AI's Promise of Abundance Comes With Hidden Costs
The vision of AI-driven abundance suggests limitless free goods via centralized systems. However, those who own the energy sources and AI infrastructure will determine how resources are allocated and freedoms exercised.
Opinion by: Merav Ozair, PhD, blockchain and AI senior advisor.
The notion that "everything will be free" has champions in Elon Musk and Peter Diamandis. Their belief centers on AI abundance eliminating poverty and delivering universal high income.
Within the broader mega tech landscape, others echo similar sentiments about impending abundance. Demis Hassabis, for instance, suggests AI has the potential to ignite a "renaissance" characterized by "radical abundance."
At the World Economic Forum 2026 in Davos, politicians embraced Musk's vision enthusiastically. The prospect of their economic challenges being "set free" delighted them. This narrative holds obvious appeal. After all, who would turn down free goods and services?
But what is the actual meaning behind these claims? Will all economic transactions carry zero cost? Will corporations transform into purely altruistic entities abandoning profit motives?
Let's examine this narrative more closely.
Production costs may plummet, but zero cost remains impossible
Let's establish context. During an era of AI-driven abundance, goods and services won't materialize from "thin air." Production will continue to require labor, raw materials, energy resources and supporting infrastructure.
Technological advances in AI and related emerging fields could drive energy costs remarkably low while enabling highly automated manufacturing processes. These developments will push the marginal cost of both digital and physical products toward zero.
Three primary drivers explain this phenomenon. Labor automation represents the first factor, with machines and AI systems managing nearly all production processes, logistics operations and numerous service functions. Advanced manufacturing techniques and AI-powered distribution systems constitute the second factor, including technologies like 3D printing, robotics and AI logistics platforms that dramatically minimize waste and inventory requirements, making universal provision technically achievable. The third factor is energy abundance — whether through fusion power or extremely inexpensive solar energy that becomes so affordable it ceases to constrain growth.
Since energy fundamentally underpins all physical processes, reducing its cost causes all other production expenses to decline proportionally.
Strategic initiatives are already underway. Elon Musk has shifted his focus toward lunar manufacturing and AI development, targeting more than 1,000 gigawatts of solar power generation. Choosing solar over nuclear power aims to drive energy costs near zero. The challenge: establishing the necessary infrastructure on the moon demands enormous initial investment and must overcome substantial technical obstacles.
From a physics and engineering perspective, when genuine constraints — energy availability and automation capability — become plentiful, costs drop precipitously, though they never reach absolute zero.
The infrastructure layer remains conspicuously absent from discussions
For robotics and energy systems to generate "abundance" for all at scale and speed requires infrastructure.
Robotic systems and automation operate on what Jensen Haung terms "AI factories." This represents AI infrastructure, signaling a transformation toward approaching AI development as an industrial manufacturing process, allowing organizations to perpetually train and optimize AI models for enhanced safety and operational efficiency.
These facilities are specialized, high-performance computing data centers engineered to "manufacture" intelligence by transforming raw data into trained AI models and tokens, moving beyond simple data storage. Equipped with advanced GPUs and vast interconnected infrastructure, they power AI applications including autonomous vehicles, robotics systems and generative AI platforms.
Building and operating AI factories carries substantial costs. They require significant capital investment for construction and ongoing operations. Organizations that have already established this infrastructure will maintain their growth trajectory and continue advancing. Nvidia, for example, achieves five times the profitability of IBM during the 1980s while employing only one-tenth the workforce. Efficiency improvements from AI will drive productivity and profits higher. Capital will flow toward those controlling AI models, platforms and particularly the underlying infrastructure.
This dynamic will create unprecedented wealth concentration in human history.
Key market participants include technology leaders like Nvidia, AWS and SpaceX. These entities will maintain market dominance, creating formidable barriers for new entrants attempting to compete.
Government entities are participating as well. China leverages its extensive solar energy infrastructure to fuel the energy-intensive AI expansion. This generates a distinctive "AI and energy" ecosystem where artificial intelligence optimizes renewable energy production, while solar power sustains data center operations. China positions itself as a frontrunner in renewable energy deployment.
Low-cost energy still carries a price tag
Energy serves as the fuel powering AI factories, which function as the engine for all robotics, automation and AI applications generating abundance. Energy drives the infrastructure, while infrastructure enables AI applications. Consequently, energy represents the actual constraint. Without affordable energy, the "free" hypothesis collapses.
Presently, electricity constitutes the primary energy form powering this infrastructure. China pursues aggressive integration of renewable energy into its infrastructure networks, and other regions are similarly expanding renewable-powered energy into data center operations. However, electricity generation and grid infrastructure capable of supporting AI-scale operations remains extremely costly and lacks scalability. Achieving abundance at scale demands energy that is both extraordinarily cheap and scalable.
What alternatives exist?
Fission energy represents one form of nuclear energy. This mature technology delivers stable power generation but creates radioactive waste products. It introduces nuclear proliferation risks and safety concerns surrounding potential meltdowns. While cheaper than current fossil fuel-based electricity sources, it still involves tangible costs and, like other electricity sources, faces limitations preventing true scalability.
Fusion energy operates by fusing light atomic nuclei to generate power, replicating solar processes, whereas conventional nuclear energy splits heavy atoms. Fusion promises virtually limitless, cleaner energy without producing long-lived high-level radioactive waste.
Fusion technology is inherently safer, eliminating the risk of runaway chain reactions.
The limitation, however, is that fission remains the current operational technology. Developing nuclear fusion for energy generation requires extraordinary expense, demanding upfront investments totaling hundreds of billions of dollars, and remains experimental with commercial viability likely decades away from large-scale implementation.
Unlike nuclear fission, nuclear fusion offers scalability. While inexpensive relative to output, it doesn't achieve zero cost. Someone must finance the upfront infrastructure construction, initial development and ongoing maintenance expenses.
Musk's lunar ambitions
Solar power generation on the lunar surface offers abundant energy free from atmospheric interference. However, it entails substantial costs for launch operations, construction activities and maintenance in vacuum conditions. Musk's strategy involves relocating all production operations, including AI factory infrastructure, to the moon.
The lunar environment offers low gravity and abundant resource availability, positioning it as the most cost-effective location for AI infrastructure development.
Robotic systems will perform terraforming and infrastructure construction. Human personnel will arrive to supervise operations and drive expansion, while AI data centers will power the emerging space economy.
Through Starlink, SpaceX, Optimus robots and xAI, Musk possesses considerable advantages for executing this vision.
Nevertheless, machinery for manufacturing advanced AI chips must be transported to lunar facilities. These bus-sized machines demand extremely precise environmental conditions.
The proposed solution involves a novel approach called Atomically Precise Manufacturing (APM). This technology constructs materials atom by atom and corresponds with Musk's "first principle" philosophical approach.
If achieved successfully, this breakthrough could unlock unlimited solar energy and mineral resources from the moon and asteroid belt. Operations would face no thermal constraints or atmospheric interference.
This scenario could enable boundless AI capabilities at minimal cost. Industry experts suggest that successful lunar fabrication could generate a trillion-dollar opportunity, potentially reaching hundreds of trillions.
Who stands to capture the greatest value from this hundred-trillion-dollar prospect? Will distribution occur equitably?
The comfortable confinement of "free" offerings
Under centralized infrastructure and system architectures, infrastructure owners establish the rules of engagement. Highly centralized systems can deliver comprehensive "free" services, but frequently require substantial control over speech, mobility, data usage and economic decisions in return. Non-authoritarian welfare states might exchange some individual autonomy for security guarantees and assured services. Many contemporary "free" digital services derive funding from surveillance, user profiling and behavioral manipulation — your personal data and attention constitute the actual payment.
Within an AI abundance paradigm, infrastructure ownership may rest with governments. Corporations might own it. A public-private partnership arrangement could control it. Regardless of ownership structure, the infrastructure remains centralized, and this centralized authority will establish distribution terms — determining how AI abundance gets allocated, identifying recipients, and setting conditions. At their discretion, they can immediately "shut the valve" and halt distribution to individuals or groups. Your reliance on their services creates a "soft prison" that erodes your autonomy and self-sovereignty.
While representing a hundred-trillion-dollar opportunity, the centralized infrastructure owner will capture the majority share and determine what portion trickles down to the general population.
The saying goes that if something is "free", you become the product. This principle holds in a world of abundant resources. Within that paradigm, the product becomes your self-sovereignty.
Opinion by: Merav Ozair, PhD, blockchain and AI senior advisor.