In 1997, IBM’s Deep Blue defeated Russian chess grandmaster Garry Kasparov at chess. The world panicked about machines replacing human expertise. What happened next? Kasparov pioneered ‘advanced chess’, where human-artificial intelligence (AI) teams consistently outperformed both humans and machines working alone. The winning advantage wasn’t deeper chess knowledge. It was knowing when to trust the algorithm and when to override it.
Approximately 27 years later, we’re witnessing the same panic. Except this time, the verticals are falling faster.
The vertical illusion
For decades, the professional playbook was simple: go deep, get certified and build a moat. Tax accountants spent years mastering arcane regulations. Junior designers obsessed over kerning and colour theory. Strategists memorised frameworks. The deeper your trench, the safer your career.
That world ended on November 30, 2022.
When ChatGPT launched, it didn’t just autocomplete emails; it collapsed the value of isolated expertise. Today, a good prompt gets you tax optimisation strategies, brand positioning frameworks and Python code in seconds.
And it’s getting better. In December 2024, OpenAI launched ChatGPT Health. Andrew Beam, an assistant professor in the Department of Epidemiology at the Harvard T.H. Chan School of Public Health, noted that “the system can now interpret complex medical imaging and laboratory results at a level approaching specialist physicians”.
Deep, narrow knowledge used to be expensive. Now it’s a subscription fee.
The generalist advantage
Author David Epstein’s Range reveals something counterintuitive: breadth beats depth in unpredictable environments. He profiles everything from Renaissance polymaths to modern athletes, showing that late specialisers with diverse experiences consistently outperform early specialists when problems get complex and ambiguous.
Consider American businesswoman and writer Frances Hesselbein, who transformed the Girl Scouts not because she was a nonprofit expert, but because she brought thinking from business, military strategy and community organising. Her advantage wasn’t depth; it was the ability to connect patterns across domains.
This matters now more than ever because AI hasn’t eliminated the need for human judgment. It has eliminated the need for human information retrieval. What remains is sense-making across contexts – exactly where specialists struggle and generalists thrive.
T-shaped talents, recalibrated
The T-shaped model still works, but the proportions have flipped. Ten years ago, you needed 80 per cent depth, 20 per cent breadth. Today, it’s reversed.
The vertical bar of your T, your specialism is table stakes, delivered efficiently by large language models (LLMs). The horizontal bar, your ability to connect, contextualise and translate across domains is where human value concentrates.
I watched this play out with a client brief last month. The task: position a financial services brand for Gen Z. The junior strategist delivered a thoroughly researched deck on digital banking trends. Solid vertical thinking. But the breakthrough came from someone who had worked in gaming, understood Discord culture, and saw the connection between clan dynamics and financial trust. That synthesis and that horizontal leap, is what LLMs can’t yet replicate.
The most valuable people in any organisation aren’t the deepest experts. They’re the translators: between data and narrative, between technical capability and customer need, and between what’s possible and what matters.
The paradox of progress
Here’s what’s strange: we’re not eliminating specialisms. We’re democratising them. Legal research, medical diagnosis and code debugging remain critical. But they’re becoming utilities, not differentiators.
The question isn’t whether you can do tax accounting or competitive analysis; it’s whether you can spot when the tax strategy connects to the pricing strategy, when the competitive analysis should inform the org design.
OpenAI CEO Sam Altman said something revealing: “The jobs that will thrive are those requiring creativity, empathy, and cross-functional thinking.” Vertical depth is necessary but insufficient. Horizontal integration is the new scarcity.
The way forward
If you’re desperate to remain relevant – and the fact that you’re reading this suggests you might be – here’s your roadmap:
Build your vertical efficiently. Use AI to accelerate your domain expertise, not replace it. Spend what used to take ten years in three.
Then invest the surplus time horizontally. Learn adjacent fields. Work on projects outside your job description. Read outside your industry. The goal isn’t superficial dabbling, it’s building genuine understanding across multiple domains so you can spot connections others miss.
Because when the verticals become commodities, the horizontals become priceless.
Kasparov didn’t become irrelevant when machines learned chess. He became invaluable by learning to think with machines across multiple dimensions. That’s the pattern: depth gets you in the room, but breadth wins the game.
The question isn’t whether you’re being replaced. It’s whether you’re becoming more valuable by connecting what can’t yet be connected. In a world where every vertical is a commodity, the horizontal thinkers set the direction.
By Valli Lakshmanan, President, BPG UAE & BPG X.
