Human brain and classical computer can generate true random numbers through algorithmic process

Arindam MitraarXiv 2005 · arXiv

Abstract

Traditionally, a sharp distinction has been drawn between Pseudo-Random Number Generators (PRNGs), which rely on deterministic mathematical formulas, and True Random Number Generators (TRNGs), which harvest entropy from physical phenomena. This paper challenges that dichotomy by proposing that true randomness can be achieved through specific algorithmic processes—selection, shuffling, and attachment—rather than arithmetic calculation. By shifting the focus from computation to extraction, this research suggests that both classical computers and the biological architecture of the human brain are capable of generating true random sequences.

Introduction

In the field of computational randomness, the prevailing wisdom is often summarized by John von Neumann’s famous observation: "Anyone who considers arithmetical methods of producing random digits is, of course, in a state of sin." This refers to the fact that arithmetic operations (addition, multiplication, etc.) are inherently deterministic; given the same seed, they will always produce the same output.

However, researcher Arindam Mitra argues that this "sin" only applies to mathematical operations. By utilizing algorithmic processes based on position-based selection, it may be possible to bypass the limitations of PRNGs. This paper explores a theoretical framework where classical hardware and biological systems function as true random number generators by acting as sophisticated filters for existing entropy.

Methodology: The Shift from Calculation to Selection

The core innovation of this research lies in three tiers of algorithms that prioritize the arrangement and selection of bits over their numerical calculation.

Algorithm-I: The Foundation

The basic model utilizes two strings of random bits. String A serves as a positional map, determining where 1s and 0s should be placed, while String B provides the actual values to be inserted at those positions. By using one random sequence to index another, the algorithm creates a new sequence that maintains a strict 50/50 probability for any given bit, effectively decoupling the output from a simple mathematical formula.

Algorithm-II: Iterative Feedback

To generate arbitrary amounts of data, the author introduces a "control sequence" ($S^c$). This sequence dictates the choice between a primary random sequence and its complement (where all bits are inverted). This iterative feedback loop allows the system to generate longer strings while maintaining the statistical properties of the input entropy.

The Generalized Algorithm

This model scales binary logic to $N$ equally probable numbers. By utilizing $N!$ (N-factorial) possible permutations and complementary sequences, the generalized algorithm ensures that the output remains statistically unpredictable. In this framework, the algorithm acts as a shuffler of entropy rather than a generator of "synthetic" randomness.

Key Insights and Theoretical Implications

The Brain as Biological Hardware

A significant conjecture of this research is that the human brain meets the biological demand for randomness not through complex mathematical computation, but through these simple selection algorithms. This frames the brain as a "living random number generator," capable of probabilistic behavior by performing selection-based logic on internal neural noise.

Meaningfully Meaningless Correlation

The paper introduces the concept of a correlation that is "meaningless" to an external observer but "meaningful" to the researcher. This allows for reproducible randomness—a critical requirement for Monte-Carlo simulations and debugging. If the input strings are known, the path can be repeated; if they are hidden, the output is indistinguishable from true physical randomness.

Redefining Classical Computers

The existence of these algorithms suggests that classical computers should not be viewed strictly as deterministic machines. Instead, they are systems capable of true probabilistic behavior if programmed to "select" and "extract" rather than "calculate."

Results and Statistical Validity

The theoretical framework is supported by several key findings:

  • Probabilistic Proof: Mathematical proofs demonstrate that if the input strings are random, the probability of the algorithm generating a 0 or a 1 remains exactly $1/2$.
  • Sequence Diversity: For a $2n$-bit output, the algorithm can generate $2^{2n}$ different sequences. The probability of any specific sequence occurring is $1/2^{2n}$, which satisfies the fundamental requirements for true randomness.
  • Cryptographic Potential: Because it is mathematically impossible to uniquely identify the specific pair of input sequences used to create a given output, these algorithms could theoretically generate unbreakable cryptographic keys, PINs, and passwords.

Limitations and Future Work

Despite the promise of algorithmic randomness, several bottlenecks remain. The algorithm is not a creator of randomness from a vacuum; it requires initial random bits (seeds) to function. While it can increase the "gain" (producing more bits than it consumes), it still relies on an external entropy source to initiate the process.

Furthermore, there is a theoretical risk of the algorithm halting or entering a cycle if an output string becomes identical to an input string, though the probability of this is statistically low ($1/2^{2n}$). Future research will focus on empirical validation, subjecting these algorithms to standard statistical batteries such as the NIST SP 800-22 or DIEHARD tests to compare their performance against state-of-the-art physical TRNGs.

Summary updated

BibTeX
@misc{mitra2005human,
  author = {Arindam Mitra},
  title = {{Human brain and classical computer can generate true random numbers through algorithmic process}},
  year = {2005},
  eprint = {cs/0505009},
  archivePrefix = {arXiv},
  url = {https://arxiv.org/abs/cs/0505009},
}
Figure