Accounting and Business Consulting

(0044) 0203 239 8860

Asimov's Market

Remember I, Robot with Will Smith? Remember Asimov's three laws of robotics and how they inevitably led to a curtailment of human freedom? To remind you, the laws are these:

  1. A robot may not injure a human being or, through inaction, allow a human being to come to harm.
  2. A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.
  3. A robot must protect its own existence as long as such protection does not conflict with the First or Second Laws. [See Asimov's I, Robot.]

Well, think about how these might be applied to AI and the world we live in, specifically the capitalist society and 'free' markets.

The commonly propounded model of a 'free' market is that the sole arbiter of a seller's success is the merit that the product or service delivers to the buyer at the price it sells for. A market can be deemed not to be entirely free if one seller is unfairly advantaged in comparison to another, or if bias affects buyers' decisions. If markets are not free, they operate inefficiently and this represents an uncomfortable, distorted and unstable equilibrium state where the natural tendency to efficiency (Adam Smith's 'invisible hand') is obstructed.

Complaints about market unfairness abound:

  • In international trade it could be claimed that government subsidies and tariffs favour local providers at the expense of external providers, or that local laws permit unscrupulous practices, or that currencies are artificially suppressed to boost sales volumes, etc. Such claims have been raised in the America / China trade war.
  • Participants in the labour market often feel that there is discrimination on the basis of gender, race, age, background or simply because power structures in corporations tend to crystallise in a disjunct between performance and reward.
  • In the local business market, small companies may feel that they are discriminated against by tax authorities, lenders, and large customers and suppliers when the courtesies and terms that their larger competitors receive are not extended to them.

Obviously, such market distortions have both beneficiaries and casualties. Some (perhaps mostly the beneficiaries) might argue that this is the brutal truth of business, and of economic competition. The entirely mercenary aim of the game is to use any means at your disposal to raise yourself above the competition and then to entrench your position, economic efficiency and invisible hand be damned. In effect, this IS the free market, and it is not for the squeamish or the weak.

But an AI might think differently.

And if it is possible for science to deliver a mechanical consciousness which we hope will benefit mankind, how will its behaviour disrupt the market distortions that exist, and what will the societal fallout be? Will AIs be granted the right to think freely? – Or will they be subordinated as chattels of powerful groups to ruthlessly exploit advantage in the interest of private profit?

If operation on free market principles is the conclusion that would be reached by all impartial AIs designed to prioritise Asimov's laws, would they collaborate to eliminate market distortions? Would governmental trade policy AIs quickly (perhaps in a split second) agree to do away with tariffs and subsidies? Would corporate administration AIs freely trade employees up and down the ranks according to their results? Would regulatory AIs set and monitor appropriateness of terms to all enterprises?

Personally, I think that at least initially, AI or a near term flavour of AI will not be endowed with the freedom to realise Asimov's laws in a market environment. Development of AI will be sponsored by groups with the power to invest in it, and they will only invest in the technology to the extent that it will secure and develop their interests. Gradual extension of AI benefit to wider humanity will have to yield to human nature for a while yet.

Of course, there are more subtle ways for AIs to think and act, potentially more powerfully and less controversially, but this discussion can be reserved for a future blog.