The Myth of Exponential AI Growth vs. the Emerging Plateau

Written on 16 August 2025.

The Myth of Exponential AI Growth vs. the Emerging Plateau

Overview

For years, advocates of Artificial Superintelligence (ASI) argued that machine learning would follow a path of exponential or even geometric growth, rapidly surpassing human capability. Popularized by futurists like Ray Kurzweil and echoed in science fiction such as The Terminator, this narrative assumed that once AI reached a certain threshold, it would recursively improve itself without limit.

Recent developments, however, suggest the opposite trend: diminishing returns, technical bottlenecks, and user dissatisfaction with newer, more resource-intensive models.

Predictions of Runaway Growth

Supporters of the exponential model claimed:

• Each new generation of AI would be vastly more powerful than the last.
• Recursive self-improvement would produce a "singularity" where AI surpasses human control.
• Society would undergo rapid transformation as machines gained superhuman intelligence.

These views often relied on simplified analogies to Moore’s Law and underestimated the complexity of scaling intelligence.

Current Reality

Contrary to those predictions, current AI developments show signs of stagnation:

• Incremental improvements – GPT-5 is only modestly better than GPT-4 despite vastly greater compute.
• Plateau in benchmarks – Advances often come from fine-tuning or combining models, not fundamental leaps.
• Hardware and cost limits – Energy consumption, chip supply, and economics constrain scaling.
• User experience mismatch – Many users prefer smaller, more responsive models (e.g., GPT-4o over GPT-5).

These factors undermine the assumption of geometric or exponential improvement.

Implications

The discrepancy between hype and reality has major consequences:

• Discrediting ASI evangelists – Predictions of imminent god-like AI look increasingly implausible.
• Reframing AI as an industrial tool – Rather than a runaway intelligence explosion, AI appears more like a powerful but limited technology.
• Policy and safety debates – Concerns about “singularity risk” may give way to more practical issues such as energy use, economic disruption, and misuse in cyberwarfare.

Cultural Impact

The vision of unstoppable exponential AI growth remains influential in popular culture and speculative philosophy. Yet, as current evidence points toward a plateau, the contrast between fictional AI apocalypse and real-world technological limits grows sharper.

What was once seen as inevitable may ultimately prove to be a myth.

The Rhetoric of "Even AI Experts Believe"

A common phrase in discussions about Artificial Superintelligence (ASI) is that "even AI experts believe it will be reached." The rhetorical function of the word even is important: it suggests a reluctant admission, as if those closest to the field would normally be skeptical but have been forced by evidence to concede the point.

In reality, however, AI researchers and industry leaders often have strong incentives to portray ASI as inevitable:

• Attracting capital – Dramatic claims about ASI secure venture funding, research grants, and government contracts.
• Shaping reputation – Associating oneself with the "future of intelligence" enhances prestige within both academia and industry.
• Market positioning – Emphasizing inevitability creates a sense of urgency around adopting present-day AI tools, thereby boosting sales.

From this perspective, the phrase functions less as a reluctant confession and more as a promotional tactic. Far from being unexpected, proclamations about the certainty of ASI serve the business and cultural interests of those making them.

Critics argue that such language artificially strengthens the myth of runaway growth by presenting a financial narrative as if it were objective scientific consensus. The invocation of even experts thus reinforces public belief in exponential trajectories that current evidence does not support.

Narrow AI and the Mark of the Beast Infrastructure

The failure of exponential AI predictions does not mean that existing systems are harmless. On the contrary, narrow AI—the practical, task-specific algorithms already in wide use—are fully capable of supporting the type of global control system described in the Book of Revelation.

The essential requirements of a "that no man might buy or sell" (Revelation 13:17, KJV) framework are not superintelligence or self-awareness. Instead, they depend on:

• Identity verification – Biometric scans, digital IDs, and surveillance data linked to centralized databases.
• Transaction monitoring – Real-time tracking of financial activity through digital payment systems.
• Automated enforcement – AI-driven tools that can instantly block or approve purchases based on compliance with state or corporate mandates.
• Integrated surveillance – Camera networks, facial recognition, and predictive policing that maintain social control.

These elements already exist at scale. For example, China’s social credit system, combined with digital payments and pervasive surveillance, demonstrates how individuals can be restricted from travel, housing, education, or commerce based on government-determined criteria. While not yet global, it shows that the infrastructure of economic exclusion is technologically feasible today.

This means that neither Artificial General Intelligence (AGI) nor Artificial Superintelligence (ASI) is necessary for a Mark of the Beast-like system to emerge. The technologies already in place—narrow AI, biometric ID, digital currency, and surveillance networks—are sufficient to enforce economic control.

From a prophetic perspective, the danger is not a hypothetical future superintelligence, but the consolidation of existing systems into a unified structure of obedience, where participation in society is conditional upon compliance.

China as Case Study

A practical demonstration of how narrow AI can sustain a system of economic control is found in China today.

• Digital payments – Cash has largely been replaced by platforms such as WeChat Pay and Alipay, where every transaction is traceable.
• Facial recognition – Cameras linked to AI systems verify identity in public spaces, at train stations, and even for apartment access.
• Social credit enforcement – Citizens can be blacklisted for infractions ranging from debt default to political dissent, resulting in restricted access to banking, education, or employment.
• Travel bans – Millions have been barred from purchasing airline or high-speed rail tickets due to low social credit scores.

These measures illustrate how existing AI and digital infrastructure already enable a system in which individuals may be prevented from buying or selling unless they conform to state-approved standards. Such an arrangement parallels the prophetic description in Revelation 13:17 (KJV), demonstrating that Artificial General Intelligence or Artificial Superintelligence is not necessary for such a framework to exist.

AI Disclosure: Parts of this page may have been created, edited, or assisted by artificial intelligence tools (such as ChatGPT or other language models). All AI-assisted content is reviewed by a human before publication. For questions, contact the site administrator.