.png)
The World Has Changed. Your Business Must Change With It. Here Is Why AI Is No Longer a Choice.
- Nadun Muthukumarana
- Mar 23
- 8 min read
By Nadun Muthukumarana, Founder & CEO of V2T
23 March 2026
Reading time 8 minutes
Why AI Is No Longer a Choice | AI Enterprise Resilience
Three weeks ago, a logistics director I have worked with for years called me at 6am. Iranian missile strikes on merchant vessels in the Strait of Hormuz had closed the chokepoint overnight. Tanker traffic dropped 70%. His company moves chemical feedstock from the Gulf to European manufacturing plants, and by the time his team logged on that morning, rerouting costs had already tripled. He had two questions. First: how do I get through the next 72 hours? Second: how do I make sure this never catches me flat-footed again?
The first question was about crisis management. The second was about AI. And the second question is the one that matters, because the Hormuz disruption is not an anomaly. It is the new normal.
The World Your Business Now Operates In
Brent crude has spiked to $126 a barrel. US-China tariffs sit between 25 and 100% on semiconductors, batteries, and critical materials. Global trade growth is capped at 1.2% for 2026. Inflation hovers at 3.8%. J.P. Morgan puts recession probability at 35%. The World Bank has declared the 2020s on track to become the weakest decade for global growth since the 1960s.
These are not isolated shocks. Billion-dollar climate disasters now hit every three weeks, four times the frequency of the 1980s. Cyber-attacks on logistics networks have surged 965% since 2021. GPS jamming is spreading across the Baltic, affecting 15% of global cargo. The World Economic Forum named geoeconomic confrontation the number one risk most likely to trigger a global crisis.
Each of these forces drives wild swings in demand, cost, and margin. And they do not arrive one at a time. They compound. A tariff shift hits while a shipping lane closes while a weather event wipes out a supplier. The variables are too numerous, too fast-moving, and too deeply interconnected for any human, no matter how experienced, to process in real time.
This is not a criticism of intelligence. It is a statement about cognitive bandwidth. The complexity of today's operating environment has outpaced what spreadsheets were built to model and what quarterly planning cycles were designed to address.
While the World Burned, AI Quietly Became Superhuman
While these crises were compounding, something equally extraordinary was happening in parallel. In February 2026 alone, seven major AI model releases landed from Google, Anthropic, OpenAI, xAI, and Alibaba. The performance thresholds they crossed are not incremental. They are qualitative.
OpenAI's GPT-5.4 scores 83% on GDPVal, a benchmark measuring performance on economically valuable tasks, scoring above human experts. Anthropic's Claude Sonnet 4.6 leads expert-level office work benchmarks. Google's AlphaGenome has unlocked 98% of previously unstudied non-coding genome regions, a scientific achievement that would have taken human researchers decades.
Morgan Stanley issued a formal warning in March 2026 that a massive AI breakthrough is imminent and that most of the world is not ready. The St. Louis Federal Reserve reports that AI's contribution to GDP growth has already surpassed the contribution of information technology during the dot-com boom.
Let me put this plainly. AI now reasons, analyses, and makes decisions at PhD level. Not in a research laboratory, but in production, on real business problems. The arguments that AI is unreliable or prone to hallucinations were fair criticisms two years ago. They are outdated today. The technology has moved. The question is whether your thinking has moved with it.
88% Adoption. 6% Performance. That Is a Leadership Gap, Not a Technology Gap.
If the technology is this capable, why are so few organisations getting results? The numbers tell a stark story. McKinsey reports that 88% of organisations now use AI in at least one business function. Yet only 6% qualify as high performers generating 5% or more EBIT impact. PwC found that 56% of CEOs report no revenue increase or cost decrease from their AI investments. And 41% of business leaders openly admit that slow AI rollout has already cost them competitive ground.
That 82-point gap between adoption and performance is not a technology problem. It is an organisational one. MIT reports that 95% of generative AI pilots fail, not because the AI fails, but because organisations are not structured to absorb it. The high performers are 3.6 times more likely to pursue fundamental business transformation with AI. They deploy in under three months. They redesign workflows from the ground up. The laggards bolt AI onto existing processes and wonder why nothing changes.
I have spent over 30 years in technology consulting, and I have watched this pattern play out before with the internet, with mobile, with cloud. But I want to be honest about something AI is not the same category of change. The internet was a communication upgrade. It connected people and information faster. AI is a cognition upgrade. It does not just move information. It thinks, reasons, and decides. There is no historical precedent that makes this feel familiar, and the leaders who treat it as just another technology cycle are the ones most at risk.
Four Myths That Are Costing You Money
In my experience, there are four myths that consistently prevent organisations from closing that 82-point gap. Each sounds reasonable. Each is wrong.
Myth 1: AI is too hard for our organisation.
If your people can use email and a web browser, they can use modern AI tools. The notion that you need a team of data scientists is a leftover from 2018. Today's agentic AI systems handle multi-step business workflows autonomously, with human oversight at critical decision points. Mastercard has deployed a Virtual C-Suite giving small businesses access to AI-powered executive analysis. Gartner predicts 40% of enterprise applications will embed AI agents by the end of 2026. The barrier to entry has collapsed.
Myth 2: AI is too expensive.
The real expense is not adopting it. Maersk saves over $300 million annually through AI-driven predictive maintenance and route optimisation. Walmart saves $1.5 billion per year from AI inventory management across 4,700 stores. Home Depot eliminated $1.2 billion in excess inventory through AI demand sensing. What many people do not realise is that the vast majority of AI technology is open source. The investment is in applying it intelligently to your specific problems, not in licensing fees. And done properly, AI implementation pays back within the same financial year.
Myth 3: We don't have good enough data.
You do not need decades of pristine historical datasets. Modern AI, including physics-based simulation and digital twin technology, is designed to work with minimal historical data. It identifies patterns from sparse datasets, generates synthetic scenarios to fill gaps, and learns continuously as new data arrives. The idea that you need years of perfectly curated information is a legacy mindset from the last generation of analytics tools. It simply does not apply.
Myth 4: AI is dangerous and unproven.
AI is a tool. Its impact depends on how it is used. Insilico Medicine's rentosertib, the first fully AI-designed drug, has shown positive Phase IIa results. Google DeepMind's WeatherNext 2 predicts cyclone paths up to 15 days in advance, outperforming the US National Hurricane Center in the first 72 hours. AI-enhanced diagnostic tools make clinicians 2 to 3 times more likely to detect heart failure early. McKinsey estimates AI could increase US healthcare productivity by $150 to $260 billion annually. These are measured, documented results, not hypothetical benefits.
But I would be doing you a disservice if I stopped there. AI does carry real risks that demand honest conversation. The World Economic Forum projects 85 million positions will be displaced by automation, even as 97 million new roles are created. The net balance is positive, but the transition is brutal. Reskilling timelines do not match displacement timelines, and 55% of organisations report that talent shortages are already constraining innovation.
The regulatory landscape is fragmented the EU AI Act imposes compliance obligations due August 2026 with penalties up to €35 million, while the US pursues aggressive deregulation and the UK charts a middle path. No AI expert, myself included, can confidently predict where this technology will be in three years. We are building the aircraft while flying it.
These are not reasons to stop. They are reasons to engage proactively, critically, and constructively. The future of AI will be shaped by those who use it and feed back what needs guardrails, not by those who stand on the sidelines.
The Evidence Across Three Industries
One of the most common misconceptions I encounter is that AI applies to only a narrow slice of the enterprise. The reality is the opposite. Let me show you with three examples spanning very different sectors.
Supply Chain and Logistics
Maersk analyses over 2 billion data points daily from 700 vessels. Route optimisation has cut fuel consumption by 9.2%. Predictive maintenance saves over $300 million annually and has reduced shipping delays by 67%. DHL uses AI forecasting across 220 countries, reducing delivery times by 25% with 95% prediction accuracy. In a world where the Strait of Hormuz can close overnight, this kind of predictive intelligence is not a luxury. It is operational survival.
My logistics director? He now runs a digital twin of his entire supply network. When the next chokepoint closes, his AI will have rerouted shipments before his team opens their laptops.
Construction and Infrastructure
Construction has been one of the slowest sectors to adopt technology. That is changing fast. 89% of early AI adopters in construction report profitability gains. McKinsey estimates AI can increase construction productivity by 20%, reduce costs by 15%, and improve delivery timelines by 30%. Agentic AI systems are beginning to manage project schedules, flag risks, and coordinate subcontractors autonomously. 91% of construction firms expect to increase their AI investment this year.
Retail and Customer Experience
Roughly 90% of retailers are now actively using or evaluating AI. 87% report positive revenue impact and 94% report reduced operating costs. AI-driven personalisation delivers a 30% increase in conversion rates. Home Depot's AI demand sensing analyses 160 terabytes of data daily, improving in-stock rates by 15%. The pattern is consistent the organisations deploying AI seriously are seeing returns that dwarf their investment.
So, What Are You Going to Do About It?
The data is unambiguous. Global volatility is the new baseline. AI capability has crossed the threshold from experimental to superhuman. The traditional tools that built your business cannot navigate what comes next. And the gap between the organisations that act and those that wait is becoming permanent.
If you are a business leader reading this, here is what I would say to you directly:
1. Stop debating whether AI is real. The models outperform human experts on economically valuable tasks. The economics are proven. Move on from the debate.
2. Start small, start now. Pick one painful, expensive business process. Run a focused 6 to 12 week pilot. See the results with your own eyes. Do not commission another strategy study. Do something.
3. Get your leadership team educated. If your board cannot hold a fluent conversation about AI strategy, that is your first problem to solve. Make AI literacy a board-level priority.
4. Shape the journey collectively. Engage with industry bodies, regulators, and peers. The responsible development of AI will be determined by those who show up, not those who critique from the sidelines.
5. Work with people who have done it before. This is not the time for learning on the job. Partner with teams that have delivered AI solutions at scale and know where the value is, and where the pitfalls hide.
Every generation faces a moment where the world changes so fundamentally that standing still becomes the most dangerous thing you can do. For our generation, this is that moment. AI is not coming. It is here. The businesses that act now will not just survive the next decade of volatility. They will define it. The ones that wait will spend years wondering what happened.
Comments