Whose future is it anyway? Part 4: ACE — Asymmetric Corporate Expansion

In the previous posts, I looked at how AI is reshaping access and dependency.

Capability is unevenly distributed across countries, platforms are becoming control points within them, and what starts as leverage can easily become dependency.

What happens when capability, access, and dependency all concentrate in the same place?

The concentration effect

AI has a natural tendency to concentrate, not because of a single decision, but because of how the system works.

The organisations building the most capable models benefit from a reinforcing loop:

• More capability attracts more users
• More users generate more data
• More data improves models
• Better models attract more capital
• More capital enables more compute

This isn't new — we've seen similar dynamics with search, social media platforms, and cloud services.

AI operates at a deeper layer — it's not just a service sitting on the surface, it's becoming fundamental infrastructure.

ACE: Asymmetric Corporate Expansion

If MAD is Massively Accelerated Disruption, and MAP is Mutually Assured Prosperity, there is a third dynamic emerging, which I have named "ACE":

ACE — Asymmetric Corporate Expansion

Asymmetric, because capability is not evenly distributed, and the advantages compound unevenly.

Corporate, because the organisations best positioned to capture AI capability are companies, not states or communities.

Expansion, because once a lead is established, it can scale rapidly across markets, domains, and geographies.

We risk a small number of organisations benefiting disproportionately, while some people — or possibly many people — benefit very little (or not at all).

To put it more bluntly, we're setting ourselves up for a future where a handful of boardrooms dictate global productivity.

First-mover advantage, but faster

First-mover advantage has always existed, but AI changes the speed at which it compounds.

If an organisation establishes a meaningful lead in:

• Model capability
• Infrastructure
• Distribution
• Integration

Then that lead can expand rapidly — faster than competitors can realistically catch up, faster than regulators can respond, and faster than markets can rebalance.

This is where the comparison to earlier technology waves starts to break down.

The cycle is shorter, the feedback loop is tighter, and the gap widens more quickly.

From platform to system

We tend to think of companies as participants in a system, but AI companies are increasingly shaping the system itself.

They influence:

• What gets built
• What is possible
• What is permitted
• What is economically viable

That is a different level of power, and it is closer to infrastructure.

Infrastructure, once established, is hard to displace.

Alignment, incentives, and reality

It's tempting to assume that the organisations building these systems will act in the broader public interest.

Some explicitly position themselves that way, but even within that group there are meaningful differences.

OpenAI and Anthropic, for example, both talk about safety, alignment, and responsible development, yet they have taken different stances on military usage, government partnerships, and deployment constraints.

Those differences are not accidental or incidental — they reflect funding structures, governance models, strategic priorities, legislative leverage (or laxity), interpretations of risk, and, ultimately, incentives.

This isn't a criticism of any individual organisation, it's a reminder that:

These are companies operating within highly competitive economic and political systems.

They are not neutral actors, and they are not insulated from the pressures of capital, competition, or national interest.

That means the trajectory of AI is not just shaped by what is technically possible, but by what is economically, politically, and strategically advantageous.

Ownership of intelligence

There is a subtle shift happening.

In previous waves of technology, companies owned:

• Software
• Platforms
• Distribution

With AI, they increasingly own:

• Capability
• Reasoning systems
• Interfaces to knowledge and decision-making

That raises a different kind of question, not just:

Who owns the platform?

But:

Who owns the capability to think, decide, and act at scale?

That may sound abstract, but it isn't.

The market amplification effect

One of the less discussed aspects of this is financial.

AI does not just improve products. It can improve the ability to analyse markets, execute trades, optimise pricing, and identify opportunities.

If a small number of organisations have access to more capable systems, they may not just build better products, they may also allocate capital more effectively, capture value more quickly, and reinforce their advantage across multiple domains.

This is where expansion becomes self-reinforcing, not just technically, but economically.

This is not science fiction

It’s tempting to frame this in terms of science fiction.

Runaway systems. Autonomous intelligence. Cyberdyne with better branding.

Those kind of scenarios are interesting, but they are also a distraction from something more immediate:

We don't need AI to become sentient to create risk, we just need it to become concentrated and shaped by the wrong incentives.

AI systems are not neutral in practice — they are built, trained, and deployed by humans, both individually and collectively.

That means they reflect:

• The data they are trained on
• The objectives they are optimised for
• The constraints and policies imposed on them
• The commercial and strategic incentives of the organisations behind them

Even when the underlying models are broadly capable, the interfaces we interact with are not raw intelligence, they are curated, guided, and constrained.

In some cases, they are steered quite deliberately.

We’ve already seen glimpses of this in how different systems behave:

• What they are willing to answer
• How they frame responses
• What they prioritise or avoid
• What they permit or preclude
• Where and how they apply guardrails

This is observable reality

Occasionally, those hidden behaviours become visible through leaks or reverse engineering — revealing the system prompts, policies, and heuristics that sit between the user and the model.

That shouldn’t be surprising, in fact it should be expected, because these systems are products, and products are shaped by market incentives.

If the underlying incentives prioritise growth over accuracy, engagement over truth, profit over public good, or national interest over global stability and cooperation, then those priorities don’t just influence outcomes, they become embedded in the systems themselves.

All of this is already happening at a scale we have never experienced before, and moving faster than moderation, monitoring, legislation, or existing safety nets can adapt to.

Which brings us back to the earlier point — concentration of AI capability amplifies this.

If a small number of organisations control the most widely used AI systems, then their assumptions, biases, incentives, and intentions don’t just affect their own products — they begin to shape how information is generated, interpreted, and acted upon more broadly.

Not (necessarily) because of malice, but because of reach.

At that point it’s not "neutral" technology, it’s real-world influence embedded in software.

A familiar pattern, at a different scale

If this dynamic plays out, the end state may look familiar.

A small number of dominant organisations.

A large number of dependent participants.

Increasing asymmetry in power and value capture.

We've seen versions of this before, but what is different here is the layer at which it operates — not just software, not just platforms, but fundamental computational capability itself.

Real-world consequences

The risk is not that AI replaces humans wholesale, the risk is that it amplifies existing structures of power faster than we can adapt to them. That it accelerates concentration before we have meaningful ways to counterbalance it, and it turns advantage into dominance before alternatives can emerge.

Ultimately, the risk is not that AI develops its own agenda, it’s that it inherits ours and then scales it.

What comes next

In the final post, I’ll look at what it would actually take to move towards something closer to MAP.

If the current trajectory leads towards disruption, dependency, and concentration, then the question is not whether a better outcome is possible, it’s what would need to change to make it likely, and whether we are actually willing to make those changes and pay whatever the accompanying price may be.