It seems like professionals across East Africa cannot make it through a single day without discussions mentioning artificial intelligence. AI has become nearly as ubiquitous in the workplace as the internet, email, M-Pesa, Word, or Excel.
Certainly, there exist exciting potential uses for AI including scientific breakthroughs in medical research curing diseases or protecting ecological diversity to name a few. But more realistically, we face severe global distortions to economic and social stability as AI generates more and more revenue for billionaires while regular people get further and further left behind.
Additionally, most of us in the academic world lament something deeper in the collective dramatic decline in motivation, self-efficacy, and rigorous thought commensurate with AI usage. We are in danger of a generation that will be unable to think critically, analyse, or innovate on their own.
So, it begs the question. What is the value of AI in workplace or academic competition when everyone is utilising it? After everyone adopts the same tools, who still thinks differently?
At the macro and firm-level, recent research from Boston Consulting Group (BCG) gives business leaders plenty of reason to rush toward AI. Of course, one must look skeptically at any consultant published research since it is often used to scare clients into hiring their firm.
But, in a global survey of 1,250 senior executives and AI decision makers, BCG placed only five percent of companies in its most advanced category. Another 35 percent had started scaling AI. The remaining 60 percent reported minimal gains in revenue and costs despite substantial investment.
The companies at the front reported impressive results because of AI investments. BCG found that its most advanced companies achieved five times the AI related revenue increases and three times the ‘cost reductions’ of other companies.
Now with AI cost reductions, even though explicit explanations to changes in the nature of work are not super clear, one obviously assumes that massive job losses account for most of the cost savings. But compared with laggards in the study, the top firms also recorded 1.7 times the revenue growth, 3.6 times the three-year shareholder return and 1.6 times the EBIT margin.
BCG seems to want business leaders to easily draw one main conclusion from these numbers in that they should move faster.
Yet a new article published by Ryan Trimberger, Haresh Vaishnav, and Eric Yuen in the Harvard Business Review raises an awkward problem with the above BCG advice. While their article was sponsored by AWS and 4MINDS, it asks what happens when every company gains access to AI.
The authors point out that a competitor can now reproduce an AI capability that another company spent 18 months and billions of shillings developing. Companies increasingly draw on the same foundation models of AI, cloud platforms, and pretrained systems.
Therefore, the resulting problem becomes an agentic convergence trap as organisations adopt similar technologies, they can end up developing similar capabilities.
So, let us assume that a Kenyan bank adopts a certain new AI model early. It uses AI to analyse customers, improve credit decisions, answer customer queries, identify fraud, prepare marketing campaigns and help managers make decisions. For a short while, the bank gains an advantage.
But then every other bank starts to do the same.
They may buy services from different vendors, but many of the underlying models draw upon enormous overlapping bodies of human knowledge that large language model Ais comb the internet to find.
Then managers begin asking AI similar questions in each and every bank. Marketing departments ask for campaign ideas while human resource teams ask it how to improve retention all while strategy teams ask which markets to enter as senior executives upload reports and ask for recommendations.
While each AI answer may differ slightly, the thinking can start converging toward a middle similar consensus.
Companies have traditionally gained advantage partly because human beings do not think alike. You can have real creativity, innovation, and directions. Two executives can study the same market and reach completely opposite and yet fascinating conclusions.
Similarly, one entrepreneur sees an opportunity that everyone else dismisses. A product manager pursues an idea that looks foolish according to historical data. A marketing director understands a peculiar customer behaviour that no spreadsheet captures. An employee challenges a practice that everyone else accepted for twenty years.
While some of those differences sometimes can produce terrible decisions, they also produce originality and unique value addition that can send a company into the stratosphere.
Artificial intelligence works by learning patterns from enormous quantities of existing material. Companies rightly value that capability because patterns help with forecasting, analysis, writing and decision support. But when millions of people increasingly turn to overlapping systems for advice, another possibility deserves attention.
AI may improve the average quality of thinking while reducing the distance between one organisation and another. So, do not be so quick in your firms to terminate every single analyst, strategist, and innovator just to replace them with generic AI. Competitive advantage requires some form of difference.
A company wins because it does something competitors cannot do, notices something competitors miss, understands customers differently, moves earlier or makes a judgement others reject.