Tag Archives: labor

Notes on Endnotes on class decomposition and technical labor

As summarized in my last post, the main essay series of Endnotes Issue 4, “A History of Separation”, is all about dismantling orthodox Marxist mythology around the importance of the industrial worker as the main protagonist of overthrowing capitalism and establishing socialism.

A key part of this argument, which I didn’t touch on in the summary, is relationship between technology, class, knowledge, and political identity.  Back in the earlier stages of industrial production and industrial labor in the late 19th and early 20th centuries, when revolutionary workers’ movements were blossoming, it was very easy for a worker to see themselves as part of a class that was the actual force that built and maintained society.  Workers sweated and toiled alongside each other in massive centralized plants, stitching clothes, building cars, refining oil, chopping trees, etc.  Workers were key to the production process, and they understood the process just as well as, if not better than, the capitalists who pocketed the profits.

However, as development and industrialization proceeded (under both capitalist and Marxist-Leninist regimes), worker knowledge of the means of production diminished.  The industrial use of technology fragmented the labor process, separated the supply chain across entire regions and countries, and made work repetitive and dull, and in general, destroyed the old basis for a common, positive identity around being an industrial worker.  From “The Strange Victory of the Workers’ Movement”:

In spite of the development of the productive forces, labour, it was insisted, remained the source of all wealth, its latent power and knowledge reflected precisely in that development. That turned out not to be true: knowledge of the production process was no longer located in the place of the collective worker, but rather (if anywhere), in the place of the collective technician. That was a key point because — while it upended the foundation-stone of the workers’ movement — it also finally confirmed Marx’s perspective in the “fragment on machines” (reproduced more soberly in Capital).

…It was increasingly the case that human labour was no longer the main productive force; science — often applied to the worst ends of industrial “development”— took labour’s place. That profoundly affected workers’ self-understanding, their experience of what they did and their place in the world: workers could no longer see themselves as building the world in the name of modernity or a better, more rational way of living. On the contrary, that world was already built, and it was entirely out of their hands. Modernity presented itself as this imposing thing, which workers’ confronted, not as subject, but rather, as an object to be regulated and controlled.

…Society is no longer just the means of production, a set of factories that can be taken over and self-managed by the workers who run them…The bases of social power are now much more dispersed. They are located not just in the repressive apparatuses of the police, the jails and the armed forces and the so-called “ideological” apparatuses of schools, churches, and television. They include also power stations, water-treatment plants, gas stations, hospitals, sanitation, airports, ports, and so on. Just like the factories themselves, all of this infrastructure relies on a legion of engineers and technicians, who keep the whole things running from minute to minute. These technicians possess no collective workers’ identity, nor were they ever included in the programmes of the workers’ movements.

The comments on the rise of science, and the importance of technical labor, stuck out since of course it connects with what I’ve been musing about regarding the current politicization and radicalization of workers in the tech industry in the US.

To me, what’s key here is the fact that the workers’ movement did not adapt or evolve in step with the evolution of class as it was restructured by capitalism, to acknowledge the growing importance of scientific and engineering labor in the increasingly complicated complex of production, distribution, and circulation.  Some groups did this, but had to split from the old vanguard in order to do so.  Two examples that I’ve taken a closer look at: 1) the Students for a Democratic Society’s “Port Authority” statement looking at the new class composition of the US, that included skilled technical workers, and 2) debates among French Marxists in various parties over the nature of the new classes in light of the post-1968 period.

The latter is particularly interesting in the context of the comments by Endnotes because it precisely shows how class recomposition provoked crisis and splits in Marxism and the workers’ movement.  The French Communist Party (PCF) stuck with the industrial workers as their social base, but bizarrely tried to build a reformist coalition with the new middle classes who they saw as non-revolutionary and more conservative.  This was presumably a reaction to the fact that parts of this middle class, particularly the skilled technicians, played a key role in the upheavals of 1968, even as the PCF was playing a backseat counter-revolutionary role (which begs the question why they continued to see these classes as less radical than themselves).  Meanwhile, theorists like Serge Mallet of the  an Alain Touraine adapted the old logic of the workers’ movement to uphold the skilled technical workforce as the new vanguard of technologically advanced and automated capitalism, as the only fraction of the working class which could grasp the complexity of post-industrial society.

However, at this point in time, it certainly doesn’t seem like any identity around a “collective technician” has emerged, let alone one that is revolutionary and anti-capitalist (yet!).  Nonetheless, its very worthwhile to go back and look at these older debates around the class nature of technicians and engineers and scientists, to help us in the present figure out how to go about composing a revolutionary proletarian composition that is dynamic and cohesive across an increasingly fractured landscape.

Notes on Endnotes’ “A History of Separation”

Endnotes Issue 4, “A History of Separation”, is all about situating the workers’ movement of the 1800s and 1900s into its historical context.  It continues with Endnotes  rather heretical central thesis that the industrial proletariat is not the class that will overthrow capitalism, and that in fact capitalism will not automatically create a revolutionary class in the first place.

Here is a summary of the arguments that stood out to me.


The workers’ movement of the 19th and 20th century was couched in a very specific identity, that of the industrial worker.  This identity was rooted in the conditions and experiences of laboring in the factories, but also in the institutions and social spaces built by the workers’ movement to help reproduce this identity and pave over divisions of language, religion, etc.  This was all closely connected to the orthodox Marxist valorization of the industrial worker as the producer of all value in society, and thus the rightful heir to the wealth being produced by capitalism.

This not only informed, but overdetermined the class politics of the era.  Early communists believed that capitalism, as it developed, would inexorably proletarianize the rest of society, expanding the ranks of the industrial workforce and uniting them in the factory across divisions of nation, language, religion, etc., and creating a revolutionary force that would overthrow capitalism and establish socialism, a dictatorship of the proletariat.  This seemed correct for a time as industrialization, urbanization, and the expansion of the industrial workforce proceeded; but as capitalism developed and became fully mature in the mid-20th century, the real long-term trend became deindustrialization and the marginalization of industrial workers, as automation steadily reduced the relative need for industrial labor, and produced all sorts of different sectors and strata of workers with no real shared experience or identity — and in particular, a growing population of surplus people of no use to capital.  Capitalism did not unite the masses into a homogeneous collective industrial class, but instead alienated and separated them from one another; the only unity produced by capitalism is the unity of being atomized from everybody and everything, and being dependent on the market.  Unity-in-separation.

The excessive focus on the industrial worker by the workers’ movement also explains the policies of the revolutionary Marxist-Leninist states, which took hold in societies which were backwards and agricultural.  These states saw their primary task as developing the productive forces and creating the proletariat in the first place, and so ruthlessly attacked and undermined and displaced peasants and other rural/agricultural classes, driving them into the factories.  The underlying logic was to produce the revolutionary class that the old regime, dominated by feudal elites who stifled capitalism, had failed to produce.  But this focus on expanding industrial production and industrial workers replicated in the socialist bloc the same dynamic as in the capitalist bloc: the tendency toward the technological displacement of workers and the production of alienation and separation.

Deindustrialization and the full realization of unity-in-separation also meant that the shortcomings of the workers’ movement — whose self-identification revolved largely around the figure of the semi-skilled white male worker — became obvious, thus leading to the relative increase in importance of the various social movements around gender, race, sexuality, etc. in the 1960s and afterwards.  With the identity of the worker no longer able to mobilize a strong and powerful force, other identities became more materially powerful.

None of this is about whether communists at various times had mistaken ideas in their heads about how to go about revolution.  Rather, it is an assessment about how the leading revolutionary strategies of history were emergent from their specific material context, with regards to the particular phase of global capitalism at the time.

Since the long-term tendency of capitalism is the production of a unity-in-separation, the task for revolutionary communists is to engage with the problem of composition, of how to add up different fractions of the exploited masses into a revolutionary force.  What’s necessary, then, is not “class consciousness”, but a “consciousness of capital” as the common enemy of all of our disparate and differentiated lives.  From the last paragraphs of the second-to-last section:

Ours is a society of strangers, engaged in a complex set of interactions. There is no one, no group or class, who controls these interactions. Instead, our blind dance is coordinated impersonally, through markets. The language we speak — by means of which we call out to one another, in this darkness — is the language of prices. It is not the only language we can hear, but it is the loudest. This is the community of capital.

When people make the leap out of that community, they will have to figure out how to relate to each other and to the things themselves, in new ways. There is no one way to do that. Capital is the unity of our world, and its replacement cannot be just one thing. It will have to be many.


So how much of this do I agree with?  A lot, I think.  Its absolutely necessary to not be weighed down with the traditions and myths of the past, and moving beyond the heroic figure of the industrial worker in favor of examining capitalism and class as it actually exists today is a vital task.  In this regard the arguments of Endnotes echos that of Autonomist Marxism, and indeed “A History of Separation” starts with a quote from Mario Tronti.  There is no doubt that the working class today is heterogeneous and stratified, encompassing everybody from software developers to teachers to meatpackers to waitresses.  Building a basis for unity must be the point of organizing, not something assumed or asserted.

The points on automation and decomposition are also on the mark.  This essay goes into more depth on the question of automation and employment, and reaches more or less the same conclusion: that while automation expels the working class from some segments of production, there is a corresponding increase in labor demand in other sectors, mainly lower-wage and more precarious service work, or what Endnotes terms as “surplus population” (surplus to the core sectors of capital but not to the capitalist economy).

I’m not nearly as pessimistic as the essay is with regards to the problem of composing a revolutionary force.  At the risk of sounding like a simpleton I think it is pretty easy to assert that there is more that binds us than separates us, and despite the mass production of unity-in-separation its not at all a hard sell for most people that regardless of our occupations, national backgrounds, etc. we all gotta eat, and we all have to take shit from some kind of boss.  Pushing this generic observation to actual material alliances built around class struggle is another story, of course, but I’m optimistic about these prospects based on the organizing I’m seeing around me, particularly in the tech industry where workers across strata are working together and slowly grasping toward a common identity built through struggle.

Indeed reading Endnotes makes me outright optimistic, despite them being labelled as “bleak” by magazines like N+1, possibly because it echoes what I’ve been hypothesizing on my own in recent years: that the key task of revolutionary communists is to act as catalysts for different fractions of the working class to come together in common struggle, by facilitating communication and collaborative efforts, unveiling avenues of concrete solidarity, and doing whatever else it takes to recompose the working class.

Amazon, the decimation of warehouse worker wages, and a warehouse inquiry

In the latest issue of Economist, there is an article with some pretty stunning analysis about the wages of warehouse workers in US counties where Amazon sets up shop — specifically on how they collapse.  The following two graphs speak for themselves.

What’s behind this?  According to the analysis cited by Economist, it appears to be a combination of Amazon workers being younger, more inexperienced, and more unskilled than in other warehouses, and generally not able to find alternative jobs due to there only being a few employers in the area.  Technology also plays a role here, with cutting-edge automation allowing the company to hire younger and less skilled workers in the first place, which is deeply related to the argument that automation doesn’t eliminate the need for work, but rather helps generate the need for less skilled work.

Amazon is growing fast, already worth more than all the major brick-and-mortar retail companies put together.  Its combination of retail, logistics, and tech is allowing it to devour large swathes of the US economy into itself.  Amazon and its low-wage, cyborg workforce is the future — and the present, for that matter.  Engaging in militant labor struggles in Amazon warehouses will only become more and more critical for all those interested in rebelling against the rule of capital.  Thus inquiries and workplace reports, like this one just released by Angry Workers of the World, are a valuable resource that worker militants should use and produce themselves, in order to pick apart the nature of the workplace and reveal the ambient level of worker unrest and struggle.  To quote from the conclusion of the first report:

This all sounds bad, but don’t believe that the workers just sat and took it like the good victims the newspapers like to write about every now and again. In the beginning a lot of people had high expectations of working with Amazon, but after a few weeks they started to realise what working for Amazon really meant. So after a few weeks you started to hear more and more angry and incensed discussions amongst workers around the aisles of the pick tower. Workers who in the beginning tried to run themselves into the ground trying to reach their targets, now having realised it didn’t make a difference in terms of getting a long term contract, stopped stressing about targets and deliberately worked slower than they could. In the beginning we were all worried about even going to the toilet because we might get a warning for “time off task” but after a few weeks when we started to realise we would all be fired soon anyway more and more people started to take “extra” breaks, spending time talking with colleagues, wandering around the warehouse, going to the canteen to grab a cup of coffee, playing a game of ping pong, and of course not giving a toss about ‘power hour’. The permanent staff already know that Amazon don’t care about the workers and the temps quickly learn it, and a lot of us start to do minor individual acts of resistance. That is all a good start, but if we want to change the way Amazon treats us we have to work and resist together!

Moving beyond Marx’s fetishization of science and technology

One of the starting points of my attempts at arguing the importance of analyzing and organizing techno-scientific workers has been a perceived shortcoming in existing Marxist theory on the matter.  McKenzie Wark, in the in introduction to General Intellects: Twenty-One Thinkers for the Twenty-First Century (2017) has some excellent comments on this shortcoming, starting from the arguments of Karl Marx himself.

Wark focuses on Marx’s Fragment on Machines, a set of notes that speculate that technological development is the transferring of skills and knowledge of workers into machines in a process mediated by and for the interests of capital.  However, there is a gap here in terms of the actual process of how this information is turned into machines in the first place.

For Marx, science will appear to the laborer as something alien to him.  Science appears in the form capital dictates.  Science is a productive force: “all the sciences have been pressed into the service of capital”.  But who makes science?  “Invention has become a business,” says Marx, but who does the inventing?…

…The problem is that as actual organized social activities, science and technology do not fit so neatly into the schema of labor and capital.  Hence in Marx they simply come from without as a reified thing called “science” which then becomes part of the machine system as fixed capital. (8-9)

In order to move past this fetishized understanding of science and technology, Wark offers analysis by both J.D. Bernal and himself.  In Bernal’s Science in History series, he argues that contemporary techno-scientific labor (as of the 1950s) was a fusion of high-skill technical labor and the bourgeoisie “gentlemanly culture” of leisurely philosophical-scientific inquiry.  Thus, the new scientific workers who make the science and do the inventing constitute a new class with hybrid origins.  Wark, in A Hacker Manifesto (2004), similarly argues that there is a new techno-scientific class of workers that he dubs “the hacker class”, who today mainly labor with information and are dealing directly with capital’s commodification and enclosing of information and knowledge.  (It would be interesting to compare the arguments of Bernal and Wark with those in “The Californian Ideology”, which seems to take a much more pessimistic view of the new techno-scientific classes).  All of this echoes other efforts at a class analysis of techno-scientific workers, such as analysis by student leftists in the US in the 1960s, debates among French Marxists in the 1970s, and of course recent writings on the new “Tech Left”.

Wark’s comments on this can be summed up in this paragraph that ties the above comments to the project of the book overall:

One task for general intellects might be to imagine a kind of common hacker class interest among those whose efforts end up being commodified as some sort of intellectual property: artists, scientists, engineers, even humanist and social science academics.  We could imagine all of them as belonging to the same class from the point of view of the commodification of information.  We all process information that is part of a complex natural-technical-social-cultural metabolism.  But nearly all of us get to see a ruling class of a rather unprecedented kind extract most of the value from the combined efforts of hackers and workers worldwide.  As general intellects, maybe we should stick our heads above our little cubicles, look around, and figure out how to cooperate with others who understand different parts of the labor process. (11)

Communist strategy, international coordination, and the pillar of the Gulf monarchies

Around a year ago Angry Workers of the World published an excellent document around the question of a workers’ insurrection, that looks at the matter in a very concrete and material way.

There is a huge amount in the document that can be discussed, but one thing in particular that stood out was this comment about how to tackle questions of international integration (emphasis added):

Maybe because of the generalisation of the ‘proletarian condition’ of being wage dependent and of the generalisation of ‘parliamentary democracy’ across the globe it now seems obsolete to talk about the impact of uneven development. Everything appears at the same time so similar (global village) and so different, once we look into details. The problem is that we clearly see the effect of regional differences on global class struggle, but:

a) we tend to explain these differences geopolitically or out of ‘national economies’ or even ethnically (oil producing nations, BRIC states, Arab Spring);
b) we celebrate a crude pluralism (‘patchwork of free and unfree labour; all sorts of proletarian income etc.);
c) we don’t develop revolutionary strategies of how regional struggles or struggles within certain stages of development relate to others.

That last bit is key.  This question of how struggles in one part of the world affect other parts of the world is a fascinating and important area of study, and something that I personally started thinking about an awful lot during and after writing an analysis of Saudi Arabia and its historical roots in imperialism.  It turned out that Saudi Arabia and the other Gulf petro-monarchies have played a major and pivotal role in the functioning of global capitalism, particularly in the restructuring toward neoliberalism after the 1970s, as described in detail in Adam Hanieh’s Capitalism and Class in the Gulf Arab States (2011).  If the left-wing movements in the area had been successful in the ’50s and ’60s, it would have changed the course of world history.

Today the Gulf states’ massive oil resources are tightly integrated with global finance capital, as well as with a regional market of precarious migrant labor.  A resurgence in communist struggle in this area would almost certainly destabilize global markets, and such a resurgence would almost certainly be embedded in either struggles by migrant workers from India/Pakistan/Bangladesh/Philippines, or in the struggles of the marginalized Shia populations of Saudi Arabia and Bahrain (who have historically been the backbone of leftist movements in the area).

In the case of migrant worker organizing, this would mean that radicals in South Asia and the Philippines have a critical role to play.  Revolutionary organizing among the migrant workers of the Gulf will require deep connections with the homelands, and the establishment of some kind of “home bases” away from the ruthless police states of the Gulf.

The supply lines of the Gulf’s repressive apparatus are also a key target for disruption, and arguably a necessary condition for successful communist resurgence.  Much of this apparatus is underwritten by the Western military-industrial complex and related surveillance and security organizations and companies.  In the modern era, many of these surveillance/security companies are integrated with the tech industry.  This gives another front on which radical tech workers can fight on.

And speaking of the tech industry, we can “close the loop” on the above analysis by looking at how many Indians migrate to the US to work in the tech industry (especially its core nexus in the San Francisco Bay Area), including in and around security firms.  Perhaps a connection can be made between these migrants, and the lower-skilled migrants in the Gulf; after all both categories tend to hail from the southern Indian states of Karnataka, Kerala, and Tamil Nadu.  The material and social terrain exists here for a triangle of revolutionary class struggle to be developed between south India, the SF Bay Area, and the Persian Gulf.

This is all of course just one thread in the kind of analysis and strategizing needed to develop an international vision for class struggle.

Data and AI, the core of contemporary capitalism

Last week’s Economist had an excellent report on the data economy and its growing importance to contemporary capitalism.  Data is now the most valuable commodity in the world, ahead of oil, and the five most valuable firms in the world are all technology companies: Alphabet (Google), Amazon, Apple, Facebook, and Microsoft.

It is generally understood that all the free online websites we use aren’t actually “free”, but rather are paid for via data collection on our activities on these sites, which are used for targeted advertisement.  Less understood is the fact that increasingly, the real value is not in data and advertising, but in using the data to improve artificial intelligence programs that are much more valuable and create much more complex services, like language translation and self-driving vehicle algorithms.  Hence, why Tesla is valued so high, despite selling so few cars relative to companies like Ford: the real value is in their massive pool of data that they collect from Tesla drivers, and their increasingly proficient self-driving vehicle AI.

Data-collection and AI has spread across the entire economy, beyond just what we think of as “tech”. Transportation is the obvious sector that is fusing with data/AI, but less obvious sectors include the health care industry, and traditional sectors in manufacturing and industrial processing.

GE, for instance, has developed an “operating system for the industrial internet”, called Predix, to help customers control their machinery. Predix is also a data-collection system: it pools data from devices it is connected to, mixes these with other data, and then trains algorithms that can help improve the operations of a power plant, when to maintain a jet engine before it breaks down and the like.

Of course, creating a system for data acquisition and monitoring is hardly new for plants, its not like you can run a modern plant without some kind of control system; what is actually new here is the application of AI to analyze plant data and suggest improvements that plant operators might not realize could be done.

The finance-data intersection is also important to grasp.  A primary point of discussion in the Economist report was around data markets — specifically, the lack of one.  Despite data being so valuable, there exists no real market for companies to trade data-sets and data-streams; instead, they simply buy up the whole company so that they can silo off data for themselves.  There doesn’t appear to be much enthusiasm from tech companies to create such a market; on the other hand, one would imagine that Wall St. is watering at the mouth at the prospect of financializing data, in the same way that manufactured commodities were in the ’80s and ’90s.  Wall St. already makes use of financial data and AI, to the point where engineers and programmers seem more important than traditional traders, but to bring the data itself into the realm of trading and financial speculation is whole other story.

Lastly, and perhaps most importantly, there is the issue of the amount of control that we average schmucks have over our data.  The report discusses many ideas around how to give us back control, such as tools and regulations that allow us to view and control all the data that is out there about ourselves, and even get compensation for its use.  Amusingly, the report’s final section is titled “Data workers of the world, unite!” and discusses arguments for a “digital labor movement” by those of us (which is basically all of us) whose data is being fed into increasingly powerful AIs, and monopolized by massive and unaccountable corporations.  References are made to Jaron Lanier’s Who Owns the Future?, and the way immense amounts of value is being spun out of day-to-day activities– mostly online, but increasingly offline as well.

This idea is a callback to the Marxist analysis of machines and labor, where technology is said to be “dead labor” insofar as it is a mechanized interpretation of what living labor (the human worker) does.  The power loom took the labor of weavers and interpreted it through a machine, locking in the knowledge and skills of the workers into a system controlled by capitalists.  Machine-tool automation did the same for machinists and metal-workers.  Today, we can imagine this trend reaching toward its logical conclusion, where capital soaks up human knowledge and skill in general and uses it to build AI that could perhaps for people entirely — and beyond.

All of this should emphasize the importance of organizing within the tech industry.  All of these data and AI-centered processes are not driving themselves.  They are dependent on armies of engineers, programmers, scientists, and all the other workers within the tech industry.  They are also increasingly political, even radical.  Contemporary movements to overthrow capitalism will necessarily have to synthesize the workers who are laboring at its core.

David Watkins vs. Peter Frase vs. Norbert Wiener on workers, jobs, and technology

How should workers respond to the introduction of labor-saving technology into their workplaces, and the inevitable and potentially negative labor restructuring that follows?

In the context of automatic self-checkout stations at grocery stores, David Watkins argues that we should cheer on Luddism; that is, workers should resist technological changes that will erode their jobs and wages.  He directly responds to the increasingly popular argument that automating away work would theoretically be a good thing, by pointing out that under capitalism this potential benefit will only ever be theoretical.  Therefore this kind of utopian imagining is not useful for workers whose jobs are being undermined and displaced by technology today, and whose living standards are falling.  Watkins also pushes back against the idea that some jobs are inherently alienating and dehumanizing, and argues that workers have always found value in all kinds of labor, regardless of how menial or grinding — including, presumably, grocery store clerks.

Peter Frase pushes back against both of these arguments.  He argues that while it is important to think of pragmatic and/or defensive strategies like resisting technological restructuring, it is also important to move past a permanently defensive posture and imagine a future where the benefits of automation can actually be distributed to workers, instead of being monopolized by capital.  He points to the example of dockworkers, who over the past half-century accepted dramatic technological restructuring of ports — but also engaged in militant struggles against management so that their high wages were protected, and even secured a certain level of profit redistribution for displaced workers.  Frase also scoffs at Watkins’ argument about workers finding value in all sorts of work, pointing out that this totally misses a basic point about how coercion functions in capitalism — that is, workers don’t have free choice to engage in work they personally find valuable, since they have to find some sort of work in order to survive.  Thus, the whole idea that we ought to prevent technology from disrupting work and allow people choice in the work they pursue is meaningless when people never had a real choice in the first place.

Between these two, I lean much more toward Frase’s position.  I do agree with Watkins’ point about the need to not let utopian theories override what we need to do today to protect our interests and needs as workers.  But I think Frase’s example of how dockworkers approached the issue of technology and job loss/change is a much more compelling alternative to neo-Luddism.  I’d also argue that if we were to accept Watkins position that technology restructuring needs to be resisted, then we’re going to need a powerful and rejuvenated radical left — and a radical left that is powerful enough to halt technological restructuring is also powerful enough to coopt and deploy technology on workers’ terms.  So why not just go with the latter?

Practically speaking, local conditions will be the ultimate determinant of whether workers and leftists decide to either resist or coopt technology.  In some cases, it’ll be easier to smash up and sabotage robots, surveillance systems, etc.; in other cases, it’ll be feasible to force management to cut into capital’s share of profits and divert it to workers, or even to research, develop, and deploy new tech on workers’ terms.

And it is worth remembering that this debate is not new.  Norbert Wiener, an MIT professor widely viewed as the “father of cybernetics”, raised the resistance vs. cooption dichotomy all the way back in 1949, in a letter he wrote to the president of the United Auto Workers (UAW) on the dangers of automation to organized labor:

What I am proposing is this. First, that you show a sufficient interest in the very pressing menace of the large-scale replacement of labor by machine on the level not of energy, but of judgment, to be willing to formulate a policy towards this problem. In particular, I do not think it would be at all foolish for you to steal a march upon the existing industrial corporations in this matter; and while taking a part in production of such machines to secure the profits in them to an organization dedicated to the benefit of labor. It may be on the other hand, that you think the complete suppression (sic) of these ideas is in order. In either case, I am willing to back you loyally, and without any demand or request for personal returns in what I consider will be a matter of public policy.

At the end of the day, intellectualizing about what machines could do must take a back seat to political practice and concrete movements, and how we view technology must be subordinate to strategies grounded by on-the-ground reality.