3.0 Singularity Vision

The New Scarcity: Power, Infrastructure, and the Economics of the AI Age

One of the most misunderstood aspects of AI-driven abundance is the assumption that abundance eliminates scarcity altogether.

However, all these does not eliminate scarcity altogether. The economy begins shifting from one constrained primarily by cost, resourcefulness, biological limits and scarcity of human cognition toward one constrained by energy, compute, infrastructure, and governance - resources that are relatively cheap, scalable, optimizable, and potentially far more expandable than biological intelligence itself. AI changes the structure of scarcity by shifting economic value away from labor and toward ownership of automated productive systems i.e. intelligence itself and digital infrastructure that enables the intelligence - the new capital.

In the AI age, scarcity may increasingly concentrate around: - energy, - compute, - semiconductor supply chains, - robotics infrastructure, - data, - governance, - and ownership of intelligence infrastructure itself.

Which means AI may simultaneously: - democratize intelligence access and capability, - while also creating new concentrations of infrastructural power that powers the intelligence.

Because even if intelligence becomes abundant, the infrastructure behind intelligence may remain highly concentrated. Hence abundance discussed in previous chapters alone does not guarantee equitable outcomes.

The key question is:
Who owns the systems that produce intelligence?

If intelligence infrastructure, compute, energy systems, robotics, and AI platforms become heavily centralized, AI could also produce new concentrations of economic and institutional power.

Which means the central challenge of the AI age may not simply be:

“Can humanity create abundance?”

but:

“Can humanity distribute abundance broadly enough that intelligence becomes a universal civilizational resource rather than a centralized mechanism of control?”

That question may ultimately determine whether AI becomes the greatest force multiplier for human flourishing in history - or simply the next engine of concentrated power.

In previous eras:

• landowners controlled agricultural economies,
• industrialists controlled factories,
• and information platforms controlled digital networks.

In the AI era, power may increasingly accumulate around ownership of:

• compute infrastructure,
• energy systems,
• AI platforms,
• autonomous production systems,
• robotics ecosystems,
• and intelligence networks themselves.

Today majority of AI & AGI efforts are unfolding through centralized without democratic oversight, without public participation.

The real question is: - In a centralized, monolithic, proxied, undemocratic, and non-inclusive setting, who decides what AI becomes? - Who controls the changes AI brings to the world? - And for whose benefit?

The New Capital

One of the most important transitions of the AI age is that economic value increasingly shifts away from labor and toward ownership of automated productive systems.

Historically: - humans were the primary carriers of intelligence, - labor was the dominant productive force, - and capital amplified human capability.

In the AI age: - intelligence itself becomes capital, - autonomous systems become productive assets, - and cognition becomes embedded directly into all infrastructures and systems powering civilization therefore potentially leading to pervasive control

This creates a new form of capital: The intelligence infrastructure.

As AI proliferates, the owners of advanced AI systems may increasingly control all forms of skills, knowledge, digital labor, production, coordination, optimization, distribution, research, and decision-making itself.

That has enormous implications for inequitable abundance and wealth concentration.

If intelligence infrastructure becomes widely accessible, AI could dramatically expand economic participation and reduce historical inequalities. But if intelligence infrastructure becomes highly centralized, AI could instead amplify asymmetries at unprecedented scale.

The defining economic question of the 21st century may therefore become:

Who owns the intelligence infrastructure of civilization?

The economy therefore does not simply become “post-scarcity" unless the intelligence itself and key digital infrastructures - if not all - powering the intelligence becomes truly distributed, inclusive and open.

Some of these systems, however, can be intentionally democratised to prevent excessive concentration of power, ensure equitable distribution of abundance and to strengthen individual and collective agency, sovereignty, and resilience.

This may include: - open and interoperable AI models, AI frameworks, agentic systems, - distributed intelligence networks, - decentralized public intelligence infrastructures - user owned, community-owned intelligence and intelligence infrastructure, - open robotics ecosystems, - decentralized identity, data, networks, ownership, - and cooperative AI platforms.

The long-term shape of the AI era therefore depends not only on technological capability, but on how AI is produced, ownership, access, governance, and participation are structured.

If intelligence infrastructure becomes concentrated in the hands of a few actors, existing power asymmetries may deepen. But if critical layers of AI and digital infrastructure are made more open, distributed, and participatory, the AI era could expand human agency rather than diminish it.

In the AI age, economic leverage may increasingly shift toward those controlling the foundational layers of digital infrastructure rather than consumer-facing applications alone.

Historically, much of the digital economy concentrated around: platforms, interfaces, distribution, and user networks.

But as AI systems become deeply integrated into all sectors of civilization, the underlying infrastructure powering intelligence itself may become far more strategically important.

This includes: distributed compute infrastructure, inference infrastructure, semiconductor supply chains, AI networks, agentic webs, data hubs and energy infrastructure.

In particular, control over inference infrastructure, AI networks, agentic webs may become just as important as control over model training itself. Training frontier models may remain expensive and centralized, but inference infrastructure, AI networks, agentic webs, harnesses is what operationalizes intelligence continuously across civilization at scale. We hypothesise that the entities controlling large-scale inference networks, AI networks, agentic web, harnesses may increasingly control the operational layer of the AI economy itself.

Over time, economic power may therefore accumulate less around individual models or applications and more around the digital infrastructure layers capable of delivering intelligence ubiquitously across society.

AGI: A New Operating System of Power

• Today, Artificial Intelligence & tomorrow AGI will emerge as a power of powers, a force multiplier capable of influencing economies, governance, and societies at scale effortlessly. That makes AI more than a tool. It is becoming an operating system of power in itself.

• And with that comes a choice:

  • AI or AGI can intensify existing injustices - reinforcing bias, widening inequality, and further concentrating control in the hands of a few.
  • Or, it can be the greatest equaliser & help rebalance the system - redistribute access to knowledge, production, skill, opportunity and wealth, increasing agency, creating space for self-determined futures, create fairer systems, supporting more just, plural and inclusive outcomes.
• Which path it takes depends on how AI or AGI is designed, built, who it serves, and what values guide its design.

Reimagining Possibility

Even if the technology is neutral, the outcomes are not - the potential of AGI to help rebalance power doesn’t come from the technology alone. It depends entirely on: - how AI is designed, - who shapes it, - what idealogical, cultural, social, political, and philosophical norms guide it, - and how it’s governed and incentivized.

It’s not magical, but rather reasonable and optimistic to imagine that if well designed, AGI will play a fundamental role in addressing systemic problems in modern society like inequality, oppression, exploitation, corruption, and poverty within a foreseeable horizon.

In conclusion,

The design, ownership, access, distribution of AI is not technical question. That's a political choice.