Friday, February 22, 2019

Artificial Intelligence as a "Road to Serfdom"?

Even if so, AI would only be one of many factors which could exacerbate further losses of freedom in the years ahead. In particular, AI should not be scapegoated for the kinds of equilibrium defining actions which humans have been responsible for, all along. What's really at stake, is how individual actors and associations choose to implement artificial intelligence in markets and workplaces. Might its use become mostly limited to the augmentation of professional functions, for instance? Why haven't we thought more about the right to actively participate in applied knowledge, as a valid component of economic and personal freedom?

With complex issues such as these, it helps to recall how supply side considerations also impact economic outcomes - especially for the high skill time based product of service sectors. Given the fact non tradable sector activity has become largely responsible for the dynamism of advanced economies, how might existing opportunities for scale be encouraged? When basic aspects of domestic aggregate output are purposely limited, demand deficiency results. And unfortunately, these general equilibrium conditions don't readily respond to either fiscal or monetary stimulus, when basic supply side factors create most of the existing imbalance.

Today's best opportunities for scale, lie in the greater inclusion of all individuals in economic participation. How so? Our dominant sectors are heavily linked to time and place. Once assets or services are specifically time and place related, the greatest potential for gains in scale is in terms of aggregate participation. This reality is radically different from centuries of tradable sector dominance, in which gains in scale (and progress) were determined by a growing output trajectory which gradually required fewer labour inputs over time.

Plus, no society can remain free, should too many citizens find themselves excluded from the most basic forms of domestic economic activity with connections to time and place. In an article for Project Syndicate, Robert Skidelsky wonders whether AI might contribute to a "road to serfdom", in part because of the slowdown in wage growth:
Studies around the world show that people want secure jobs. At the same time, they have always dreamed of a life free of toil. The rise of the robots has made the tension between these impulses palpable.
Skidelsky also emphasized the fact that technology has in fact been able to bolster wage capacity for a long time. What I believe has not been highlighted enough, is the fact that non tradable sector dominance has proven responsible for much of the present wage conundrum. Yet while technology has already provided millions with workplaces essentially free of physical toil, AI is beginning to enter territory which is increasingly unappealing to professionals. After all, deep learning processes enable AI to supplant some aspects of work which are part of intellectual challenge. Most high skill work has always been free of toil in a "beast of burden" sense.

This is why the high skill work of the present, needs to be recognized for the intellectual challenge rewards it can actually provide, for anyone fortunate enough to take part. Most important, is that much of this is work people would actually be willing to perform even without pay, if the circumstances of their lives allowed them to do so. Just the same, the costs of living in our most prosperous regions not only require one to do desirable work for pay, it needs to be substantial pay to live in these settings as well. Consequently, we still have high hopes which - alas - won't be met, to somehow gain higher incomes and stable jobs for everyone who tries hard enough to obtain them.

Fortunately, good economic options are on the horizon. However, there's still plenty of social turmoil to get through before those options start to become more obvious. And the biggest hurdle which has led to so many dashed hopes, is the myriad of ways society has defined what success supposedly looks like. Presently, we are being ground under by a massive accumulation of societal expectations - even though many of them are little more than lifestyle illusions.

Once we recognize the possibilities of innovating the domestic parts of our economy - that is, the ones which so often manage to make our incomes appear too small, we can begin to redefine the non tradable sector equilibrium conditions which have made the maintenance of our present lifestyles so fragile. AI can help us in this challenge, by radically reducing today's human capital investment costs. Instead of struggling to bring income capacity closer to supposed general equilibrium "necessities", it makes more sense to allow non tradable sector equilibrium dynamics to reflect a wide range of local wage capacity. Eventually, variations on defined local equilibrium, could also give citizens the ability to ensure that AI improves the prospects of all human capital.

No comments:

Post a Comment