Book Review: MACHINES THAT BEHAVE BADLY - The Morality of AI

By Toby Walsh

The celebrated author, Toby Walsh, asserts that we cannot build machines that match even the intelligence of a two year old child. Yet we can program computers to perform narrow, focused tasks that human beings would want to avoid. Oddly enough, the author adds, AI is already a part of our lives but is somewhat hidden from sight. Ultimately, AI will permeate all aspects of our daily lives just like electricity.

AI, like human intelligence which is a collection of different skills, is also a collection of different technologies such as machine learning, natural language processing and speech recognition. However, if we hype up our expectations of what AI can do or not do, we will simply fail because AI cannot surpass human intelligence.

It is impossible, Walsh asserts, to list all the narrow applications that AI is being used for but it is clear that AI makes a significant promise of transforming our society. The potential advantages of AI will cover almost all sectors of society including agriculture, banking, defence, education, entertainment, finance, government services, insurance, law, manufacturing mining, politics, retail and transportation.

However, AI presents significant risks such as displacement of jobs, increase in inequality, within and without countries, the transformation of war, the undermining of public discourse and the erosion of privacy and other human rights. Therefore, the unexpected consequences of AI would be the most devastating in human history.

Machines are now routinely directing humans mechanically, and controlling populations at scale, at speed and with minimum cost. However, the real world does not conform to the rules of the artificial universe created by AI simply because computers cannot match the social, cognitive and emotional intelligence of human beings.

There are now some philosophic trends now emerging that affect AI. They assert that our moral purpose is to follow individual self interest blindly even when that is at the expense of sacrificing the collective interest of society as a whole. It follows therefore that the only social system to support individualism is laisez faire capitalism and free markets. This only leads to disproportional levels of inequality between the poor and rich therefore plunging society into endless cycles of civil strife and violence.

Another related trend holds that governments should make no attempt to regulate, censor and  do anything else that would stand in the way of a free technological future. This also means ensuring the free flow of goods and services across all national boundaries.

A third ideological trend believes in the dream of transcending the limitations and imperfections of the human body by imparting human brain tissue in machine minds which carried further would gain us digital immortality. Can we now introduce some ethical behaviour in autonomous machines and if so how can we program them to do the right thing? This is still a question that needs to be resolved quickly because the stakes are simply too high to be ignored.

There is some worry about the close relations between the Big Tech companies found in Silicon Valley and the Big Oil companies found in Texas, specifically Houston. This is an unhealthy relationship that augurs ill for curbing the gas emissions from fossil fuels. The big oil tycoons want Silicon Valley to supply them with the AI tools to extract oil and gas from existing and new resources faster and more cheaply. This means that just 100 oil companies account for 71 percent of global emissions.

However, the greatest problem facing humanity today that is also playing an important role in our daily lives is the threat of climate change. Most people mistakenly believe that AI powered devices consume most of the world's energy resources. Research now shows us less than 1 percent of the world's energy consumption is spent on computing. Indeed all the big computing providers such as Google Cloud, and Microsoft Azure have pledged to reduce their current carbon emissions to zero net emissions by the year 2030 or 2040 at the latest.

However, Toby Walsh is quite optimistic that the future of mankind will be much brighter than it is today in spite of the warnings by the naysayers who predict impending doom. He reasons that given the fact that ongoing research across all countries, rich or poor, there is bound to be a breakthrough in the findings that will dispel the gloom and doom scenario painted by the naysayers. Surely then, the benefits of AI will clearly outweigh the disadvantages.