Syndicate content

When Robots Attack!

Tanya Gupta's picture

Robots have been a part of our mythology for thousands of years, the emphasis alternating between their positive transformative power over human society and acting as agents of great destruction.  Our image of robots has been shaped to a large extent by Hollywood and literature.  Celluloid robots in Star Wars, 2001 Space Odyssey, Robocop, Star Trek and many of Isaac Asimov’s novels have become a part of the human story.  Off-celluloid, robots have been helping our society in concrete ways (for example police work (bomb disposal), infrastructure projects etc.).  However when Watson won Jeopardy it brought artificial intelligence and robotics a new kind of attention.  People started to wonder if robots could replace humans.  When we think of robots we think of self driven cars, household robots or even warrior robots.  However, in our view, the influence of robots and Artificial Intelligence (AI) is more subtle and their presence more ubiquitous than one would think. One such impacted sector is the agriculture sector (in the US) which is on the cusp of a massive transformation, as it moves from mechanization to automation. When rolled out and commercialized (soon) this massive scale of automation will have a significant impact on US farming and on immigration for sure.  But does this also impact the development landscape? If so how?

Agricultural robotic systems have been implemented in fruit and vegetable harvesting, greenhouses and nurseries. Harvest Automation, for example, has developed the the HV-100, a 90-pound robot for commercial nurseries that can pick up and rearrange potted plants. There are quite a few silicon valley startups that are contributing to this revolution in the region known as “America’s Salad Bowl”, around Salinas Valley. California, where Salinas Valley is located, produced $1.6 billion dollars worth of lettuce in 2010 and 70%+ of all lettuce grown in America. Lettuce Bot, a new robot developed by Stanford engineers Jorge Peraud and Lee Redden, both from farming families from Peru and Nebraska, can “produce more lettuce plants than doing it any other way” (Yahoo Finance).  Lettuce Bot’s innovation is that while attached to a tractor, it takes pictures of passing plants and compares these to a database. When the weed or a lettuce head that is too close to another one is identified, a concentrated dose of fertiliser is sprayed. A close shot of fertilizer kills the errant weed or lettuce head but actually feeds the further off crops at the same time.

Lettuce Bot

Similar companies include San Diego-based Vision Robotics, and Agrobot.  In Europe, a new project that is a part of the European Union Seventh Framework Program (FP7) cRops (Clever Robots for Crops) is focusing on creating robots to harvest high value crops for site-specific spraying, and selective harvesting of fruit. Robots are at the cusp of changing not just the entire farming sector, but also other sectors. In fact, Moshe Vardi, Professor of Computer Science at Rice University argues that "by 2045 machines will be able to do if not any work that humans can do, then a very significant fraction of the work that humans can do".  If this is true, the world will be significantly changed.  What does this mean for those of us working in development?  How can we start to respond now to this changing world?  What will it mean for education and training if millions of workers are displaced by robotics and automation?
Here are at least two ways that development organizations can start to respond effectively and quickly to the flash-flood style technological changes that are transforming key sectors radically:

  • Cross sectoral studies: what will robotics and automation of sectors such as agriculture mean for key sectors and issues such as migration? The answer to this question will not be given by any single sector.  To understand the dynamics underlying this question (technology, robotics, sector speciality, social sciences, immigration), development organizations need to leverage partnerships and multi sector data collection and analysis.  One organization that is doing this is the Global Knowledge Partnership on Migration and Development (KNOMAD).  KNOMAD is based on multidisciplinary knowledge and evidence collection and is a collaboration between the World Bank, the Swiss Agency for Development and Cooperation (SDC) and the German Federal Ministry of Economic Cooperation and Development (BMZ).  More such partnerships should be developed and leveraged to explore issues such as the following:
    • Impact of robots on key sectors, migration and remittances:  In the US for example 72% of farm workers are foreign born. What if they are replaced by robots ? What will it mean, for example for the Mexican economy, a source of 68% of US farmworkers? What will it mean if the flow of unskilled labor coming from Mexico, Central America, and the Caribbean to the US (38% of immigration) stops ? What similar patterns exist in other countries? (interestingly, this will not be the first time that technology and immigration may have contributed to a fundamental shift in the agriculture sector.  In the 1870s-1890s, the growth in agriculture and the financial success of Salinas Valley was linked to both the Southern Pacific Railroad that came to Salinas in November of 1872 as well as the contributions of Chinese and Japanese labor labor).
    • Remittances to developing countries are estimated to reach $372 billion in 2011. Many of these remittances are earned from sectors that are vulnerable to automation.  What if a bulk of these remittances were to dry out as a result ?  What impact would it have on the global economy and consequently on development?
    • Helping countries with robotics-triggered changes: can development organizations help countries assess how to leverage the promise of robotics and other technologies? Can they assist on how to cope with this kind of creative destruction and change, and help them plan for a changed future, potentially in the process also enabling synergies across countries (such as South-South learning) ?
  • Include technology focused foresight as a core development function:  In future studies, the term "foresight" has become widely used to describe activities such as: “critical thinking concerning long-term developments, debate and effort to create wider participatory democracy, shaping the future, especially by influencing public policy”.  Public sector foresight, as Nancy Donovan, U.S. GAO's Applied Research & Methods team (and co-founder of Public Sector Foresight Network along with Dr. Clem Bezold), says, can involve activities such as “recognizing the long-term implications of today's decisions; identifying key trends and opportunities as well as emerging challenges before they reach crisis proportions; and informing government's future role and responsibilities”. As Nancy notes, “the rapid pace of "high clockspeed”(pdf) technologies pose particular challenges to government organizations”.  There is a need for an equally rigorous and integrated review of the development sector.  In particular it would be useful to look at the information that foresight exercises, supplemented by big data, can provide in terms of the development pivots that need to be taken. (pivot of the lean startup model defined as a “structured course correction designed to test a new fundamental hypothesis about the product, strategy, and engine of growth”  as applied to development)

While robots are not bringing about an equal society as Aristotle thought, nor are they taking over the world, robots are indeed engaging in the creative destruction of jobs in many sectors including agriculture.  Using their large data processing capabilities and in some cases mimicking human actions, robots are contributing to a world that may soon be automated.  We, in development, can serve our client countries well by looking more systematically at the impact of robotics on development and how we can respond to this challenge.

Photo Courtesy: Victor Habbick/freedigitalphotos
Follow PublicSphereWB on Twitter


Submitted by Michal on

Thank you for a very interesting and thought-provoking article! Reading about your first set of recommendations regarding the cross-sectoral study of consequences of mechanization/robotization of agricultural sector, a classical anthropology study came to my mind. It's Scott's "Weapons of the Weak" who wrote about the impact of market and mechanization on peasants in Malaysia. He shows how these new forms of labor transform organization of societies and local communities with potential social clashes emerging therefrom.
Transformation of labor in agriculture is indeed most significant and ought to be among issues of primary interest of decision makers precisely due to consequences you put forward in your article. Sadly, it's not the case yet.

Michal we appreciate your thoughtful comment. Yes, the transformation of labor in agriculture should be a key discussion item in national (and international) policy dialog. Thanks also for the intro to James Scott (who interestingly enough interpreted anarchism as the "everyday resistance to the rule of technocratic elites" thus pointing to the futility of top down planning) As you very well know long term (and subtle) consequences are sometimes overlooked in national policy discussion in favor of issues which generate more political buzz. For example, a national level project on “digitizing” county x (of course without really understanding what that means) will garner a whole lot more attention from policy makers in most developing countries.

To counter this, development organizations have to do a better job in establishing a “deep” partnership with countries as they jointly develop assistance strategy. This deep partnership must be focused on knowledge sharing (with sectoral loans and assistance taking a back seat). We must engage with the country partners early on and develop a collective understanding of the profound consequences of these subtle changes in sometimes overlooked sectors

Submitted by Daniel on

This is a very interesting article and as someone who works on rural development in the EU pre accession sphere, I think it poses very pertinent questions. What will the primary sources of income be for rural dwellers when mechanization and automation replaces much of human of farm labor? Europe and the US have become susbtantially mechanised/automated, and employment in agriculture as a consequence has reduced to a few percent of the active population. The technology revolution is capital intensive and challenging to keep up with as a small farm. Looking at this, the question I ask myself is not how we will feed ourselves in the future, I think the technology to do that exists today already, but how will we keep our rural spaces as vibrant connected parts of the national economy? In some richer countries subsidies increasingly aim at maintaining the rural space connected with urban activity and payments that focus not only on farming but the overall quality of life, ensuring similar availability of services, diversification of economic activity, and sustinable environmental management. But what of poor countries where farming employs as much as a fifth or even nearly half of the active population working on tiny farms mostly for subsistence?
Avoiding mechanization/automation is not really an option. Foreseably there will be 9 billion people to feed by the end of the century, technological advances will be the primary factor that will allow us to do so. In addition, sustainably feeding the world also means to keep maintaining environmentally sustainable spaces for our fellow living creatures of this earth (animals and plants). Policy makers need to think of the rural space today and develop strategies that look at rural areas beyond simply farming. These strategies need to cut across sectors, with the aim of maintaining a relatively even quality of life and availability of services in a future of fewer but larger farms and agroprocessors that will need fewer working hands. The models we see so far in many developing countries where rural areas largely depopulate resulting in the emergence of huge underserved slums with high unemployment and often abject living conditions can hardly be the solution of the future.

Daniel thanks for a very thought provoking and thoughtful comment. You have highlighted some excellent points.

Overall -- the development and deployment of advanced technology is both inevitable (market forces will opt for efficiency) and desirable (we will have 9 billion people to feed). So the Heideggerian option of the destruction of modern technology and a return to an earlier agrarian world is neither possible nor feasible
More specifically there will be some negative externalities, particularly the destructive impact of large scale mechanization and automation on dispersed and small scale subsistence farming
The disappearance of subsistence farming will lead to urban migration, which in turn will lower the quality of both urban life (resource depletion) and rural life (not enough labor).

The unprecedented pace of technology diffusion and its impact on market and social forces are reshaping sectoral boundaries. The resulting obsolescence of the models we we use in understanding the world are leading to an unfilled need for newer models. It also leads to technology avoidance. Rather than a Luddite development response to a world being re-shaped by technology, we need a compelling symbiotic planned engagement between the two. This engagement/relationship cannot be in abstract terms but using concrete on the ground examples and invoking precisely the kind of engagement we just had.

Submitted by Anonymous on

Dear Abir and Tanya, thank you for your replies. I do not want to get esoteric, but - as we are talking about future scenarios - let me question two of your suggestions; 1) the inevitability of market forces that will decide about the future of agriculture, 2) the necessity of large-scale farming to feed everyone.
My basic idea is that agriculture is related to the soil, which has powerful symbolic meanings attached to it. The soil symbolically represents one's belonging, one's identity and it has very strong roots in narratives of nations, tribes and other social formations. Furthermore, due to historical, generational attachment of collectivities to the particular soil (the 'motherland'), specific cultures including labor divisions has been created in these collectivities. Disrupting the relations attached to and managed within the cycle of food production may cause detrimental ruptures in societies. I'm not saying we should not make our agriculture more efficient by introducing robotization into farming (we have already done that), but I'm thinking about the extent that it is necessary and - what is more important - how to cope with what you termed 'externalities', i.e. mass migration, rising unemployment, rising urban poverty, creation of slums, etc. that are consequences of increased mechanization in rural agriculture.
This question is of course not merely economic/political, but ethical as well; given the statistics on the average nutrition intake of people in Western and non-Western countries (or countries of the Global South and North). What happens very often is that we in the North make agriculture more efficient, so that we can eat much more than we actually need. This is of course not a good model. Making agriculture more efficient is crucial, but this cannot be done if we are in the same time producing mass rural-urban migration that further leads to other problems. To find the middle ground, the correct relation between development and tradition is, I believe, the most crucial and difficult task.
Btw. Heidegger did not argue for destructing modern technology, but rather opted for using it in such way that would dismantle the Cartesian subject-object dichotomy, since he understood ourselves to be in intimate relationship with the world that surrounds us. But that's just a side note.

Submitted by Claudia on

Tanya and Abir - thanks for an interesting post. Re - assessing how to leverage the promise of robotics and other technologies - it might be worthwhile to look at a STEPS Centre working paper on new models of technology assessment, available here

Add new comment