AI regulation overshoots, destroying productive applications necessary to prevent existential catastrophes

Although scientists have calculated the significant positive welfare effects of Artificial Intelligence (AI), fear mongering continues to hinder AI development. If regulations in this sector stifle our active imagination, we risk wasting the true potential of AIs dynamic efficiencies. Not only would Schumpeter dislike us for spoiling creative destruction, but the AI thinkers of the future would also rightfully see our efforts as the ‘dark age’ of human advancement. This article provides a brief philosophical introduction to artificial intelligence; categorizes artificial intelligence to shed light on what we have and know now and what we might expect from the prospective developments; reflects thoughts of worldwide famous thinkers to broaden our horizons; provides information on the attempts to regulate artificial intelligence from a legal perspective; and discusses how the legal approach needs to be to ensure the balance between artificial intelligence development and human control over them, and to ensure friendly artificial intelligence.

Our technology, our machines, is part of our humanity. We created them to extend ourselves, and that is what is unique about human beings. – Ray Kurzweil1

1. Introduction

The Chinese cardboard game “Go” is one of the most complex strategy games humankind invented. Go was considered so important, there are myths indicating that ancient kings played Go between their armies in the battlefield to resolve the conflict in peace. Computers prevailed against humanities best in many zero-sum, perfect-information, partisan, deterministic strategy games2 before, with the exception of Go, which was something to be proud of.

The strategy aspect of Go is very complex and emphasizes the importance of balance on multiple levels and has internal tensions. A game of Go cannot be won by using brute force: calculating every possible move, similar to what IBM®’s then state of the art AI, Deep Blue® used to win over Gary Kasparov. To manoeuvre through the countless possible moves on the Go board and chose the most efficient path, one requires capabilities beyond the conventional computing powers; capabilities only our minds have (or so we thought), such as extremely accurate image and pattern recognition and insight, all of which we thought granted us superiority over the artificial minds we created.

In October 2015, a software called “AlphaGo®” became the first computer to beat a professional human Go player in an un-handicapped game of Go (Silver and Hassabis, 2016). AlphaGo’s victory is probably one of the most significant demonstrations of the capabilities of an AI. Firstly, it shows that AIs are beginning to surpass us at things where success is dependent on strategy as well as calculation. Things we classify as a “game”, from stock exchange to conflicts, from contract negotiations to hostage situations. Second, AlphaGo developed strategies on its own, through playing millions of games against itself. These feats sent the chills down the spines of those who fear that AIs will overpower us in the future.

We humans accelerate the future with our minds. This is a strength and a weakness. Often, our predictions of the future are highly inaccurate. Based on predictions from a book called ‘The World in 2010’, published in 1976, we should have been living above and below the surfaces of three planets as of five years ago. Predictions regarding the future of AI are equally likely to be off base.

To avoid premature regulation over AI, we should be studying and searching for the meaningful point in time when a broader anxiety about AI becomes a genuine concern. The study of a point of ripeness, a ‘threshold ability test,’ asks when AI could really bring about concrete disadvantages that might counter-balance the demonstrated contribution to economic efficiency and welfare.

In the absence of such an objective benchmark marking the point in time when AI becomes a competitor with the human mind, regulators could easily jump the gun in regulating AI, which would lead to irreparable harm in total welfare of human societies.

Most of what we consider AI today is really our own intelligence re-formatted and re-cycled, with the help of computers lacking any skill of learning or consciousness of being. Regulation at this stage would be perverse. The economic efficiency potentials of AI should be set entirely free at this point in time, allowing us to actively and aggressively research appropriate goals for them which would not result in the extinction of humankind.

If you think our future robot overlords will one day thank us for ignoring the risks and under regulating, think again. On the one hand, any issues we may face from AIs will likely result from humanity failure to effectively direct AIs to our needs, not because we switched to a defensive AI regulation regime too early. On the other hand, at some point of time in the not too distant future, natural, human-related or external factors may threaten the fate of the Earth, and we may need AI to save the planet and us. One hopes that society has not pulled the hand brakes on the wheels of AI too early, fearing our own active imagination.