Jolting Technologies: Superexponential Acceleration in AI Capabilities and Implications for AGI

David Orban\orcidID{0009-0004-4954-1147 · 2025

Read Jolting Technologies: Superexponential Acceleration in AI Capabilities and Implications for AGI by David Orban\orcidID{0009-0004-4954-1147 (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2507-06398.txt.

Introduction

What if AI progress is not just exponential but accelerating in its acceleration? David Orban formalizes that jolt hypothesis in a solo 2025 paper and builds the tools to detect it if it ever shows up in the data.

His framework names the pattern precisely: growth whose third derivative stays positive, a superexponential climb. Monte Carlo simulations test whether the proposed detection methods can spot such a pattern reliably. The author is candid that long-run empirical data to settle the question do not yet exist. Drivers he examines include shrinking gaps between idea and action, and AI systems improving their own successors. The paper aims at readiness: measure the jolt early, and its governance implications can be faced with numbers.

Catalog record

CREATORDavid Orban\orcidID{0009-0004-4954-1147
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE27 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2507-06398

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Jolting Technologies: Superexponential Acceleration in AI Capabilities and Implications for AGI David Orban\orcidID{0009-0004-4954-1147 Abstract This paper investigates the Jolting Technologies Hypothesis , which posits superexponential growth (increasing acceleration, or a positive third derivative) in the development of AI capabilities. We develop a theoretical framework and validate detection methodologies through Monte Carlo simulations, while acknowledging that empirical validation awaits suitable longitudinal data. Our analysis focuses on creating robust tools for future empirical studies and exploring the potential implications should the hypothesis prove valid. The study examines how factors such as shrinking idea-to-action intervals and compounding iterative AI improvements drive this jolting pattern. By formalizing jolt dynamics and validating detection methods through simulation, this work provides the mathematical foundation necessary for understanding potential AI trajectories and their consequences for AGI emergence, offering insights for research and policy. 1 Introduction The trajectory of technological advancement has been widely studied. Although exponential models like Moore’s law have been useful, observations suggest technologies like AI may exhibit superexponential growth—a changing rate of acceleration, termed "jolt." This paper formally explores the Jolting Technologies Hypothesis , positing the third derivative of capability metrics C as a critical positive indicator of transformative change, crucial for a more accurate forecasting the emergence of Artificial General Intelligence (AGI). The motivation for this research stems from the profound societal and economic implications of accurately anticipating AI development. Misjudging AI’s pace could lead to significant unpreparedness. The mathematics of jolts, from basic calculus (velocity v=dC/dt , acceleration a=d^{2}C/dt^{2} , jolt j=d^{3}C/dt^{3} ), allows for a nuanced understanding beyond simple exponential curves. A sustained positive jolt implies an increasing acceleration, potentially outpacing exponential projections. Drawing on public academic literature, industry reports, and empirical and synthetic data, this paper investigates the theoretical basis and mathematical modeling of jolting technological progress; examines empirical evidence for or against this hypothesis in AI using benchmarks, milestones, and expert analyses; explores contributing factors such as compressed idea-to-action cycles, resource allocation, and AI research dynamics; and analyzes implications for AGI timelines, phase transitions, governance, and societal adaptation based on external research. Synthesizing external, verifiable information within a formal analytical framework, this paper robustly examines the Jolting Technologies Hypothesis, offering insights into AI’s future. The remainder of this paper is structured as follows: Section 2 details the theoretical framework of technological jolts. Section 3 presents empirical evidence supporting the hypothesis in AI, including key results from our Monte Carlo simulations. Section 4 provides a case study on AI agent capabilities. Section 5 discusses the broader implications for AGI, governance, and safety. Finally, Section 6 concludes with a summary of our findings and directions for future research. 2 Theoretical Framework The Jolting Technologies Hypothesis extends beyond qualitative observations of rapid progress, proposing a specific mathematical structure to describe and quantify superexponential growth in technological capabilities, particularly in AI. This framework centers on the concept of a "jolt," defined as the third derivative of a capability metric with respect to time, indicating a change in the measure of acceleration. 2.1 Formalizing Technological Jolts The term "jolt" used in this hypothesis is analogous to the concepts in classical mechanics, representing the rate of change of acceleration (the third derivative of position). While sharing this mathematical foundation, the Jolting Technologies Hypothesis (JTH) specifically applies this concept to the trajectory of technological capabilities. It is important to distinguish the JTH from broader concepts such as Kurzweil’s "Law of Accelerating Returns" [ 11 ] . For clarity, we define our terminology precisely: In this paper, ’superexponential growth’ specifically refers to growth characterized by a positive third derivative ( C^{\prime\prime\prime}(t)>0 ), rather than the colloquial usage of ’faster-than-exponential.’ This differs from standard exponential growth where C(t)=e^{kt} has constant relative growth rate. Our definition implies that the growth rate itself is accelerating, mathematically distinct from other forms of superexponential growth such as C(t)=t\cdot e^{kt} . This precise definition is critical for our quantitative framework and should not be confused with general descriptions of rapid growth. We acknowledge that technological development is not uniformly continuous. As Kuhn [ 10 ] demonstrated, scientific progress often proceeds through paradigm shifts, and Kondratiev [ 9 ] identified long-wave economic cycles that affect technology adoption. Our framework does not assume smooth, uninterrupted acceleration. Instead, the detection of jolts may coincide with these discontinuous transitions. The mathematical formalism allows us to identify and quantify these periods of rapid change within the broader context of uneven technological development. Let C(t) represent an aggregated measure of technological capability at time t . The first derivative, C^{\prime}(t) , signifies the rate of improvement (velocity), and the second derivative, C^{\prime\prime}(t) , represents the acceleration of this improvement. Standard exponential growth, while showing positive C^{\prime}(t) and C^{\prime\prime}(t) , typically assumes a statistically constant acceleration rate or a constant growth exponent. The Jolting Technologies Hypothesis focuses on the third derivative, C^{\prime\prime\prime}(t) , which captures the rate of change of acceleration itself. A sustained positive third derivative signifies that the system is not only accelerating but is becoming increasingly effective in accelerating its progress. To provide a standardized and comparable measure across different domains and scales, the jolt magnitude , J(t) , is formally defined as: J(t)=\frac{1}{C(t)}\frac{d^{3}C}{dt^{3}} (1) This definition normalizes the third derivative by the current capability level C(t) , ensuring dimensional homogeneity and allowing for more meaningful comparisons of jolt intensity across diverse technological trajectories. A sustained J(t)>0 provides quantitative evidence for a system operating in a jolting regime, where its capacity to improve is itself improving at an accelerating pace. In addition to this, a dimensionless jolt magnitude, J_{N}(t) , can also be defined to facilitate comparisons, particularly when the scales of C(t) , C^{\prime}(t) , and C^{\prime\prime}(t) vary significantly: J_{N}(t)=\frac{C^{\prime\prime\prime}(t)\cdot C(t)}{C^{\prime}(t)\cdot C^{\prime\prime}(t)} (2) To illustrate conceptually, consider a technology in which the capability C(t) is measured by the number of problems solved per day. If C(t) increases from 10 to 20 (velocity C^{\prime}(t)>0 ), and this rate of increase itself grows (acceleration C^{\prime\prime}(t)>0 ), a positive jolt C^{\prime\prime\prime}(t)>0 would mean that the rate at which acceleration increases is positive . For example, not only is the technology getting faster, and not only is it getting faster at getting faster, but the rate at which it is getting faster at getting faster is itself increasing. This signifies a deeply embedded pattern of accelerating improvement, distinct from simple exponential growth, where acceleration might be constant or proportional to velocity. An exponential rate of change reflects a consistent pace of innovation, and a technological jolt implies an increasing rate of innovation , where the capacity to generate novel solutions and improvements accelerates. Industry-wide learning curves, which describe steady efficiency gains, continue to play a role; however, they alone may not fully account for the superexponential characteristics indicative of jolting magnitudes. 2.2 Jolt Dynamics and Shrinking Doubling Times The phenomenon of jolting is intrinsically linked to the observation of shrinking doubling times for capability metrics. In traditional exponential growth, the time required for a capability to double, \Delta t_{\text{double}} , remains relatively constant. This is because the instantaneous relative growth rate, \alpha(t)=C^{\prime}(t)/C(t) , is constant. The doubling time can be expressed as: \Delta t_{\text{double}}(t)\approx\frac{\ln(2)}{\alpha(t)} (3) It is important to note that while a positive jolt ( C^{\prime\prime\prime}(t)>0 ) indicates increasing acceleration,

Read the complete text