Digifesto

Tag: max horkheimer

Horkheimer and Wiener

[I began writing this weeks ago and never finished it. I’m posting it here in its unfinished form just because.]

I think I may be condemning myself to irrelevance by reading so many books. But as I make an effort to read up on the foundational literature of today’s major intellectual traditions, I can’t help but be impressed by the richness of their insight. Something has been lost.

I’m currently reading Norbert Wiener’s The Human Use of Human Beings (1950) and Max Horkheimer’s Eclipse of Reason (1947). The former I am reading for the Berkeley School of Information Classics reading group. Norbert Wiener was one of the foundational mathematicians of 20th century information technology, a colleague of Claude Shannon. Out of his own sense of social responsibility, he articulated his predictions for the consequences of the technology he developed in Human Use. This work was the foundation of cybernetics, an influential school of thought in the 20th century. Terrell Bynum, in his Stanford Encyclopedia of Philosophy article on “Computer and Information Ethics“, attributes to Wiener’s cybernetics the foundation of all future computer ethics. (I think that the threads go back earlier, at least through to Heidegger’s Question Concerning Technology. (EDIT: Actually, QCT was published, it seems, in 1954, after Weiner’s book.)) It is hard to find a straight answer to the question of what happened to cybernetics?. By some reports, the artificial intelligence community cut their NSF funding in the 60’s.

Horkheimer is one of the major thinkers of the very influential Frankfurt School, the postwar social theorists at the core of intellectual critical theory. Of the Frankfurt School, perhaps the most famous in the United States is Adorno. Adorno is also the most caustic and depressed, and unfortunately much of popular critical theory now takes on his character. Horkheimer is more level-headed. Eclipse of Reason is an argument about the ways that philosophical empiricism and pragmatism became complicit in fascism. Here is an interested quotation.

It is very interesting to read them side by side. Published only a few years apart, Wiener and Horkheimer are giants of two very different intellectual traditions. There’s little reason to expect they ever communicated (a more thorough historian would know more). But each makes sweeping claims about society, language, and technology and contextualizes them in broader intellectual awareness of religion, history and science.

Horkheimer writes about how the collapse of the Enlightment project of objective reason has opened the way for a society ruled by subjective reason, which he characterizes as the reason of formal mathematics and scientific thinking that is neutral to its content. It is instrumental thinking in its purest, most rigorous form. His descriptions of it sound like gestures to what we today call “data science”–a set of mechanical techniques that we can use to analyze and classify anything, perfecting our understanding of technical probabilities towards whatever ends one likes.

I find this a more powerful critique of data science than recent paranoia about “algorithms”. It is frustrating to read something over sixty years old that covers the same ground as we are going over again today but with more composure. Mathematized reasoning about the world is an early 20th century phenomenon and automated computation a mid-20th century phenomenon. The disparities in power that result from the deployment of these tools were thoroughly discussed at the time.

But today, at least in my own intellectual climate, it’s common to hear a mention of “logic” with the rebuttal “whose logic?“. Multiculturalism and standpoint epistemology, profoundly important for sensitizing researchers to bias, are taken to an extreme the glorifies technical ignorance. If the foundation of knowledge is in one’s lived experience, as these ideologies purport, and one does not understand the technical logic used so effectively by dominant identity groups, then one can dismiss technical logic as merely a cultural logic of an opposing identity group. I experience the technically competent person as the Other and cannot perceive their actions as skill but only as power and in particular power over me. Because my lived experience is my surest guide, what I experience must be so!

It is simply tragic that the education system has promoted this kind of thinking so much that it pervades even mainstream journalism. This is tragic for reasons I’ve expressed in “objectivity is powerful“. One solution is to provide more accessible accounts of the lived experience of technicality through qualitative reporting, which I have attempted in “technical work“.

But the real problem is that the kind of formal logic that is at the foundation of modern scientific thought, including its most recent manifestation ‘data science’, is at its heart perfectly abstract and so cannot be captured by accounts of observed practices or lived experience. It is reason or thought. Is it disembodied? Not exactly. But at least according to constructivist accounts of mathematical knowledge, which occupy a fortunate dialectical position in this debate, mathematical insight is built from embodied phenomenological primitives but by their psychological construction are abstract. This process makes it possible for people to learn abstract principles such as the mathematical theory of information on which so much of the contemporary telecommunications and artificial intelligence apparatus depends. These are the abstract principles with which the mathematician Norbert Wiener was so intimately familiar.

Horkheimer, pragmatism, and cognitive ecology

In Eclipse of Reason, Horkheimer rips into the American pragmatists Peirce, James, and Dewey like nobody I’ve ever read. Normally seen as reasonable and benign, Horkheimer paints these figures as ignorant and undermining of the whole social order.

The reason is that he believes that they reduce epistemology to a kind a instrumentalism. But that’s selling their position a bit short. Dewey’s moral epistemology is pragmatist in that it is driven by particular, situated interests and concerns, but these are ingredients to moral inquiry and not conclusions in themselves.

So to the extent that Horkheimer is looking to dialectic reason as the grounds to uncovering objective truths, Dewey’s emphasis on the establishing institutions that allow for meaningful moral inquiry seems consistent with Horkheimer’s view. The difference is in whether the dialectics are transcendental (as for Kant) or immanent (as for Hegel?).

The tension around objectivity in epistemology that comes up in the present academic environment is that all claims to objectivity are necessarily situated and this situatedness is raised as a challenge to their objective status. If the claims or their justification depend on conditions that exclude some subjects (as they no doubt do; whether or not dialectical reason is transcendental or immanent is requires opportunities for reflection that are rare–privileged), can these conclusions be said to be true for all subjects?

The Friendly AI research program more or less assumes that yes, this is the case. Yudkowsky’s notion of Coherent Extrapolated Volition–the position arrived at by simulated, idealized reasoners, is a 21st century remake of Peirce’s limiting consensus of the rational. And yet the cry from standpoint theorists and certain anthropologically inspired disciplines is a recognition of the validity of partial perspectives. Haraway, for example, calls for an alliance of partial perspectives. Critical and adversarial design folks appear to have picked up this baton. Their vision is of a future of constantly vying (“agonistic”) partiality, with no perspective presuming to be settled, objective or complete.

If we make cognitivist assumptions about the computationality of all epistemic agents, then we are forced to acknowledge the finiteness of all actually existing reasoning. Finite capacity and situatedness become two sides of the same coin. Partiality, then, becomes a function of both ones place in the network (eccentricity vs. centrality) as well as capacity to integrate information from the periphery. Those locations in the network most able to valuably integrate information, whether they be Google’s data centers or the conversational hubs of research universities, are more impartial, more objective. But they can never be the complete system. Because of their finite capacity, their representations can at best be lossy compressions of the whole.

A Hegelian might dream of an objective truth obtainable by a single subject through transcendental dialectic. Perhaps this is unattainable. But if there’s any hope at all in this direction, it seems to me it must come from one of two possibilities:

  • The fortuitously fractal structure of the sociotechnical world such that an adequate representation of it can be maintained in its epistemic hubs through quining, or
  • A generative grammar or modeling language of cognitive ecology such that we can get insights into the larger interactive system from toy models, and apply these simplified models pragmatically in specific cases. For this to work and not suffer the same failures as theoretical economics, these models need to have empirical content. Something like Wolpert, Lee, and Bono’s Predictive Game Theory (for which I just discovered they’ve released a Python package…cool!) may be critical here.

Eclipse of Reason

I’m starting to read Max Horkheimer’s Eclipse of Reason. I have had high hopes for it and have not been disappointed.

The distinction Horkheimer draws in the first section, “Means and Ends”, is between subjective reason and objective reason.

Subjective reason is the kind of reasoning that is used to most efficiently achieve ones goals, whatever they are. Writing even as early as 1947, Horkheimer notes that subjective reason has become formalized and reduced to the computation of technical probabilities. He is referring to the formalization of logic in the Anglophone tradition by Russell and Whitehead and its use in early computer science, most likely. (See Imre Lakatos and programming as dialectic for more background on this, as well as resonant material on where this is going)

Objective reason is, within a simple “means/ends” binary, most simply described as the reasoning of ends. I am not very far through the book and Horkheimer is so far unspecific about what this entails in practice but instead articulates it as an idea that has fallen out of use. He associates it with Platonic forms. With logos–a word that becomes especially charged for me around Christmas and whose religious connotations are certainly intertwined with the idea of objectivity. Since it is objective and not bound to a particular subject, the rationality of correct ends is the rationality of the whole world or universe, it’s proper ordering or harmony. Humanity’s understanding of it is not a technical accomplishment so much an achievement of revelation or wisdom achieved–and I think this is Horkheimer’s Hegelian/Marxist twist–dialectically.

Horkheimer in 1947 believes that subjective reason, and specifically its formalization, have undermined objective reason by exposing its mythological origins. While we have countless traditions still based in old ideologies that give us shared values and norms simply out of habit, they have been exposed as superstition. And so while our ability to achieve our goals has been amplified, our ability to have goals with intellectual integrity has hollowed out. This is a crisis.

One reason this is a crisis is because (to paraphrase) the functions once performed by objectivity or authoritarian religion or metaphysics are now taken on by the reifying apparatus of the market. This is a Marxist critique that is apropos today.

It is not hard to see that Horkheimer’s critique of “formalized subjective reason” extends to the wide use of computational statistics or “data science” in the vast ways it is now. Moreover, it’s easy to see how the “Internet of Things” and everything else instrumented–the Facebook user interface, this blog post, everything else–participates in this reifying market apparatus. Every critique of the Internet and the data economy from the past five years has just been a reiteration of Horkheimer, whose warning came loud and clear in the 40’s.

Moreover, the anxieties of the “apocalyptic libertarians” of Sam Franks article, the Less Wrong theorists of friendly and unfriendly Artificial intelligence, are straight out of the old books of the Frankfurt School. Ironically, todays “rationalists” have no awareness of the broader history of rationality. Rather, their version of rationality begins with Von Neummann, and ends with two kinds of rationality, “epistemic rationality”, about determining correct beliefs, and “instrumental rationality”, about correctly reaching ones ends. Both are formal and subjective, in Horkheimer’s analysis; they don’t even have a word for ‘objective reason’, it has so far fallen away from their awareness of what is intellectually possible.

But the consequence is that this same community lives in fear of the unfriendly AI–a superintelligence driven by a “utility function” so inhuman that it creates a dystopia. Unarmed with the tools of Marxist criticism, they are unable to see the present economic system as precisely that inhuman superintelligence, a monster bricolage of formally reasoning market apparati.

For Horkheimer the formalization and automation of reason is part of the problem. Having a computer think for you is very different from actually thinking. The latter is psychologically transformative in ways that the former is not. It is hard for me to tell whether Horkheimer would prefer things to go back the way they were, or if he thinks that we must resign ourselves to a bleak inhuman future, or what.

My own view is that a formalization of objective reason would allow us to achieve its conditions faster. You could say I’m a logos-accelerationist. However, if the way to achieve objective reason is dialectically, then this requires a mathematical formalization of dialectic. That’s shooting the moon.

This is not entirely unlike the goals and position of MIRI in a number of ways except that I think I’ve got some deep intellectual disagreements about their formulation of the problem.