Seventy years ago, IBM introduced the IBM 650 as its first mass-produced computing machine, selling 450 units. Last year, the global information-technology industry sold an estimated $5.7 trillion of goods and services. I believe this demonstrates that access to digital information is now as important for success as intellectual property or brand.
More than size, the key element of today’s IT is its incalculable and ever-increasing complexity. The growth and scale of connected systems create more relationships among people, objects, and lines of code. It’s a nonlinear phenomenon, explosive and essential.
I love it. Near-continual information sharing enables our digital lives, with benefits that are the envy of any other time before. I come from a small city in India, and for billions of people like me your ideas and your drive couldn’t reach many people before today. It’s how we’ve had a revolution in biosciences, how anyone can start a business anywhere, and how we’re able to predict and cope with natural disasters better than before. Data velocity is wonderful. Complexity is growth.
But I also hate it. In a recent survey from Rubrik Zero Labs, 90% of companies surveyed said they now manage hybrid cloud environments, such as in their own facilities, offsite, and up to five different cloud platforms. On top of the hassles, I believe cyberattacks have at least 10 different ways into a corporate system.
Cryptocurrencies have provided opportunities for cybercriminals to carry out ransomware attacks, and remote work has increased opportunities for identity-based hacks and social engineering security attacks. A shocking 86% of companies reported paying a ransom once they got an extortion demand, another finding from Rubrik Zero Labs. This complexity can place a huge tax on the economy.
Love it or hate it, though, digital complexity is not going away. People bear the cost to have the benefit. AI won’t simplify everything, nor will quantum computing. Complexity has to be managed in entirely new ways as systems grow ever more complex.
The best practitioners embrace the changing reality. “I stay in this business because it keeps changing,” says Steve Pugh, chief information security officer at Intercontinental Exchange, which among other things owns the New York Stock Exchange, futures and commodities markets, and risk management practices. Nowadays in security, he says, complexity shows itself in “a preponderance of automated attacks, more access to hacking tools, and an increasing speed of attacks.”
In response, he’s focused on core values for his marketplaces, like continuous uptime, and built a risk-oriented system of researchers who stay up on the newest hacking technologies and techniques, plus a “red team” of employees who deliberately try to break his security. Continuous protection means continuous learning.
How much complexity am I talking about here? Besides traditional IT, there are more smartphones than people, 92% of them on fast networks. Apple manages 2 million different apps, from tens of thousands of developers. There are billions more sensors, and AI models munging petabytes. There now are over 1,100 hyperscale data centers, each with anywhere from 5,000 to well over 1 million servers.
It’s going up from here. In current dollars, the Manhattan Project to create the first nuclear weapon cost $36 billion. This year Microsoft alone will spend twice that on AI. Google, Amazon, and Meta are all spending at that level, or higher. That means more data, more frameworks, more use cases, more velocity of information, more complexity.
I believe what is being built now isn’t just bigger. It’s different. Is the data about the color of a car being used for online sales, for an insurance report, for an influencer post, for a self-driving program, or as part of a much larger question? Once something that was mostly stored and occasionally fetched, data is now in near-continual motion, creating a new tax on networks. For analysis, it needs to be of the highest quality. If there’s a problem with an AI hallucination, it’s more likely to be a problem with data than a problem with code. Managing data complexity like this is entirely new.
There are similar challenges ahead for supposedly “simplifying” technologies, like generative AI that writes code. These will change workflows, and for a while at least will have to be checked by humans, or checked and tested by still unknown AI-based management programs—another level of complexity. Gen AI may likewise make it simple to create new images, but these will have to be shared, secured, and stored, taxing systems and networks.
Part of the solution, where security and much else are concerned, says Mr. Pugh, is to focus on some human things that don’t change. “We always manage how much risk we take,” he says. “We try to understand our environment and our adversaries. We keep people who are willing to prove us wrong, spotting our vulnerabilities. We have trusted allies.”
It’s a challenging future, but on balance, it’s healthy. We come up with new technologies that make things more complex for a time, but then they are simplified, allowing us to do even more complex things. Leaders must recognize that complexity isn’t the enemy. It’s the consequence of increasing opportunity in our interconnected, complex world. Great leaders will turn complexity from an obstacle to a catalyst that continues to enrich and improve quality of life.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.
This story was originally featured on Fortune.com
