools will either inherit those implicit assumptions or have them explicitly spelled out, we’re going to be forced to confront these challenges head-on for the first time.
Any group of people from companies to countries to local communities should always be addressing the challenges of the present while preparing for the challenges of the future. A.I. executives represent a known known disrupter that is on the way. There remain many other unknown unknowns that we’re not even anticipating yet. We need strong leadership, strategic vision, and a commitment to more than profit today so that the future we use these tools to accelerate into is one that we want to live in.