The following is a fragment of a wider argument around machine learning and its deployment within the limited anthropological horizon of Western thought that I wrote about two years ago. It perhaps works best as a development of some of the ideas in the previous blog post on Star Trek the Motion Picture.
----------------------------------------------------------------------
A distinction deserving of clear explication is between Reason and processes of rationalisation. On the surface, distinguishing between these two things should not be problematic. Reason, commonly taken as a universal faculty for human beings, should not be confused with the local historical movements and contingent processes through which it manifests in particular cultures towards particular ends.
This statement (unavoidably Platonic in formulation) is however contradicted forcefully by the continued prevalence of theories of modernisation which hold that the unfolding of man’s universal reason in economy, politics, and social organisation, follow a common structure and lead to common measurable outcomes, independent of the locality or historical traditions within which they are situated. According to this view Modernity itself is a universal, the revealed expression of that inner universal which is Reason; whereas culture and history (with a small ‘h’) are ephemeral peculiarities that are increasingly relativised by the underlying forces of modernisation, which paradoxically are thus construed as expressing a more authentic Idea.
As we know, the ideological conviction that Western modernity was the vanguard expression of a universal human truth became a cudgel wielded by the most technologically and militarily advanced nations in support of colonial violence. According to this logic the Civilising Mission, propelled by the inescapable forces of Progress, was, via the European Empires, an imposition of processes of rationalisation upon peoples less advanced down Reason’s royal road.
Despite much reassessment of the historical legacy of colonialism and so-called Western Universalism, some key characteristics have remained deeply ingrained in contemporary theories of development and democratisation, not least when they are applied to Western foreign policy. It is still common to associate economic development both with secularisation (itself a poorly understood concept) and the appearance of democratic government, despite mounting evidence to the contrary - most powerfully from the example of China. The underlying cause for this, I would claim, is the continuing association of universal human reason with Western processes of rationalisation.
We should not I think view as coincidental the fact that the theories of rationalisation and modernisation developed by Karl Marx and Max Weber emerged at the height of the Age of Empire (1875-1914 according to Hobsbawm’s chronology). Putting aside the vexed question of whether rationalisation actually leads to the flourishing of human freedom (Weber certainly had his doubts) it continues to be the case that development and modernisation are associated with the disappearance of certain social forms, in particular centralization, authoritarianism, and clientelism within governmental organisation. According to this criterion China is still very much a developing nation despite its technological prowess and its economy soon to overtake the US as the world’s largest.
This distinction also alerts us to a problem inherent in contemporary uses of the word rational. The present state of much of the Western world testifies to the fact that a society may become increasingly rationalised without necessarily becoming more reasonable. Indeed, the former may actively frustrate the latter. Bureaucratic procedures, market mechanisms and technological systems can each possess an impeccable internal logic whilst producing outcomes that appear unreasonable when considered from outside the limited purposes they have been designed to serve. Efficiency, calculability and predictability are not synonyms for reason; they are values generated within particular historical processes of rationalisation. The danger comes when the apparent success of these systems in organising one domain of social life encourages us to extend their categories into others. We then begin to mistake what is calculable for what is valuable, what is measurable for what is real, and what is administratively legible for what is socially desirable, and anything that escapes the grasp of these categories is either passed over or actively militated against.
The rapid development of artificial intelligence perhaps brings this problem into its most acute and potentially destructive form. Much of the rhetoric around AI assumes that progressively greater computational capacity represents an unambiguous augmentation of human reason. But this assumption risks reproducing precisely the confusion between reason and rationalisation thus described. Computation constitutes one extraordinarily powerful means of formalising certain kinds of reasoning, but it does not follow that everything rationally intelligible can therefore be rendered computationally intelligible. More importantly, the ends towards which computational systems are directed remain culturally, economically and historically determined.
A machine intelligence developed within an economy like our own, organised around accumulation, efficiency and competition, will inevitably reflect those imperatives, however technically sophisticated its operation may become. The question is therefore not simply whether machines can reason, but what historically particular conception of reason has already been embedded within the systems we are constructing. Just as nineteenth-century Europeans were liable to mistake the institutional forms produced by their own history for the destination of humanity as a whole, we should be wary of treating the technological forms produced by twenty-first-century capitalism as the inevitable culmination of human intelligence itself.
China is developing along lines that diverge from the model established by Western history. Or to put it another way, China shows that there are other paths, other processes of rationalisation which don’t necessarily lead to democratisation, or Western style impersonal rational-legal governance. China is of course integral to the global economic system, but its internal reflections, policy documents and centralised management of AI deployment within society, implies at least a different developmental pathway to the West, even if a metaphysics unique to that pathway has yet to come into view.
With the waning of Liberal Democracy in the West and the reappearance of charismatic power on the political stage, it may be time to reassess much of the dogma around universal history as it continues to form the backbone of so many missteps on the international stage, not least the foreign policy of Western States. Challenging such an ingrained self-image is perhaps the work of generations.
No comments:
Post a Comment