In the latest edition of the Expert Series, RAC member, Dr Jeanette Rowley, comments on the stated duty to consider nonhuman animals in our development of trustworthy and responsible artificial intelligence.
This article comments on the stated duty to consider nonhuman animals in our development of trustworthy and responsible artificial intelligence. It explains that European ethics guidelines for the development of trustworthy and responsible AI include consideration for nonhuman animals because they can be harmed by the development and use of autonomous artificial intelligence. It argues, however, that the commodity status of nonhuman animals betrays the stated duty to recognise them as benefactors of institutional moral concern. It explains the impact of the data economy on nonhuman animals in terms of data collection and data-driven artificial intelligence in Precision Livestock Farming (PLF) and argues that the animals subjected to this ‘livestock production process’ are not animals for the regulation of artificial intelligence to avoid harm.
The article explains that the inclusion of nonhuman animals in guidelines for the responsible development of AI recognises our moral concern for other animals, but that the use of artificial intelligence in Precision Livestock Farming denies some animals their legitimate stakeholder status for the purposes of developing and utilising ethical artificial intelligence. The article concludes, therefore, that Precision Livestock Farming further objectifies some nonhuman animals.
An extended introduction explains the data-driven economy, existing concerns and proposed regulation, to foreground a discussion of how data collection and the use of artificial intelligence impacts upon nonhuman animals in Precision Livestock Farming.
1. The data economy
In 2017, The Economist reported that the most highly valued commodity was no longer oil but electronic data.[1] The report is, perhaps, unsurprising following a claim in 2014 that private companies identify, use and sell up to 75,000 separate data points for every individual consumer.[2] This new ‘data economy’ is driven by our use of smartphones, the internet and the ‘Internet of Things’ (IoT): connected gadgets used for our pleasure and convenience, but which register a digital trace of every aspect of our lives. What we watch on TV, what we buy, read and eat, places we visit, how long we stayed, how much electricity we use and what we use it for, are bits of data, among others, held by machines which then predict our preferences, our choices, and when and where we are likely to exercise them. These 75,000 data points are a digital representation of individual lives: points of reference to our personalities, our characters and how we act out our personal and practical life circumstances. All this data is mined, processed, used overtly for the purposes of targeted consumer marketing, and exists for other unknown, potentially covert purposes.
The many thousands of data points are extracted from initial raw data through the use of a branch of artificial intelligence known as machine learning. In this process, simple data points are processed, by machines, to produce ‘big data’, which refers to the way computers capture and process large volumes of various types of data to infer and detect patterns and make predictions. For example, algorithms can be used to predict our buying patterns, when we are likely to visit the supermarket, what we might want to watch on TV, places we might like to visit, and if we are likely to suffer illness. This use of artificial intelligence is driving a new era of commercial enterprise in which algorithms are also used by businesses to enhance efficiency and reduce waste, but also to automate decision-making processes.
The new data economy is widely acknowledged as immensely profitable, but it can affect consumers and result in unfair practices. For example, if a person applies for a credit account and a computer decides to refuse the application based on disqualifying but out-of-date personal data about the applicant. The new era of data collection, the Internet of Things and the applied machine learning that underpins predictive technologies have advanced rapidly in an absence of general ethical guidelines, despite generating important questions about consent and privacy, bias and discrimination, potential loss of autonomy, disrespect for rights and the possibility of unfair outcomes.[3]
2. The development of social responsibility for the development of AI
It is no surprise, therefore, that in the rapid growth of the now advanced data economy that automatically strips every individual of the right to privacy, there has developed social awareness and concern with data collection, its relationship with machine learning and the role and accountability of those developing profit-focused artificial intelligence. In 2014, the Future of Life Institute was formed to promote the need for safety research in the development of AI and make research mainstream.[4] By 2016, six of the world’s biggest technology companies, Google, Facebook, IBM, Microsoft, Amazon, Apple and DeepMind, joined forces to consider how best to develop and use AI,[5] now reassuringly presenting their portfolio for best practice in six socially orientated ‘Thematic Pillars’.[6] Around the same time, various regional organisations and charities started to consider the commercial utility of data-driven artificial intelligence and ethical issues: one of which in the United Kingdom is the Alan Turing Institute, which, from 2017, started to examine the social implications, ethics and legal issues in relation to the development and use of AI.[7] The overriding concern is that our development of AI should be used for social good and not lead to the creation of autonomous artificial intelligence that causes harm.
European Union (EU) legislation on the processing of personal data has been in place since 1995 in the form of Directive 95/46/EC, which attempted to balance the commercial benefits of free-flowing personal data with privacy rights. The recent EU General Data Protection Regulations are the latest in a series of interventions aimed at protecting consumers.[8] According to Jacob Turner, governments have not made much effort to regulate the commercial development and use of data-driven artificial intelligence,[9] but there does now appear to be some movement in this area. Following setting out its vision in 2018 for the development of AI, the European Commission (EC) has now published its ‘Ethics Guidelines for Trustworthy Artificial Intelligence’ (EC Guide)[10] and is currently piloting its recommendations for the responsible development of AI.
The EC Guide looks to human rights standards in its development of an ethical framework for AI and is produced by DG Connect, which is the Commission department responsible for developing a digital single market to generate smart, sustainable and inclusive growth in Europe.[11] It specifically aims to support and develop the new digital economy while addressing its ethics guidelines to all ‘stakeholders’ and aiming to protect ‘vulnerable groups’. The ethics guide for the development of responsible artificial intelligence advises that it should be lawful, ethical and robust to avoid causing harm, and, critical for the purposes of this discussion, it mandates that the development of responsible artificial intelligence will consider the ‘potential impact or safety risk to the environment and to animals’. The Commission and the 52 appointees to the High-Level Expert Group on Artificial Intelligence, however, ignore the way in which data collection and artificial intelligence already impact on the lives of nonhuman beings to the point at which they are, arguably, no more than biological components in an automated, intensive, data-driven commodity production process that has the potential to control their entire lives, from the selection of their DNA to when machines in the system decide it is time for them to be killed.[12]
3. The development of data-driven technology in animal farming
Laboratory research on the development and use of data-driven technology in animal farming dates back to the early 1990s,[13] robotic milkers have been around since the mid-1990s[14] and the significant commercial benefits of Precision Livestock Farming have been promoted from at least 2003.[15] Daniel Berckmans notes that one of the early objectives of PLF was to develop online monitoring tools to support farmers and that emerging technologies offered interesting possibilities for the automatic control and management of the complex biological processes of animals.[16] Berckmans promotes PLF as innovative technology that offers endless possibilities for the automatic control of animals in the livestock production process, including welfare advantages and non-tangible benefits for farmers, such as the opportunity for a better quality of family life as a result of a less labour-intensive occupation.
Acknowledging that nonhuman animals used in farming are complex and individually different, Berckmans promotes PLF as an approach that can optimise livestock production because it uses data collection and prediction technology to maximise the commodity value of every individual animal. He explains that in PLF, animal variables, such as their weight, activity, behaviour, drinking, feeding, sounds they make, body temperature, respiration rate and blood variables, can be continuously measured, information analysed and reliable predictions made about how a given animal will respond to environmental changes. Algorithms in microchips can use this data to monitor and manage the animals automatically.
The development of a single European digital economy includes all financially profitable sectors, including farming nonhuman animals. Roberto Viola, the Director General of DG Connect,[17] blogs about his excitement for ‘new fields of operation’ for the future of farming. Viola states that “[t]echnologies such as artificial intelligence, robotics, blockchain, high performance computing, the Internet of Things and 5G have the potential to make farming more efficient and productive.”[18] What Viola refers to is the optimization of digital technologies in farming to the extent that farmers can revolutionise all aspects of the work they do and will only need to monitor automated processes using remote devices.
Some of the current operations Viola is excited about are, among others, the ‘Happy Cow’ project, ‘Pig Farm Management’, ‘Poultry Chain Management’ and ‘Decision-making Optimisation in Beef Supply Chain’.[19] The ‘Happy Cow’ project is designed to improve dairy farm productivity through the Internet of Things and machine learning technologies. In 2015 data collected by robotic milkers was said to be “staggering”, and included information from “protein-fat ratio in the milk to white blood cell counts”, but that was regarded to be just “the tip of the iceberg”.[20] The ‘Happy Cow’ project now promises to optimise feed intake and decrease the in-between calving time of cows by combining real‐time sensor data, gathered from neck collars with GPS, with machine learning technologies and cloud-based services. The stated objective is to create more value in the dairy chain.
The use of similar technologies in ‘Pig Farm Management’ will optimise pig production management through the use of interoperable on-farm sensors and slaughterhouse data.
The ‘Poultry Chain Management’ project relies on IoT technology, with data collected on a central cloud-based platform linked to a Smart Data-Analytics tool. The main objective is to optimise the production of chickens to reach a desired and accurate end weight.
The ‘Decision-making Optimisation in Beef Supply Chain’ project utilises a technological framework based on IoT, Open Data, Big Data and Blockchain to (among other objectives) increase the reproduction rate by 90%.
4. The extent and use of data collection in Precision Livestock Farming technologies
Precision Livestock Farming, then, is a method of processing nonhuman animals that relies on various types of data fed into, and generated by, computers. Connected data technologies involved in PLF include, but are not limited to, a variety of sensors, cameras, microphones, satellite positioning systems, automated robots, such as robotic milkers,[21] respiration monitors and radio frequency ID ear tags, all of which provide information on the entirety of an animal’s existence: their feeding habits, weight gain, movements, blood pressure, temperature, heart rate, milk production, meat composition and quality; while internet connections, wireless communication tools and cloud storage offer remote monitoring.[22] The use of technology and artificial intelligence can be applied to animals throughout their lives from the selection of their DNA[23] to deciding the most profitable time for them to be killed, all with minimal human intervention, in an optimised, lean production process to maximise profit.[24]
Operational systems can be small or large scale but necessarily depend on forensic observation, continuous monitoring, data mining and the use of algorithms to make predictions, choices and decisions about how the automated systems should function and respond. These technologies are already inspiring bespoke engineering design services for the development of ‘smart farms’.[25] Further, data obtained from nonhuman animals locked into the PLF system is shared beyond the confines of the site of production through ‘datahubs’, allowing farmers to share data throughout the supply chain.[26] The data collected and shared, however, is not recognised as that of the animals but that of the farmer. Stokkermans states that “[t]he farmer determines who is allowed to use his data and for what…”[27]
Precision Livestock Farming is, therefore, correctly described as a data and information-intensive method of livestock production.[28] It places personal data, obtained from nonhuman animals, at the centre of a process for optimised production of their commodified bodies.
5. Harm to animals in the PLF system
Since its inception, PLF has been promoted as beneficial for both nonhuman animals and the environment because it is said to offer enhanced animal welfare benefits and be less harmful to the environment than conventional farming methods. Indeed, the literature and projects cited above mention general and specific benefits in varying degrees, such as early detection of lameness in cows[29] and ‘swine cough’ in barn-housed pigs.[30] There are, however, serious risks associated with the use of artificial intelligence in farming that are less well documented. These risks indicate that nonhuman animals are in danger of suffering direct harm from AI.
All technology employed in livestock production can malfunction. Sensors, remote controllers and other devices can fail and feed incorrect data into automated systems. For example, a remote feeding sensor could indicate that animals have been fed when in fact they have not; robotic milkers, which identify the milking patterns required for individual cows, could malfunction and deny a milk-heavy cow a milking session; the laser beam used by the robot to line up the milking apparatus could fail, and information collected about the quality of milk or health of the cow could be inaccurate.[31] Any of the algorithms used in machine learning can make mistakes and make bad predictions or decisions, such as reporting healthy animals as unhealthy or unhealthy animals as healthy, and then take inappropriate action based on those false decisions. Cyber-attacks, natural disasters and malicious alteration of algorithms are all possibilities that could result in nonhuman animals suffering from the development and application of data-driven artificial intelligence.[32]
Clearly, these threats, inherent in the PLF system, could potentially harm animals and are serious concerns in themselves, but they are not the primary focus of this discussion. The primary concern of this discussion is the permissibility of the development and use of various forms of artificial intelligence that are programmed to optimise a commercial process of giving life and bringing about profitable death. In the PLF system, harm to animals is the specific intention of the development and use of AI despite a mandatory requirement for the development of responsible artificial intelligence to avoid harm to people, animals or the environment. The potential for harm to animals as a result of technological or man-made failures in the PLF system is, however, tertiary: I want to argue that the primary harm of PLF is the additional objectification of the species of nonhuman animals categorised as livestock.
6. The objectification of nonhuman animals
To reiterate, PLF is defined as “the management of livestock production using the principles and technology of process engineering”.[33] This process relies on the collection of data taken from animals, some of which is fed into machines, for ‘learning’ calculating, making predictions and taking decisive action. It is a system that exists in a society that specifically mandates a culture of ethics in the development of data collection and artificial intelligence to prevent harm to humans, animals and the environment. Precision Livestock Farming is, therefore, a system in which nonhuman animals are described as absent referents because in both the definition and system of PLF, the individual, literal living, sentient being has disappeared and they are treated as, and turned into, objects. [34]
The objectification of nonhuman animals is not, of course, brought about specifically by Precision Livestock Farming. In traditional farming, the animals designated ‘livestock’ are already turned into, and treated as, objects, and are historically objectified[35] because they are denied their subjectivity, the self-authoring of their own lives in their own environments, and are invisible components of an imposed system and property owned as commodities.[36]
7. PLF and further objectification of nonhuman animals
Whether PLF imposes additional objectification on animals designated livestock is questioned in recent literature because it is proposed that “[o]ne way of viewing PLF is … to see it as a ground for improving the wellbeing of animals, because the technology highlights the care for the individual animals and their quality of life.”[37] Adopting an ethics of care approach, this view argues that Precision Livestock Farming disrupts the traditional debate on the objectification of animals in the livestock business because “PLF redefines the notion of care, in terms of data transparency, standardisation of methods for analysis, real-time collection and processing of data, remote control, and the use of digital platforms. Therefore PLF requires a redistribution of responsibilities within a wider scope of relations in the value chain.”[38] This redistribution of responsibility is also said to include an obligation to consider the way farmers themselves feature in the new PLF paradigm of processing nonhuman animals as commodities.[39] The suggestion is, thus, to regard Precision Livestock Farming as creating opportunities for new types of discussions, including the potential objectification of farmers who “subjected to the PLF system are reduced to being a part of an encompassing farm management system that is embedded in markets and agricultural policy.”[40]
The overriding notions of welfare and care in PLF currently only relate to their utility in optimising the production process.[41] That the use of AI will produce better welfare insights glosses over the underlying problem that the moral duty to animals, as recognised in the EC Guide for the ethical and responsible development of AI, does not apply to PLF, and, further, that the development and use of artificial intelligence to process (some) nonhuman animals is enthusiastically encouraged as a profitable commercial sector.[42] Using an ethics of care approach to commodified sentient beings implies that the duty to consider the impact of AI on nonhuman animals is automatically imperfect. But given that it is well established that animals matter morally, as is also recognised in the EC Guide, then the stated duty, which recognises we must also take care to ensure our developing data and AI systems do not harm animals, cannot be imperfect.
Precision Livestock Farming, despite its claims for better welfare outcomes and the existence of literature suggesting how we might redefine the exploitation of animals in this process, further objectifies these animals because they are denied their animal stakeholder status in the framework for protection from harm inflicted by autonomous artificial intelligence. Their exclusion from this belated, but essential, moral initiative for AI that does no harm and is utilised for social good betrays our recognised moral concern for animals and further objectifies them because their right to exist as legitimate animals has now also disappeared.
8. Conclusion
The data economy and development of associated artificial intelligence are pervasive. They are the new driving force of the global economy and include all sectors. In this new era, increasingly referred to as the fourth industrial revolution (industry 4.0), people are raising concerns about data collection and their human rights. Ethical guidelines state that we have a duty to control artificial intelligence, ensure that we use it for social good and, also, take into consideration animals and the environment. However, the European Commission and the 52 appointees to the High-Level Expert Group on Artificial Intelligence overlook a serious issue of justice because they fail to consider the ethical issues in the development of artificial intelligence in Precision Livestock Farming that seeks specifically to gather and exploit the data of nonhuman animals, monitor and assess their biology, their optimum condition for slaughter and, ultimately, ‘decide’ when it is time for them to die. Perhaps this oversight is not surprising, given the fact that, historically, considerations for the ethical development of AI have been (perhaps inadvertently) humancentric[43] and current EC ethics guidance for the development of AI are grounded by human rights and inadequately address our general and specific duties to nonhuman animals in the development of responsible AI.
The EC’s framework for ethics to govern artificial intelligence is, therefore, fundamentally flawed to a point of injustice. While it registers our moral obligation to consider how we might harm nonhuman animals in the development of artificial intelligent technologies that drive a new digital economy, it is produced by a Commission department that emphasises only human rights and specifically encourages digitalisation and the use of artificial intelligence in nonhuman animal farming. DG Connect has a clear and detailed understanding of how artificial intelligence is applicable to the animal farming sector, but does not include these animals as relevant, vulnerable stakeholder subjects in the creation of a safe digital AI economy. This dichotomy further reduces animals as objects because they are denied their natural status as animals for the purposes of developing and regulating AI that does no harm.
The views expressed by our Research News contributors are not necessarily the views of The Vegan Society.
[1] Leaders. (2017, May 6th). ‘Regulating the internet giants: The world’s most valuable resource is no longer oil, but data’. The Economist. Retrieved 20-09-2019 from https://www.economist.com/leaders/2017/05/06/the-worlds-most-valuable-resource-is-no-longer-oil-but-data
[2] Simon, S. & Gerstein, J. (2014, May 14th). ‘What the government isn't doing’. Politico. Retrieved 20-09-2019 from https://www.politico.com/story/2014/05/big-data-beyond-the-nsa-106653
[3] Leslie, D. (2019). ‘Understanding artificial intelligence ethics and safety: A guide for the responsible design and implementation of AI systems in the public sector’. The Alan Turing Institute. https://doi.org/10.5281/zenodo.3240529 Retrieved 20-09-2019 from https://www.turing.ac.uk/sites/default/files/2019-06/understanding_artificial_intelligence_ethics_and_safety.pdf
[4] Tegmark, M. (2018). Life 3.0: Being human in the age of Artificial Intelligence. UK: Penguin.
[5] See Turner, J. (2019). Robot rules: Regulating artificial intelligence. Switzerland: Palgrave Macmillan, p. 209.
[6] Partnership on AI. (2016–2018). Retrieved 23-09-2019 from https://www.partnershiponai.org/about/ This particular alliance now lists over 90 partners in 13 countries, a large proportion of which are not for profit. See https://www.partnershiponai.org/partners/
[7] See also, the UK government’s ‘Data Trust’ intentions at https://www.gov.uk/government/publications/artificial-intelligence-sector-deal/ai-sector-deal#key-commitments and Machine Intelligence Garage, ‘Ethics Framework’ at https://www.migarage.ai/ethics-framework/ Retrieved 20-09-2019.
[8] Regulation (EU) 2016/679 of The European Parliament and of the Council of 27 April 2016 on the protection of natural persons with regard to the processing of personal data and on the free movement of such data, and repealing Directive 95/46/EC (General Data Protection Regulation). Available at https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=OJ:L:2016:119:TOC
[9] Turner, n. 5 above.
[10] European Commission. (2019). ‘Ethics guidelines for trustworthy AI’. Retrieved 20-09-2019 from https://ec.europa.eu/digital-single-market/en/news/ethics-guidelines-trustworthy-ai
[11] European Commission. https://ec.europa.eu/info/departments/communications-networks-content-and-technology_en
[12] For a discussion about the invisibility of nonhuman animals in intensive animal production, see Lynning Harfeld, J., Cornou, C., Kornum, A., and Gjerris, M. (2016). ‘Seeing the Animal: On the Ethical Implications of De-animalization in Intensive Animal Production Systems’. Journal of Agricultural and Environmental Ethics, 29(3), pp. 407–423. For a concise industry overview of technology used in the breeding selection process, see Bourdon, R. (2019). ‘Cattle breeding technologies in perspective’. Beef: Farm Progress America, available at https://www.beefmagazine.com/mag/beef_cattle_breeding_technologies All Retrieved 20-09-2019.
[13] Berckmans, D. (2018). ‘The Importance of Animal Welfare: European precision sustainable livestock farming’. Open Access Government https://www.openaccessgovernment.org/the-importance-of-animal-welfare-european-precision-and-sustainable-livestock-farming/45158/ Retrieved 20-09-2019.
[14] Heyden, T. (2015, 7th May). ‘The cows that queue up for themselves’. BBC News available at www.bbc.co.uk/news/magazine-32610257 Retrieved 20-09-2019.
[15] See Cox, S. (2003). Precision Livestock Farming. Available at https://doi.org/10.3920/978-90-8686-515-4. For a short, concise explanation of PLF, see Bartzanas, T., Amon, B., Calvat, S., Mele, M., Morgavi, D., Norton, T., Yanez-Ruiz, D. & Vandongen, C. (2017). ‘EIP-AGRI Focus Group, Reducing livestock emissions from Cattle farming, Mini-paper — Precision Livestock Farming’. EIP-Agri, Agriculture and Innovation, European Commission, available at https://ec.europa.eu/eip/agriculture/sites/agri-eip/files/fg18_mp_precision_livestock_farming_2017_en.pdf All retrieved 20-09-2019.
[16] See Berckmans D. (2004). ‘Automatic On-Line Monitoring of Animals By Precision Livestock, International Society for Animal Hygiène Saint-Malo. Retrieved 23-09-2019 from https://www.isah-soc.org/userfiles/downloads/symposiums/2004/Berckmans.pdf
[17] Note that this is also the department responsible for producing a Europe-wide ethical standard for the development of artificial intelligence.
[18] Viola, R. (2019, April 12th). ‘The future of farming: how digital can make a difference’. Retrieved 21-09-2019 from https://ec.europa.eu/digital-single-market/en/blogposts/future-farming-how-digital-can-make-difference
[19] For information on all examples cited in this section, see Internet of Food and Farm 2020 and embedded page links at https://www.iof2020.eu/ Retrieved 20-09-2019.
[20] Heyden, n. 14 above.
[21] For a description, see Postma, R. (2012). ‘A healthier cow for one cent more’. NRC.NL. Retrieved 20-09-2019 from https://www.nrc.nl/nieuws/2012/02/17/een-gezondere-koe-voor-een-cent-meer-12167524-a1114151
[22] See Bartzanas, T., Amon, B., Calvat, S., Mele, M., Morgavi, D., Norton, T., Yanez-Ruiz, D. and Vandongen, C. ‘EIP-AGRI Focus Group Reducing livestock emissions from Cattle farming, Mini-paper — Precision Livestock Farming’. EIP-Agri, Agriculture and Innovation, European Commission 2017. Retrieved 20-09-2019 from https://www.nrc.nl/nieuws/2012/02/17/een-gezondere-koe-voor-een-cent-meer-12167524-a1114151
[23] Ishmael, W. (2019). ‘Cattle breeding technologies in perspective’. Beef. Retrieved 20-09-2019 from https://www.beefmagazine.com/mag/beef_cattle_breeding_technologies
[24] See, for example, http://smartcattle.net/; Coffeen, P. (ed.) (2019). ‘Big data, big opportunities: How artificial intelligence is transforming dairy farming’. Progressive Dairy. Retrieved 20-09-2019 from https://www.progressivedairy.com/news/event-coverage/big-data-big-opportunities-how-artificial-intelligence-is-transforming-dairy-farming, and Tournier, B. (2017). ‘Tracking devices for livestock increase farm profits’. Sierra Wireless. Retrieved 20-09-2019 from https://www.sierrawireless.com/iot-blog/iot-blog/2017/10/tracking_devices_for_livestock_increase_farm-_profits/ and Maltz E., Livshin N., Antler A., Edan Y., Matza S., Antman A. (2003). ‘Variable milking frequency in large dairies: performance and economic analysis - models and experiments’. In Cox, S. Precision Livestock Farming. Available at https://doi.org/10.3920/978-90-8686-515-4
[25] Smart Cattle. http://smartcattle.net/ Retrieved 20-09-2019.
[26] Stokkermans, P. (2017). ‘Data hub improves digital traffic for dairy farmers’ Nieuwwe Oogst. Retrieved 20-09-2019 from https://www.nieuweoogst.nl/nieuws/2017/05/16/datahub-verbetert-digitaal-verkeer-melkveehouder
[27] ibid, Stokkermans, n. 26 above.
[28] Bartzanas et al, n. 22 above.
[29] Alsaaod, M., Fadul, M., & Steiner, A. (2019). ‘Automatic lameness detection in cattle’. The Veterinary Journal, 246, pp. 35–44. Available at sciencedirect.com/science/article/pii/S109002331830234X Retrieved 20-09-2019.
[30] Shike, J. (2017, 17th April). ‘Swine cough monitoring technology offers early detection and treatment’. Farm Journals Pork. Available at www.porkbusiness.com/article/swine-cough-monitoring-technology-offers-early-detection-and-treatment Retrieved 20-09-2019.
[31] For example, a robotic milking machine obtains information from a device worn by a cow and is able to automatically lead her out of the robot if it does not recognise that she is due a milking session. See Heyden, n. 14 above. See also the process described by Postma, n. 21 above.
[32] For a discussion, see Public-Private Analytic Exchange Program. (2018). ‘Threats to precision agriculture’. Available at https://www.dhs.gov/sites/default/files/publications/2018%20AEP_Threats_to_Precision_Agriculture.pdf Retrieved 20-09-2019.
[33] Wathes, et al. (2008). As cited in Bos, J. M., Bovenkerk, B., Feindt, P.H. and van Dam, Y.K., ‘The quantified animal: Precision Livestock Farming and the ethical implications of objectification’. Food Ethics (2018) 2:77–92. Available at https://link.springer.com/content/pdf/10.1007%2Fs41055-018-00029-x.pdf Retrieved 20-09-2019.
[34] For an explanation of objectification, see Adams, C. (1990). The sexual politics of meat: A feminist-vegetarian critical theory. New York: Continuum.
[35] Harfeld Lynning, J., Cornou, C., Kornum, A., Gjerris. M. (2016). ‘Seeing the animal: On the ethical implications of De-animalization in intensive animal production systems’. Journal of Agricultural and Environmental Ethics 29(3): pp. 407–423.
[36] Martha Nussbaum is a well-known author on the wider concept of objectivity central to historical feminist literature. See Nussbaum, M. (1995). ‘Objectification’. Philosophy and Public Affairs. 24(4): 249–291.
[37] Bos, Bovenkerk, Feindt & van Dam, n. 33 above, p. 89.
[38] ibid.
[39] Werkheiser, I. (2018). ‘Precision Livestock Farming and Farmers’ Duties to Livestock’. Journal of Agriculture and Environmental Ethics (2018) 31(2): 181.
[40] Bos, Bovenkerk, Feindt & van Dam, n. 33 above, p. 89. It is acknowledged that there is a discussion to be had around whether an ethics of care approach in PLF could re-present the literal individual living being to the point at which it becomes difficult to ignore their individual subjectivity. In such a scenario, PLF could, in theory, change human–nonhuman relationships and impact on the ‘value chain’. PLF also has the potential to further illustrate that human individuals involved in the promoted ‘value chain’, farmers, for example, are victims of an objectifying, oppressive system. These discussions are out of the scope of the present paper.
[41] Providing welfare benefits has long been associated with achieving higher yields. Literature cited in this discussion will be seen to relate the two, and prior to the development of technological interventions in animal processing, the UK Farm Animal Welfare Committee explained that “providing soft bedding for dairy cattle can significantly increase milk yield”, that meeting the behavioural needs of animals can reduce costly outbreaks of injurious behaviours, such as tail biting in pigs, and generally that “sympathetic handling improves productive output”. See, for example, Farm Animal Welfare Committee. (2011). ‘Economics and farm animal welfare’, p. 14. Available at https://assets.publishing.service.gov.uk/government/uploads/system/uploads/attachment_data/file/324964/FAWC_report_on_economics_and_farm_animal_welfare.pdf Retrieved 17-10-2019.
[42] The development of responsible AI in PLF is presumably relevant only in the context of existing ‘welfare’ measures, which is a lower ethical standard than that deemed necessary to protect humanity from the development of artificial intelligence that can cause destruction, harm and kill.
[43] See, for example, Tegmark, M., n. 4 above, ‘The Asilomar AI principles’, pp. 329–331. Meanwhile, Tegmark, a primary historical player in the development of ethics for AI, does acknowledge the human ‘enslavement’ of nonhuman animals, see p. 182.
