The Looming Crises at the Nexus of For-Profit Corporations and Artificial Intelligence
At the outset, prudence would suggest an introduction to the cast of characters that will make up this short narrative, and in so doing, an attempt to describe the conundrum in which they find themselves. After familiarizing ourselves with the core dilemma of the story, a potential solution will be offered in the form of a hybrid-business model.
Profit, that mischievous little devil, has been on the scene a long time, and most of us would be familiar with some lay definition of this much maligned actor — for these purposes I refer to it as the portion of the proceeds from a business that remain after accounting for all its operating expenses. Profit is our chief protagonist in this story, and from such simple beginnings doth a host of problems emerge.
To grasp the underlying conflict of the drama, the profit problem as it were, it is necessary to understand the fundamental relevance of profit to human wellbeing. While this may sound like the type of task a philosopher would set for themselves at the outset of particularly lengthy tome, but I will contend that the relevance of profit to the human wellbeing is deceptively simple.
First, it should be manifestly evident that profit is not of biological interest to humans per say, in that the principle itself can neither nourish tissue nor keep the rain off of one’s head, as other more tangible items can. While not of any survival value in the simplest sense, I suspect profit would likely be of little interest to humans outside of one important point– it has hitherto been useful as a broad indicator of the well-being for groups of humans. That is, it can be a stand in for other things which more closely reflect human well-being.
Consider that profit can be used to summarize the success or failure of a business venture, which can then be used to summarize the relative wellbeing of at least some members of said business, and when compounded across an entire host of businesses, gives some indicator of the well-being of that entire group. If business is brisk, than the hope is that people will be paid, and with that pay, be able to procure for themselves the necessary accoutrements of survival, that is food, water, shelter etc. The chain of relevance for profit as an indicator is not a particularly lengthy one as indicators go, so it is easy to see how one might take for granted the line of connections that make it relevant to humans at a biological level.
That profit is only useful as an indicator for human wellbeing is also what makes it problematic. Because it is several steps removed from human well-being, there is always the potential that the chain will be broken and it will cease to reflect the actual things it is standing in for, that is, the ability to procure food, shelter, fresh air etc. This has arguably happened before, for instance in 19th century England at the outset of the industrial revolution when the underclass was subjected to working conditions that perhaps made their survival chances worse than had they been hunter gatherers in the Amazon basin lacking any concept of profit whatsoever. Even so, profit arguably has served as a broad indicator for human wellbeing in a way that few other indicators have, and therefore maintained its vaunted position in the drama of human affairs.
2. The Unfortunate Power of Deep Learning
3. Graph Neural Network for 3D Object Detection in a Point Cloud
4. Know the biggest Notable difference between AI vs. Machine Learning
Unless I am much mistaken, that chain of relevance that makes profit useful as an indicator is now being severed at an alarming rate. Which brings us to our second character in this brief drama — artificial intelligence. Artificial intelligence and the algorithms behind it are increasing in aptitude at an alarming rate, I should know because I help program them, and thanks to cutting edge developments in neural networks and reinforcement learning, will likely impact profit in a seismic way.
Firstly it will likely increase the value of capital in respect to labor. With the benefit of AI, capital can procure ever increasingly amounts of profit in comparison with labor, perhaps to an exponential degree. The result is that those with capitol are likely to see ever increasingly returns on it, while those with only their labor, are likely to see their share of profits shrink. The net result is a widening of disparities between those who have capitol and those who have merely their own labor. Such a process may have already been under way for a long time, but the advent of strong AI will likely push this to extremes.
The second impact of AI is that it may begin to supplant human labor altogether. I am neither the first, nor the most erudite to make this observation. The historian Yuval Harari has written about it extensively in his books Sapiens and Homo Deus when he speaks of the rise of a “useless class”. But long before Harari sounded the alarm it had been foretold by a number of economists, many of whom grasped that as technology continued to replace human labor in the ingredient mix of economic activities, there could come a time when wages would diminish in proportion to profit, and profit would no longer be the useful indicator for human well-being that it once was.
While such luddite fears have tended to be misplaced — technological advances have historically shifted labor from one sector to another rather than replacing it altogether, there is no guarantee this will hold true in the future. Once many of the mental attributes of humans can be equaled or surpassed by machine intelligence, it is hard to say what new sector of the economy will be large enough to absorb these displaced workers. Indeed the past is not always a reliable guide to the future, and it is wise to remember that many of today’s retail jobs do not require the full flowering of human intelligence to be performed adequately. An AI with merely the intelligence of an eight year old might be sufficient to replace large swaths of the labor market.
At that point, unless measures were taken to tie profit back to human well-being, it could cease to function as a useful indicator. Indeed, profit might well turn into something entirely malignant, an optimization principle for corporations operating without a care in the world for human well-being. One can even imagine an economy of AI entities buying and selling goods to each other, making tremendous profits in the process, but without a single human in the loop. The stock market itself is coming closer by the day to resembling such a market, as more and more trades are taken autonomously by algorithms.
Which is all to say, it likely that the pool of humanity whose wages actually benefit from the automation technologies driving the rapid growth of IT companies such as Google and Amazon will diminish as those companies continue to replace their human inputs with technological ones. The outsized profits these companies garner will be but dimly reflected in the wellbeing of the societies in which they operate. The largest portion of these profits could accrue in very few hands, or potentially in no human hands at all. The fact that corporations, at least in the United States, can legally function as citizens and therefore profits accrue to them even when no human is directly at the helm, augurs for such a reality.
Unless we work speedily to solve this profit problem, and ensure that it remains linked to human well-being at a fundamental level, we may find ourselves at an impasse where a large portion of the populace are impoverished, psychologically or physically, even while enormous profits are being made. The disturbing downward trend of human life expectancy in the United States over the last handful of years may be an indicator of just such a failure.
But even if one does not accept that we are on the horns of massive technological disruption, there would be good reason to seek measures that more closely tie profit to human well-being, if only to avoid the outsized disparities that could result from a labor capital imbalance.
A number of potential solutions have been advanced over the decades, from the grand to the grotesque. While many of these are no doubt laudable, each suffers from implementation problems and none have proved scalable in a way that would suggest we are now safely out of the woods. Which brings us to the point in the story where we introduce a potential hero-in-waiting; the hybrid business model.
The premise of the hybrid business model is not to do away with capitalism or the profit principle, but rather to make such alterations which are necessary to more closely link profit to human well-being. To understand one way in which this could be done, imagine an idealized basket of goods that would represent as closely as possible, those characteristics of human well-being that are universally shared. This is no easy task, but I believe there are a few simple principles that jump out straightaway. Whatever one’s creed, disposition or race — things like fresh air, clean drinking water, a strong and intimate sense of community, shelter from the elements and sufficient money in one’s pocket to procure food, would likely be considered directly reflecting human well-being more so than profit itself. We can imagine this basket of goods comprising something like an idealized neighborhood, call it the “eco-village”, for indeed, the principles of sustainability would need to be as important as any other if the new indicator is to serve any long term purpose. Now that we imagined this idealized indicator in the form of an eco-village, the last step that remains is to yoke it to the profit principle.
Fortunately we have a long history of lawyering for just such tasks. And while by no means simple, the basic premise can be conceived simply enough. The chief difficulty arises in keeping the new linkage as incentive neutral as possible, so as not to distort the principles of supply and demand that make free market capitalism so effective. In this conception, a for profit business could choose to go “profit-humane” adopting a charter in which, past a certain point, perhaps 100% return on investment, the profit produced by the company would be devolved to a neighborhood development project of a fixed and sustainable model, the so called “eco-village”.
Once the neighborhood / eco village had reached its maximum capacity, perhaps 80 members since there is evidence this is the largest group for which the humans can reliably recall social bonds, a new neighborhood would be started. In this conception, as more funds became available through increased profit of the business, different layers of the neighborhood would be added, in order of their importance, potentially culminating in a universal basic income at a neighborhood level.
Perhaps in most ways this neighborhood could function as a normal “subdivision”, with members being free to buy and sell their homes as they pleased and market forces determining the price of a plot. In such a way, membership in such a neighborhood could be seen as a stand in for owning shares of a for profit company, the amenities provided acting as a kind of dividend for holding such shares. While the formulation of such a community model, and the legal permutations necessary to create such hybrid profit structures are certainly daunting in complexity, they may be more attainable than other solutions.
Such a hybrid business model has the advantage of being small enough that it could be undertaken at the grass roots level, with individual companies opting to go “profit-humane” just as many companies now opt to go “fair-trade”. This article is clearly only the beginning of such an endeavor, the beginning of a conversation rather than the end of one. If the reader should feel motivated to help in the formulation of such hybrid business models, perhaps having some special knowledge related to business law, community development, sustainability etc, I would invite you to reach out and share your perspective. The aim is, that once formulated, such a model could be made available for free on the internet, and be adopted or experimented with in the manner of open source software.