A.I. is Corrupting the Social Contract Between the Rich and the Poor
This is the 2nd in a series of articles considering how emerging technologies are impacting socio-economic and geo-political landscapes. The 1st article in the series looks at 1) how globalisation drives populism, 2) how Donald trump got elected in a thriving economy and 3) how A.I. creates more jobs than it takes.
Good technology often re-orientates human efforts away from something mundane, repetitive or risky towards tasks that generate more utility for society as a whole. Consequently, people tend to engage in safer, more productive, more rewarding tasks, leading to longer and more satisfying lives. Whilst the adoption of new technology tends to be in the best interest of society at large, we also know that it economically disadvantages the cohort of people it is replacing.
This human displacement is generational and has traditionally been societally mitigable so long as the displaced group is small. As it gets larger, either in terms of electoral franchise or economic impact, the group becomes more difficult to deal with, e.g. the miners strikes in the U.K. in the 70s and 80s.
Small incremental technological evolutions are easy to absorb as the capabilities of human capital tends to evolve in parallel, allowing for increased hybrid human-machine productivity. But large-impact events like the mass adoption of machine learning and the API infrastructure to create custom applications could produce significant technological and possibly structural unemployment.
Typically, economists would argue that the compensation effect will outweigh the level of structural unemployment. Yet, one would expect that structural unemployment will reach a level where so many people are out of work that the net gains in GDP cannot be reconciled with the fall in aggregate social utility. In other words, aggregate compensation resembles aggregate utility less as the Gini coefficient tends towards one.
We now seem to be reaching a point where the speed of machine evolution is such that the threat of technological unemployment could remove large quantities of labour and centralise power in small pools of capital to the extent that society cannot absorb it. This means that many people may be forced out of work in a way that would traditionally result in civil unrest and fractious political ramifications.
Unstable Equilibrium Straining the Social Contract
Technological and structural unemployment is certainly consequential, as is any resulting civil unrest, but there are broader civic ramifications of automation. A century ago, the labour movement fought to ensure a social contract existed between the rich and poor. This relationship secured improved working conditions, the five-day working week, enfranchisement for women and the landless, pensions and ultimately a mechanism for social mobility.
In return, the working class satisfied the industrial need for often repetitive, low-skilled, higher-risk labour that the wealthy relied upon. Additionally, the working class was disproportionately overrepresented in the military in times of war. In effect, this social contract provided low-skilled labour and military force in return for salary, improved conditions and opportunity for social mobility through work or the armed forces. Until now.
The real threat of mass automation is the destruction of the social contract between the rich and poor. Lines of code are replacing the jobs of the poor, and robots are increasingly deploying military forces. Moreover, many of the jobs most at risk are held by specific social or ethnic groups, which are likely to be disproportionately affected by massive machine evolution.
This is the system underlying modern democracy. Every citizen has a single equal vote regardless of wealth or status. The rich are obligated to the poor through politics, and the poor are obligated to the rich through economics. But what happens in a world where machines replace the functions of the poor? What happens when the owners of capital do not require unsophisticated labour? What is the free market incentive to provide freedom, of any sort, to those who the powerful no longer require? Why maintain a contract when only one side has any capacity to enforce it?
Many economists would suggest that the free market will maintain the social contract of the 20th century. If people don’t have jobs, they can’t spend. And if they can’t spend, then the owners of capital collapse in on themselves.
However, it is not the role of private enterprise to unnecessarily employ excess labour. That implies that the state, through the public sector, would be compelled to employ or pay large contingents of the more vulnerable parts of society. Businesses will end up paying for this either through taxes, or through income tax-driven wage inflation. This also implies a dependence on the nation state for wealth redistribution that the free market typically finds disagreeable.
Historians, on the other hand, would suggest that the obvious solution for the systematically downtrodden would be revolution. But what does revolution look like in a world of constant monitoring, limited privacy and comprehensive digital financial controls? What does revolution look like when it is no longer possible for the citizenry to equip themselves with the modern mechanics, knowledge and weaponry of war? It means that only those with power can achieve power.
The Role of the State in an Automated World
Universal basic income (UBI), shorter work weeks, retraining and anti-progress legislation have all been suggested as solutions, but all fall short. Since UBI establishes income by public decree, it’s inevitably dictated and influenced by the political necessities of a particular government. Shorter work weeks are expensive and ultimately diminish workers rights by requiring multiple people for the same job. Retraining works but requires significant funds and time.
Historically, technology has created more jobs and wealth than it has cost. It has been advantageous to all, and many of the social and civic equality movements to date have depended on innovations in technology and thinking. But the rate of change has shifted dramatically. Two 20-year-olds a century apart in the Middle Ages in Europe had a more similar life experience than two 20-year-olds 10 years apart today.
Speed of change is fundamental. It enables society to digest, contextualise and respond to those changes, which ultimately creates the stability necessary for democratic equality. Accelerating innovation may shift the power, irrevocably, into the hands of the owners of capital. This will eventually deconstruct a social contract that for the first time created a form of equality between the rich and the poor.
The concept of sentient machines and simulation theory is fun to indulge in but it’s a little like worrying about ghosts when there’s a burglar in your bedroom. The imminent and increasingly apparent threat of AI is that of social displacement and the effects it will have on global governance, economics and society. We need to develop mechanisms that will allow the population to directly benefit from the co-ownership of infrastructural technologies in order to hedge against displacement.
Social welfare infrastructures need to be redesigned to incentivise desired skills and competencies, and leveraged to actively coopt vulnerable workforces into retraining programmes ahead of time. Otherwise, the motivations to maintain democracy in its modern form will become increasingly strained and unconvincing to those who own the capital through which the armies of robots produce, control and destroy value.