Artificial intelligence, animal welfare, and the ethics of smart farming | Walter Veit Podcast #9

Summary

In this captivating episode of the Walter Veit Podcast, Dr. Walter Veit, an assistant professor in philosophy, explores the intersection of artificial intelligence, animal welfare, and smart farming ethics. Emphasizing a moral obligation for animal farms to utilize AI technologies, Veit argues these systems can enhance animal welfare without compromising productivity. He advocates for precision livestock farming (PLF), a transformative approach promising individualized animal care via AI, potentially revolutionizing animal husbandry by minimizing disease and improving living conditions. By addressing skepticism and ethical concerns, Veit underscores the role of AI in enhancing both economic and welfare outcomes, positing that PLF signifies a pivotal shift towards ethically-conscious and sustainable farming practices.

Highlights

• AI can radically change animal welfare in farming by allowing farmers to monitor and treat animals with precision. 🧠
• Farms are morally obliged to adopt AI as it can enhance animal welfare without reducing productivity, a win-win for both. 👏
• Precision Livestock Farming (PLF) involves using technology like sensors and AI to improve animal welfare and productivity. 🔍
• AI-driven smart farming can preemptively address health issues, reducing animal suffering significantly. 🚑
• Economic and welfare interests align in smart farming; it's a promising path for sustainable and ethical agriculture. 🌾
• With increased AI adoption, the future could see a return to individualized care while maintaining productivity. 🔄
• Investment in smart farming technologies is surging, indicating their potential in improving welfare and economics in agriculture. 📊

Key Takeaways

• Artificial Intelligence (AI) holds the key to transforming animal welfare in farms by offering precise, individualized care. 🐄
• Animal farms have a moral duty to implement AI technologies, as they can improve welfare without reducing productivity. 🌱
• Precision Livestock Farming (PLF) uses advanced technologies to monitor and enhance animal welfare effectively. 📈
• Smart farming can minimize harmful practices and maximize animal well-being through targeted interventions. 🎯
• Economic benefits align with improved welfare in precision livestock farming, presenting a win-win scenario for farmers. 💰
• There is growing confidence in AI's ability to enhance farm productivity and animal welfare simultaneously. 🤝
• The synergy between AI and farming could mitigate animal suffering globally, especially in developing nations. 🌍

Overview

Dr. Walter Veit takes us on an intellectual journey exploring the promising avenues that artificial intelligence opens for enhancing animal welfare in farms. His discourse highlights the potential of AI to monitor individual animals meticulously, proposing a radical ethical stance where farms are morally bound to adopt these technologies.

According to Veit, Precision Livestock Farming (PLF) utilizes sophisticated technology like AI to boost both the welfare and productivity of farm animals. PLF not only anticipates health issues earlier than traditional farming but also offers a cost-effective solution aligning economic and ethical interests—a once-distant possibility now within grasp.

Veit addresses the doubts and resistance both from farmers and ethicists but insists on the unprecedented benefits PLF brings. He envisions a future where AI doesn't just automate but elevates the standards of animal care globally, especially in under-developed areas, paving the way for a more sustainable and kinder form of agriculture.

Chapters

• 00:00 - 00:30: Introduction to Walter Veit's Work This chapter provides an introduction to Walter Veit's academic background and research interests. It highlights his current position as an assistant professor in philosophy at the University of Reading and his affiliation with the Munich Center for Mathematical Philosophy at LMU Munich. Veit's research is primarily centered around the philosophy of cognitive and biological sciences, philosophy of mind, and applied ethics.
• 00:30 - 01:00: Overview of AI and Animal Welfare The chapter titled "Overview of AI and Animal Welfare" focuses on the philosophical aspects of animal consciousness, particularly in the context of artificial intelligence, animal welfare, and the ethics of smart farming. The speaker, Walter, highlights the importance of these themes and thanks Jonathan for setting the stage for the discussion. The talk promises to delve into the intricacies of these interconnected topics.
• 01:00 - 02:00: Perspective on AI's Role in Animal Farming This chapter explores a unique perspective on the role of AI in animal farming. Unlike the common skeptical view that AI contributes to animal suffering, this discussion highlights the potential positive impacts AI could have. It challenges the conventional understanding and shows AI as a tool that might help animals, offering a more optimistic outlook. An illustrative image of a robot hugging an animal symbolizes this hopeful perspective.
• 02:00 - 03:00: Smart Farming and Ethical Obligations The chapter discusses the moral obligation of animal farms to implement smart farming AI technologies. The central thesis argues for a transformative relationship between humans and animals facilitated by AI. The focus is on the ethical responsibility of farms to adopt technologies that promise improved operations.
• 03:00 - 04:00: Precision Livestock Farming (PLF) The chapter discusses Precision Livestock Farming (PLF) and its focus on improving animal welfare without reducing productivity. It highlights that there is not much of an argument against incorporating these technologies in farming. The discussion references a recent book, derived from a PhD thesis, titled 'A Philosophy for the Science of Consciousness,' which aims to answer questions regarding animal experiences, particularly in the context of their interactions with technology.
• 04:00 - 05:00: Benefits of PLF in Animal Welfare The chapter discusses the potential impact of Positive Livestock Farming (PLF) on animal welfare by considering both positive and negative states that can affect their well-being. The author also explores how frameworks used for understanding animal welfare might be extended to Artificial Intelligence Systems (AIS), which are even further removed from human consciousness. Collaborating with a colleague, they have recently published a paper titled 'Studying Introspection Animals in the Eyes,' suggesting a possibility of gaining insights into AI consciousness by studying animal minds. Originally, the focus was to elaborate on how insights from animal cognition could be applicable to our understanding of AI minds, but it seems the narrative shifted during the discussion.
• 05:00 - 06:00: Economic Implications of PLF The chapter titled 'Economic Implications of PLF' stems from a newfound research interest in animal welfare and the influence of artificial intelligence (AI) on it. Recent developments in this field, particularly a series of papers on AI ethics, have highlighted how ethicists have previously overlooked the impact of AI on animals. The focus is now shifting towards integrating ethics with animal welfare considerations. Additionally, a conference interaction has contributed to further exploration in this area.
• 06:00 - 07:00: Skepticism and Ethical Considerations In the chapter titled "Skepticism and Ethical Considerations," the discussion revolves around the advocacy for recognizing animal consciousness and improving their rights and welfare. The influence of Marian Dawkins, a prominent Animal Welfare scientist, is highlighted as a significant factor. Dawkins, known for defending animal welfare, served as an informal advisor, playing a crucial role in guiding efforts towards animal rights and ethical treatment.
• 07:00 - 08:00: Technological Developments and Challenges This chapter focuses on recent technological advancements and the challenges they pose, particularly in the realm of animal consciousness and welfare. The discussion presents a more skeptical view on the current optimistic attitudes towards the progress in animal consciousness science, highlighting the lack of certainty in the field. Additionally, it explores the role of smart farming technologies in promoting animal welfare, indicating the increasing influence of artificial intelligence across various domains.
• 08:00 - 09:00: Conclusion and Future Prospects The chapter titled 'Conclusion and Future Prospects' discusses the significant areas of influence on human lives as identified by the European Parliament. One of these emerging areas is smart farming. While the focus isn't on how technology might improve plant farming, there's an emphasis on the potential impact on animal farming. Through this discussion, possibly echoed by figures like Peter, the chapter underscores the advancements and future directions in farming technology.

Artificial intelligence, animal welfare, and the ethics of smart farming | Walter Veit Podcast #9 Transcription

• 00:00 - 00:30 hello Walter and Walter fight is assistant professor in philosophy and at the University of reading as well as an external member of the Munich Center for mathematical philosophy at the lud maximilians University of Munich and his research focuses on the intersections of the philosophy of cognitive and bi biological science the philosophy of mind and applied ethics and his latest
• 00:30 - 01:00 book is on a philosophy for the science of animal Consciousness so we're very much looking forward to your talk Walter and thanks so much for the kind introduction I'll try to talk very loud then I hope that helps uh let's see if my voice gives up by the end of the talk so I'll talk today about artificial intelligence animal welfare and ethics of smart farming thanks to Jonathan we already had a bit of an introduction here so that's a nice way of going and
• 01:00 - 01:30 moving into this topic now what I am going to talk about is a bit of a different perspective from what we've heard in most talks which was a bit of a skeptical type where AI as a source for animal suffering is typically seen as a problem right AIS aren't really seen as something that might help animals in any way so nice thing about this poster I think is that choose this robot hugging
• 01:30 - 02:00 a pig and perhaps that's nicely suggestive of a very different kind of relationship we might have between the eyes and animals now the thesis I will defend today perhaps a bit of a radical view that is that animal farms have a moral obligation to implement smart farming AI Technologies why do I hold this view why will I defend this well I will argue that since they can improve improve
• 02:00 - 02:30 Animal Welfare without diminishing productivity there really isn't much of a good argument against introducing these Technologies in a farming context now I already mentioned uh my recent book uh based on my PhD thesis a philosophy for the science of Consciousness in which I try to yeah develop a scientifically grounded approach to answering questions about what animals experience especially in relation
• 02:30 - 03:00 to positive and negative states that can incorrect their welfare I've also written about how we might expand such a framework to AIS who of course are even farther away from our conscious experiences I've worked together here often with head of drowning a recent paper that came out is called studying introspection animals in the eyes and originally I was thinking of talking about how we might learn about AI Minds from animal Minds however the has been a
• 03:00 - 03:30 recent shift in my research interest um to focus more on Animal Welfare and especially how a I can impact animal welfare and that's largely been caused by three development one was a recent series of papers by off us present here on AI ethics and how ethicists have neglected um how AI might impact animals and how we might include animals in then I've also been at a conference where I got to meet the Dal was already
• 03:30 - 04:00 mentioned here where are several speakers um were also present and he really urged us on to not only investigate animal Consciousness but also try to help get their rights recognized help to improve their welfare lastly the most important factor has been Marian Dawkins who served as somewhat of an informal advisor on my PhD she's a very influential Animal Welfare scientist and perhaps known to many here as someone who has defended a
• 04:00 - 04:30 bit of more skeptical take about animal Consciousness science uh against a lot of very positive attitudes recently about the progress of the field how we can have more certainty but that's not really what I'll focus on here today but rather her recent work on Smart farming for Animal Welfare where she has uh recently written a bunch of papers on the subject now it's of course clear that the I is creeping into all domains
• 04:30 - 05:00 of human Liv here the European Parliament has rele yeah show what perhaps the most significant areas are where I will influence our lives but interestingly here there's also smart farming mentioned as an important area now I'm of course not focusing on um how ice might be used to better Farm plants but rather the animal case um so Peter sing and make a really good case for why we
• 05:00 - 05:30 should have something of field or thei ethics that also talks about animals perhaps a kind of an interdisciplinary field here but at some places in the paper it almost reads as if there has been not any attention given to these topics and while this might be true for animal ethesis and AI eist I think it's a mistake when we think about all the attention that has already been given by Animal Welfare scientists to this uh pressing issues and Animal Welfare
• 05:30 - 06:00 science is of course not a purely scientific field it also involves value judgments where Animal Welfare scientists really try to improve the lives of animals Animal Welfare science can be seen as the science of animal wellbeing now obviously in a farming context we heard yes wild animals also constitute a I might make a difference and I've also written about these issues similar to other ORS here but I think
• 06:00 - 06:30 it's clear that when it comes to having impact right now the farming context is really where there's most animal suffering taking place that we can actually impact right we are quickly approaching 100 billion mammals and Earth being killed each year for human consumption and these numbers are only expected to increase even further right industrial large scale farming is screwing vestern and develop Nations
• 06:30 - 07:00 alike and there's not only poor welfare conditions and developing nations but they exist across the globe so there urgent need for change and I really want to motivate here the idea that smart uh farming can really help us to solve this problem in a way that nothing in the last perhaps 30 years has gone ever that close now what is Smart farming Jonathan gave us a brief introduction I'll talk a
• 07:00 - 07:30 bit more about this here so you get a bit more of an Insight after all not all of you might be familiar with this literature now it's the integration of Advanced Technologies into agricultural practices to improve efficiency productivity and sustainability we should also mention welfare here sometimes welfare is mentioned sometimes it isn't unfortunately um but smart farming yes relies on um combining a set of new technological developments think of
• 07:30 - 08:00 drones sensors robotics electronic variable trackers that can be attached to animals that might keep track of the heart rate and other factors big data and artificial intelligence the letter of which I will focus on here now we already know that smart livestock farming can impact animals and what I'm especially will focus on here is the ethics of these Technologies and yes I'll from now on talk mostly about precision livestock farming because
• 08:00 - 08:30 that's a more precise term what is precision livestock farming um often abbreviated it as plf by Engineers it's the application of Advanced Technologies and practices to enhance the efficiency and Welfare of livestock management through continuous realtime tracking right so you might think that a farmer goes through his farm uh with thousands of animals might see uh animal limping um intervenes
• 08:30 - 09:00 calls the veterinarian um tries to solve these issues but there might be hundreds of more animals that suffer from the same or different afflictions that um no one in the farm is able to keep track of perfectly um we only have a certain amount of time per day that we can spend with each animal and tii might provide us with something of a revolution where we can keep track of animal anal continuously and assess them all at the
• 09:00 - 09:30 same time so we can have this variety of devices sensors cameras microphones to Monitor and analyze Health welfare and performance of individual animals and this field has really grown rapidly in the last couple of years was somewhat of a slow burner between 200 2011 but in the last 10 years or so there has really been a rapid roof of
• 09:30 - 10:00 this fi with many papers being published it's taking up an ever increasing amount of Publications in the field of Animal Welfare science now uh because I mentioned performance and Welfare perhaps let me give a brief primer of what I mean by these terms performance efficiency can be measured in terms of food or economic output for instance milk output we can have different proxies here um and that's of course not what I'm
• 10:00 - 10:30 really interested in I'm not arguing that we that um farmers are obligated to introduce uh smart farming because it will increase their performance but rather because it will impact welfare welfare will be defined as the balance of positive and negative mental experiences across the life of animal now this definition is slowly becoming the majority view in the field and here we can then Define a good life as a life
• 10:30 - 11:00 in which positive mental experiences outweigh the negative ones um to briefly talk about health bad health will of course typically diminish welfare but it's not uh somehow conceptually part of this um it's just a causal fact that bad Health typically will make our lives go worse H nevertheless you can for instance think of a three-legged dog that is still happily uh limping around
• 11:00 - 11:30 almost running and really not having any worse um life experiences so just to make those definitional points clear now let us move on um when it comes to Precision livestock farming there is a lot of skepticism and if you think about two domains that might be optimized as Jonathan mentioned we might think well if you focus only on the economic side efficiency here then there are legitimate
• 11:30 - 12:00 perhaps that this might harm animals think of Broiler chickens Brad to be so large and growing so F that they suffer from frequent hunger metabolic problems heat stress broken bones and other health issues now this improves the performance for farmers and large corporations but of course it's not great for the welfare of individual chickens chickens that don't grow so fast or to be that large clearly have better lives they don't constantly
• 12:00 - 12:30 suffer from Hunger however what I'm will argue here is the Precision livestock farming aimed at improving Health and Welfare really has the potential to vastly improve the lives of farm animals much more than any of you in the room perhaps might predict and there are several reasons for this firstly by constantly keeping track of them and detecting problems long before pharmace Arians May notice them um we really make
• 12:30 - 13:00 a lot of progress um that we might not expect so there's a lot of arguments going around that uh smart farming will lead to the further exploitation of animals that we will have an even less individualist approach to animals so think of perhaps Farms existing in the distant past romanticist where a farmer just had perhaps a 100 animal at most took individual care for them almost had
• 13:00 - 13:30 a kind of bonding experience and here I really think that the opposite is true that really uh Precision livestock farming will almost get us partially back to this form of interaction um where the subjects will perhaps not be humans they will be non-human AI systems but nevertheless there will be more uh caregiv to really try to um yeah assess individual animal so one
• 13:30 - 14:00 example that is I think a very nice one is the reduced use of antibiotics so because diseases can rapidly spread through populations um in these very small enclosures where animals are put very close to each other and um no one is able to detect the signs of disease very early on really because there's just too many animals often Farmers just give antibiotics to all the animals to just deal with this but that's of course lead leads to antibiotic resistance uh
• 14:00 - 14:30 pathogens and leads to further problems down the line do might even be spread on to humans um but these complications are SED here we can then use a targeted approach by assessing perhaps an animal that is coughing um and then targeting them giving them antibiotics rather than just all animals so here really we take a more individualist approach and this will lead to Better Health and Longevity for individuals rather than just their active um and there are many factors
• 14:30 - 15:00 here where measuring sounds Behavior temperature growth injuries parasites as well as death will really help us to make progress with Animal Welfare uh death of course is also a welfare issue um animals typically do not enjoy the presence of dead companions um other than the fact that this will further spread disease um it's also um very uncomfortable to many animals um um
• 15:00 - 15:30 finally we might almost say that through machine learning and feedback especially from Farmers once these Technologies are widespread uh applied in almost all Farms ECI systems will likely become even better than humans at detecting these problems as sometimes we talk about these Technologies as just being tested uh on a small scale and then they're delivered to large Farms um without further progress taking place
• 15:30 - 16:00 but that's of course not true we know that AI is rapidly advancing similarly how to how human feedback to Jet gbt has further uh evolved almost jet gbt to be uh even more advanced similarly here the feedback from Farmers um and assessing the error rates we will have um be able to further and further improve um the success of these detection devices
• 16:00 - 16:30 now um because this is a very nice picture to illustrate how useful it would be some Farmers have to manually count Kettle now this process is prone uh to have errors it takes a lot of time we having automated counting is much more reliable similarly to paper based record keeping it can be very unreliable it's much more useful to hear you say all algorithms they can also detect errors
• 16:30 - 17:00 similarly manual Health assessment it should be clear will be much worse than also having continuously keeping track of animals their health States the hormone levels the stress levels and the like now there have some fur examples that I think make it very clear why we should rapidly try to bring these Technologies uh to Farms anywhere in the world uh for instance is one AI tool I won't mention the name here but you can
• 17:00 - 17:30 check it in this blog uh helps to monitor piglets for the stress and has reduced piglet crushing by 31.37% and this of course shouldn't be neglected when if you can decrease the cause of death by One Source by 30% that is a great result and any farmer would be rational to implement this as fast as possible similarly we can have um adjustments of temperature in real time
• 17:30 - 18:00 based on the body temperature that is being assessed of animals as well as other behavioral signs such as freezing overheating which of course in turn can cause discomfort disease and even death other signs for instance if you uh assess the sounds we put recording microphones up can help us as I mentioned to detect coughing as an early sign of respiratory problems this animal might be isolated from others so they won't be infected we can give them
• 18:00 - 18:30 medicine in order to ensure that they will uh win their health back rather than just die similarly it can help animals um or Farmers to much better communicate with each other uh we already know that animals are capable of distinct stress cods especially in juveniles for cold hunger loneliness among other criteria but there's just no way for a farmer to listen to all these thousands of animals individually even
• 18:30 - 19:00 if you have a lot of farmers running around um here AI can really help us to intervene fast um which is an very important issue when it comes to many of these Factor if an animal complains about being hungry or cold for a long time and we're not able to respond in hours and it might have horrible consequences now to come to the ethical question or perhaps the two most radical stances either Banning these
• 19:00 - 19:30 Technologies all together because we might fear they will exploit animals further at the other side we have obligations we might think well clearly these Technologies uh have this potential to improve animal life so much further there is an obligation for Farmers to implement these Technologies as fast as possible now your own views will probably uh differ here uh from each other and would plausibly depend on
• 19:30 - 20:00 how much you are of an optimist or pessimist both about developments in the farming industry generally as well as ethical commitments you might think AI Engineers have now we know that some companies uh have somewhat more of a lip service to EIC commitments but then when it comes down to it they might neglect their advice given by aif assists within their own companies a assists might be fired uh so this can legitimately justify some pist views about um how
• 20:00 - 20:30 much these Technologies even in the animal case might really be used to improve welfare perhaps again this is just lip service um so skepticism is very common both among animal ethicists and activists but surprisingly a strange factor is that a lot of fers themselves seem to share that very skepticism they do not want to implement Precision livestock farming
• 20:30 - 21:00 and how can this be explained is this inconsistent so a lot of animal eist seem to object to PF Precision livestock farming because it neglects Animal Welfare for a pursuit of economic benefits and then a lot of farmers seem to object to this very technology because they don't believe that these technology surpass their own skills so there has been a lot of talk about about Farmers thinking well DCI technology
• 21:00 - 21:30 companies almost frame it as um us as Farmers being incapable bad at our job bad at keeping track of the health and animals and then these new technologies that they've developed in a few years are meant to be much better than us um it's almost seen as something of an a front and also they don't see how they would provide a major benefit at least against their costs even if there might be some use um that should be of course more useful than perhaps spending time
• 21:30 - 22:00 on hiring more veterinarians instead J prooved Animal Welfare so one strong counterargument we might have here against the view that Precision life um stock farming should become obligatory for Farms is that they don't actually improve animal lives and this the first counter argument we will examine here so how do we deal with this well one response we can give is that it's just incredibly conservative stance similar
• 22:00 - 22:30 to a lot of other common denials by the public that AI could never achieve x with X just standing in for whatever ability we might imagine say AI beating um the world chess champion in chess um or other abilities such as writing philosophy essays at least on an undergrad or a Master Level and the I routinely surpasses those abilities and it's just moving on uh debunking and
• 22:30 - 23:00 again again um what we might put up as a kind of barrier to the abilities the I will never surpass especially when it comes to what we think are unique human skills um that the ice will just never achieve but at least in the last few years I think there skepticism is becoming weaker especially due to the development sgbt so really a more moderate stance I
• 23:00 - 23:30 think would be better here and that would again lead us perhaps to thinking well while these Technologies might not be in the moment as good as actual Farmers or veterinarians giving track of these animals they probably will be in the not all too far future now Jonathan mentioned that Animal Health and Welfare isn't mentioned a lot compared to production but that doesn't seem true based on some literature that have been conducted in the last years where in
• 23:30 - 24:00 fact the majority of Publications uh Focus us on the application of precision livestock farming for Animal Health and Welfare rather than production now in some cases we might think well this might just be lip service where perhaps really the interests are about production but because a lot of these Publications come from anual welfare scientists working together um with AI engineer years I think we can be fairly confident that
• 24:00 - 24:30 there is a genuine interest by the researchers developing these Technologies to improve the lives of animals after all the attitudes of humans to animals have changed dramatically over the last 50 years and they are um changing more and to more after all the numbers of vegans vegetarians are only continuing to increase um now I mentioned one counterargument that is often given here that the I will be leading to the exploitation of animals because we'll
• 24:30 - 25:00 focus on the collective rather than individuals but I think the cases and examples I've given before are nicely illustrative of that not being true at all really these tools are in fact meant to focus much more on individuals than farmers are capable of because they are thousands of individuals they just don't have the time to focus on all of that so really there is a bit of a revolution here where we move back almost at least partially to an all the way of farming
• 25:00 - 25:30 where individuals are actually given attention the only difference might be that there's not an actual human coming to the animal and examining it um and you might think well perhaps there's something lacking a bond that won't be created with an AI um but of course animals are often scared of um Bas veterinarians coming to them it might involve painful procedures um so it's not at all clear whether including
• 25:30 - 26:00 individuals that actually have to approach animals for measurements is better this might cause them more stress than having just an system keeping track of all these facts um that the animal isn't even aware of furthermore there has been an explosion of startups and government funding um into Precision livestock farming we suggest the least funders have high confidence in the eventual success of these Technologies and the goal here isn't just to improve
• 26:00 - 26:30 productivity but for a lot of funding that goes into this area really um welfare is a major um interest when you think of um organizations like open philanthropy for instance that are very interested in funding these areas what they are very much interested in is reducing suffering in the farming context and a lot of startups have been deliberately uh created to improve the lives of animals so we shouldn't just
• 26:30 - 27:00 attribute them the desire to mely maximize profits at a cost of Animal Welfare I think that wouldn't um be fair uh furthermore the usefulness of these Technologies has already largely been demonstrated in small scale studies as well as some larger Farms that are already starting to implement these Technologies and in fact when we think about the application of these on a much
• 27:00 - 27:30 larger scales where we have huge corporations running hundreds of farms um it will in fact become much cheaper for them to implement these because of Economics of scale so the argument um that this couldn't be done because it's too expensive I think won't be useful here uh nevertheless I think it does provide a strong second counterargument because we might of course think think that if Precision livestock farming is too expensive it
• 27:30 - 28:00 can't possibly be an obligation for Farms because these Farms run on very thin profit margins and thus can't afford these technology so let me make a case against this and we might think that even if Precision livestock farming indeed improves Animal Welfare like many other interventions that still might Force Farms to run losses and because this field is so competitive it might Force these Farms to go bankrupt now I think it is very unlikely that
• 28:00 - 28:30 Precision livestock farming won't improve economic output here as well as welfare because issues like disease injury and the like um which is perhaps somewhat different from other interventions such as employing chickens that are less prone to Growing rapidly into the same size um there isn't really this tradeoff here between um welfare issues and economic issues there's some
• 28:30 - 29:00 tradeoff admittedly Jonathan pointed us out and where we might place our error rates if we prefer positive um versus false errors uh nevertheless I think that is a bit overstated because there's so much going on here where um the productivity um interest and the welfare interests are really identical um so at least for the initial uh implementation of these Technologies there really isn't
• 29:00 - 29:30 um any counterargument um any attempts to try to slow down uh the use of these Technologies is much more likely to harm animals um which of course isn't something we would want and once these Technologies are widespread we might try um to focus more on welfare where tradeoffs arise but for all the initial applications such as Pig crushing
• 29:30 - 30:00 detecting diseases that will kill animals here we really have um productivity and profit um coming to barely conflict with the goal of improving the lives of animals um furthermore again as I mentioned before these technologies will become more and more affordable the more companies are using them um as they become more widespread they
• 30:00 - 30:30 will become more affordable for smaller Farms as well economies of scale should make a more clear case here finally given how mous Farms are already now part of huge businesses um it should be very easy for them to adopt these Technologies um without um having to perhaps ACR a lot of um funding or initial initial starting Capital to get this technology off the ground perhaps a
• 30:30 - 31:00 large operation might test uh Precision livestock farming in one particular farm and then track how performance improves both for productivity and Welfare and so far most of these tests um are highly suggestive um that Precision livestock farming both improves welfare and productivity um finally another counterargument we might have to this is that well well if these Farms can't exist um perhaps these Farms shouldn't
• 31:00 - 31:30 exist if they can't afford to invest in the welfare of the animals and that's of course an old argument where we might say well a lot of these um large scale industrial Farms just have very poor welfare and perhaps they shouldn't even exist and if that's the case that's not really an argument against the obligation to implement Precision livestock farming but a completely different argument all to together for the Farms that perhaps continue to exist
• 31:30 - 32:00 um that focus more on welfare there I think it is a clear obligation to implement these Technologies now to deal with a final counter objection and that is often made in the last 10 years and more generally in Animal Welfare context it's a general complaint that Animal Welfare science has in the past mostly focused on the presence of pathological conditions uh on suffering rather than positive
• 32:00 - 32:30 experiences right so a lot of the um Technologies we've discussed here largely focus on avoiding negative States um and we might think that this can be a similar problem here where these Technologies might lead us to something that has been called proxy failure where we might have or Implement a measure um in order to achieve some all think perhaps of an education context where we um introduced various
• 32:30 - 33:00 tests for TR keeping track of educational attainments but over time those measures furiously seemed to be less reliable in getting at the Target they've actively become verse as almost a kind of gamifying process here and similarly we might think well that these Technologies um would focus more and more on the health side with the positive experiences becoming more neglected but there multiple arguments against this well firstly health
• 33:00 - 33:30 conditions will remain the main driver of poor welfare in industrial Farms um for at least the next decade um so there's no doubt that even if these Technologies would exclusively deal with these problems they would still lead to Major welfare [Music] improvements and furthermore once these Technologies are widespread we might think that actually trying to tweak them in order to also use them to improve
• 33:30 - 34:00 positive experiences for instance by assessing signs of frustration loneliness boredom um here we also have a way of interfering um think think perhaps of animals being allowed to interact with each other if they say show signs of loneliness and boredom perhaps there could be an automated form of a connecting cages and um then closing them up against each other where animals can come in contact with each other really I would provide a very nice
• 34:00 - 34:30 way here ofhs allowing us to much more reliable focus on the positive experiences of animals when farmers and veterinarians just are running behind all the negative um states that they are encountering unable to really treat um all animals so really AIS will help us to keep better trick of all these negative experiences and once we've done this I think it will become much more easy to also Implement
• 34:30 - 35:00 um uh tools to improve positive experiences of animals using this existing infrastructure right um and because Animal Welfare scientists are so closely involved in trying to implement these Technologies I think we can be fairly confident that this will happen and now as we see here in this conference there's increasing demand for animal nii ethicist to become involved in these debates I think we can help uh to avoid
• 35:00 - 35:30 such a proxy failure where these Technologies would be purely used to deal with health issues but where we can really focus perhaps in the future on making the experiences better and think for instance of these rushes used um for cows um that give them very positive experiences you might say it helped them with cleanliness and the like but it's undi that a lot of cows just approach these um for the pure joy of being scratched of having this positive
• 35:30 - 36:00 feeling of being massaged so to conclude this talk um smart farming Precision livestock farming has unparalleled potential at improving animal welfare and importantly without the typical costs associated with welfare improvements in farms uh increasing cage sizes uh using um non-modified uh chickens uh obviously here there's a
• 36:00 - 36:30 trade-off between welfare and economic side of these businesses um but for many issues in smart farming that is in fact not the case where these two interests align and is only once we've um achieved the health of these animals uh that perhaps there will be a trade-off in the future but I think that's a problem perhaps for 10 years from now rather than now we shouldn't really try to stop the development of these Technologies
• 36:30 - 37:00 try to have policy constraints because I don't really see the issue here now in fact I think that opposition St we already saw that these Technologies um are able to prevent um p l to crushing uh we can very fast detect some diseases we can cure them before they lead to lethal
• 37:00 - 37:30 results and in fact the future affordability of these technologies will I think importantly increase public pressure to improve Animal Welfare further a lot of the arguments that are often given against improving Animal Welfare are that's just not economically viable but as we can see with these Technologies there is a vast range of Animal Welfare improvements we can um implement that will also help the productivity of these businesses thus
• 37:30 - 38:00 improving Animal Welfare standards and again increasing the public perception of what acceptable welfare standards are for farmed animals so I'm hopeful here that there will be something of a feedack loop um that will lead to animal welfare being taken ever more seriously now Animal Welfare scientists I think it's important to mention are already closely engaged with the development of these Technologies and that is important to mention because they are very much interested in
• 38:00 - 38:30 ensuring that these Technologies are ethical that they're not just used to exploit animals to uh further their suffering for the M sake of productivity um that is not um to say that the epist can't become involved here that is far from the truth um but it's not to say that epist are the only ones who interested in ensuring that these Tech techologies are ethical and it's important here I think to have more
• 38:30 - 39:00 exchange between Animal Welfare science and animal eist as well as AI eist to move further to the final sentence here future developments in ey technologies will largely enable these large scale um industrial farming practices to partially return to a model of individual care where we can take um preventative measures um for the spread of disease very quickly we can deal with welfare issues um very fast R
• 39:00 - 39:30 having to wait very long where it becomes very obvious where an animal shows very clear signs of something being wrong with them AIS can really help us here and I think this is quite contrary to the fear some Skeptics have expressed um that it will lead um to a kind of collectivization of these animals that are no longer seen as individuals they're just a type of product I think it couldn't be F of from the truth and dcis will in fact in a be
• 39:30 - 40:00 set in a way to r Treat these animals as individuals to be taken care of and I think that is a very nice um perspective where we can think well the I must not be used in any harmful ways in fact the II might be more interested in keeping track of the weal of these animals but that's just said in the algorithm whereas Farmers uh might be more um susceptible to having their individual
• 40:00 - 40:30 interests time and the like tiredness interfering with the welfare interests of the animal the AI won't be having such a problem uh finally I think I've made a good case here I hope you have our F arguments perhaps to counter this that animal FS really have an obligation to implement these smart farming Technologies to improve Animal Welfare I think there's little out that they will in the future make a major difference and it's just a
• 40:30 - 41:00 question of time for how fast um we can spread these Technologies once they become available in Western Nations um I hope many economists would agree that they will become more affordable spread to developing nations where there's obviously even more animal suffering where this trade-off between welfare and productivity isn't really taken serious typically productivity is the only thing that is being maximized there but because AI systems can help us to reach
• 41:00 - 41:30 so many welfare benefits that also coexist with productivity it's especially then in developing nations uh where we might have a kind of trickling down of these technologies that can improve Animal Welfare further and with that thanks so much uh for listening and we can move to the [Applause] questions