There is plenty of time at the bottom: the economics, risk and ethics of time compression
Author:
- Anders SandbergASAnders Sandberg
- 11 March 2019
- journal article
- research article
- Published by Emerald Publishing in Foresight
- Vol. 21 (1) , 84-99
Abstract
Purpose: The speed of computing and other automated processes plays an important role in how the world functions by causing “time compression”. This paper aims to review reasons to believe computation will continue to become faster in the future, the economic consequences of speedups and how these affect risk, ethics and governance. Design/methodology/approach: A brief review of science and trends followed by an analysis of consequences. Findings: Current computation is far from the physical limits in terms of processing speed. Algorithmic improvements may be equally powerful but cannot easily be predicted or bounded. Communication and sensing is already at the physical speed limits, although improvements in bandwidth will likely be significant. The value in these speedups lies in productivity gains, timeliness, early arrival of results and cybernetic feedback shifts. However, time compression can lead to loss of control owing to inability to track fast change, emergent or systemic risk and asynchrony. Speedups can also exacerbate inequalities between different agents and reduce safety if there are competitive pressures. Fast decisions are potentially not better decisions, as they may be made on little data. Social implications: The impact on society and the challenge to governance are likely to be profound, requiring adapting new methods for managing fast-moving and technological risks. Originality/value: The speed with which events happen is an important aspect of foresight, not just as a subject of prediction or analysis, but also as a driver of the kinds of dynamics that are possible.Keywords
This publication has 33 references indexed in Scilit:
- Exponential laws of computing growthCommunications of the ACM, 2016
- The Unilateralist’s Curse and the Case for a Principle of ConformitySocial Epistemology, 2016
- The Threat of Algocracy: Reality, Resistance and AccommodationPhilosophy & Technology, 2016
- Racing to the precipice: a model of artificial intelligence developmentAI & SOCIETY, 2015
- Tipping points in open systems: bifurcation, noise-induced and rate-dependent examples in the climate systemPhilosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences, 2012
- Singularity Hypotheses2012
- Ultimate fast optical switching of a planar microcavity in the telecom wavelength rangeApplied Physics Letters, 2011
- Does Algorithmic Trading Improve Liquidity?The Journal of Finance, 2011
- Universal upper bound on the entropy-to-energy ratio for bounded systemsPhysical Review D, 1981
- Validity of the single processor approach to achieving large scale computing capabilities1967