Navigating the future: AI, market evolution, and pension sustainability

Written By:

Andrea Isoni, PhD
Chief AI Officer, Director
AI Technologies

In this issue’s View from the Outside, Andrea Isoni charts the S&P 500’s evolution from industrial giants to tech titans – and explains why AI isn’t just reshaping markets

As I often find myself deep in historical data, analysing the ebb and flow of market cycles, I realised that for those managing capital with a horizon measured in decades – not quarters – the conversation needs to shift. We are standing at a unique intersection of history.

On one side, we have a technological revolution that is fundamentally altering how value is created. On the other, we face a demographic reality in the Western world that challenges the very maths of our social contracts.

In this piece, I’m attempting to synthesise two massive trends: the evolution of the companies that dominate our economy, and the role of Artificial Intelligence in sustaining the systems we rely on, specifically pensions.

Given the historical data of the S&P 500, and what I know about the tech industry today, I want to attempt to infer what the landscape will look like in 2045.

Of course, what follows is purely speculative thought, possibly grounded in reasoning, yet not to be considered as investment advice.

The rearview mirror: what drove the market?

To understand where we are going, we must look at the “ghosts” of the indexes past. As you know, the S&P 500 is a basket of the most important public companies in the world at any given time. If I go back in time, the evolution is stark.

In 1965, the top sector was industrial. The giants were General Motors, Exxon Mobil, General Electric, IBM, and AT&T. It was a physical world.

Value was created by moving atoms – building cars, generating electricity, and refining oil. The “pricing premium” belonged to those who built the physical infrastructure of the post-war boom.

By 1985, the power had shifted towards energy. Exxon Mobil led the pack, joined by IBM, GE, Philip Morris, and General Motors. Cars, machines, and consumer industries were driving massive energy demand. If you controlled the fuel, you controlled the market.

Fast forward to 2005, and the cracks in the industrial dominance were showing. The top sector was technology (or a blend of it). Microsoft stood at the top, but the list was a mix: Exxon Mobil, GE, CitiGroup, and Procter & Gamble. We saw a hybrid era – hardware companies like Intel and Cisco were huge, but software services like Microsoft and Oracle were climbing.

Today, in 2026, the transition is complete. The top sector is undeniably technology. The top five – NVIDIA, Microsoft, Apple, Meta, and Amazon – tell a clear story. Even the hardware companies (NVIDIA, Apple) are deeply integrated with software ecosystems.

The insight: demand and commoditisation

Even from the few data points above, there are many insights that can be drawn. The most obvious is that “demand” drives the index. Just as demand for Nvidia chips is huge today, so was energy demand in 1965 and 1985.

But the second, perhaps more critical aspect for the long-term view, is the shift in what we value. From 1965 to date, demand for purely “physical” goods, relative to the economy, has decreased, while demand for services and software has exploded. Why? Because of the huge reduction in the cost of producing goods due to technological improvements.

The companies that reach the top are those that resist this commoditisation. They have a “pricing premium” due to pure demand or better products that are hard to replicate. In 1965, it was hard to replicate a GM factory. In 2026, it is hard to replicate the Nvidia software/hardware stack or the Microsoft ecosystem.

So, where will the demand and pricing premium be in 2045?

The S&P 500 of 2045: what is bubbling up?

Looking 20 years into the future is a dangerous game, but if we look at what is struggling at the bottom of the market today, or what problems are becoming unsolvable without new tech, we can find clues.

I can give a few ideas on where the “pricing premium” might hide in 2045.

1. The cybersecurity imperative

At the moment, companies and consumers alike use a variety of data and softwares that are not easy to integrate and are fragmented. We are building a digital world on fragmented foundations.

I see this trend continuing, which opens never-ending gaps for cyber attacks. Also, due to AI adoption, cybersecurity is the only sector which has certainty of profitability: if a company spends on AI solutions the results are not certain but the additional spending in cyber is.

As our reliance on digital infrastructure becomes absolute, the value of protecting it becomes infinite. Cybersecurity solutions are not easy to replicate.

The demand will only increase. It is very likely that a company providing a unified, AI-driven defence layer – perhaps utilising quantum encryption – will be a top constituent (hard to say now).

My prediction: The top sector in 2045 could well be Cybersecurity. Watch companies like Fortinet or Palantir, or their future successors.

2. The integration of hardware and software

We often talk about robots, drones, and autonomous cars. But machines of any type will require a high “degree of automation”. It will not be enough to just produce the metal chassis of a car, or the propeller of a drone, or even just the chip.

The companies leading the S&P 500 in 2045 will be those that have integrated software and hardware with easy customisation for clients. The barrier to entry won’t be the hardware (which gets commoditised) or the software alone, but the seamless marriage of the two.

My prediction: Companies like Tesla, BYD, or defence-tech firms like Anduril Industries could dominate this space.

3. Energy management (not just production)

Energy production will increase – AI data centres and electric transport demand it. But the issue isn’t just generating power; it is handling different sources (battery, solar, gas, nuclear) in an “efficient” way.

Solar panels and wind farms can get “somehow commoditised”. What is difficult to “commoditise” is the management and optimisation of the sources. Even more so if we start seeing exotic sources like Space Solar. It will emerge a company that can handle energy and optimise the grid.

My prediction: A grid-optimisation giant will emerge, though perhaps just outside the top five unless energy demand massively outpaces current projections.

4. Tech-Pharma and the aging demographic

The population is getting older. This is a certainty. We have seen companies like Ely Lilly and Novo Nordisk being successful developing new GLP drugs, with AI tech coming into this space to accelerate development.

A company that uses AI to dramatically shorten the drug discovery loop, treating aging not just as a decline but as a manageable condition, will have a massive pricing premium.

My prediction: Tech-enabled pharma giants will be staples of the top five.

5. The standardised software ecosystem

Just as we have integration issues on the different software data we use today, we will eventually need a “standardised” software platform to rely on to avoid over-fragmentation. Such a software must be AI enabled, have automation, and scalability.

My prediction: Microsoft or a similar overarching ecosystem provider will remain relevant by becoming the operating system of the AI economy.

The macro view: why AI is the saviour of the pension system

While we speculate on which companies will win, there is a broader economic context that is vital for anyone looking at long-term liabilities.

It is not a secret that the birth rate is decreasing or flat at best in many Western countries. If there are fewer than 2.1 kids per couple, in practice, it means the population is decreasing.

This creates a maths problem for the pension system. Roughly speaking, the pensioners of today are paid by the taxes of the current working population. If the population is decreasing, less money is available for pensions overall.

We are already seeing the friction this causes: France recently raised the pension’s age (and French people were not happy). Italy lowered pensions on certain categories (Italians had just 370,000 born last year). This is where the technological boom intersects with macroeconomic survival.

Now there is AI technology to boost productivity. The infamous ChatGPT and the models that follow are claimed to raise US GDP by as much as two to seven percent over 10 years, according to sources from JP Morgan, McKinsey, and Goldman Sachs.

Let’s look at the maths again. Let’s assume that AI technologies overall will boost productivity by 10%. That means, if without AI software 10 working people are needed to sustain the pension of one person, with AI software boosting output, we might need just nine people working to pay for that same pension.

Effectively, a “robot” (or a software agent) is paying for a real person’s pension, acting economically like a “newborn” worker.

When we look toward 2045, we are looking at a world where the S&P 500 is dominated by companies that solve the problems of fragmentation (cybersecurity), integration (robotics), and demographics (healthcare). But more importantly, the very technology that drives these companies – AI – is the mechanism that might save the solvency of the system itself.

For the long-term observer, the opportunity lies not just in picking the next NVIDIA, but in understanding that the “pricing premium” of the future will come from solving the complexities of an automated, aging world.