AI for Animals #2: AI in factory farming
August 1, 2024
Welcome to the second edition of AI for Animals!
Roughly once a month, this newsletter will give a brief overview of a specific topic relating to AI and animals, along with the latest news and other useful resources. While the main focus will be on biological animals, we’ll also include resources relevant to digital minds.
We’re always open to ideas, questions, and feedback: just email contact@aiforanimals.org. If you’ve been forwarded this and would like to subscribe, you can do so here.
Thanks to Allison Agnello, Yolanda Eisenstein, Shaun Feakins, Ali Ladak, Constance Li, and Julio Veuthey for their contributions to this month’s edition!
Max Taylor
Index
AI in factory farming
Animal agriculture isn’t yet capitalizing on the potential benefits of AI
Imagine you’re a data analyst set loose in an intensive broiler chicken farm with a range of cameras, microphones, and other sensors (including ones that can be placed on the chicken as part of a collar or leg bracelet).
At the individual level, you can measure things like the chicken’s heart rate, blood pressure, respiration, weight, temperature, vocalizations, movements, and location. You can collect all this information using pretty simple tech: no AI required.
However, actually making sense of all these measurements is much harder. Fortunately, you can use AI’s adaptive pattern-spotting and predictive abilities to assess the interaction of all these variables. This allows you to reliably understand the bird’s sleep patterns, stress levels, growth rate, likelihood of disease, mobility issues, and engagement in natural behaviors like foraging and play.
Zooming out to the group level, you can monitor how these variables change as chickens interact, allowing you to identify hierarchies, indicators of aggression, and the importance that each bird places on social behaviors.
Zooming out further to the farm level, you can measure factors like air quality, temperature, humidity, carbon dioxide levels, feed availability, and total stocking density, as well as the way in which all these factors correlate with your individual and group-level measurements.
This matrix of data, clarified by AI, shows you how the farm would need to be run to maximize productivity, cost efficiency, sustainability, and animal welfare in line with the producer’s wishes.
Back in reality, almost all of this data is going unmeasured. Even the largest, most intensive farms typically capture just a fraction of the data that they could, with on-farm decisions still largely dependent on sporadic observations by time-strapped workers.
One reason for this is that most of the necessary technology is still in its early stages. There are exceptions: for example, there are already commercially available technologies that can monitor lameness in dairy cows, respiratory health issues in pigs, parasites in fish, and growth rates in chickens. Studies suggest potentially significant improvements over current observations, with the systems detecting lameness in cows three days before farmers noticed it, and respiratory problems in pigs up to two weeks before being picked up in farmers' and veterinarians' routine observations.
But these systems haven’t been around long enough for most producers to trust and understand them. They’re complicated (and, once you get past the many veils of slick marketing hype, they’re often pretty dull). They raise thorny questions about data ownership and security. And when they’re so expensive to install and maintain, it’s hard to see how the long-term benefits will outweigh the immediate costs.
This will probably change soon – with the most intensive sectors benefiting most
This picture is likely to change. Some notable players have already begun to use AI in their operations; for example, the poultry processing heavyweight Avara is working with tech start-up Beakbook to predict the ‘harvest weights’ of their birds, helping them to make decisions around stocking density and ensure that birds’ weights match customer requirements. Future applications at Avara could include using sensors to predict and address disease outbreaks, detect signs of distress in chickens’ vocalizations, and monitor air quality. This example of concrete AI usage by a big industry player is particularly important in a business like farming, where producers are particularly hesitant to significantly change their practices until peers have done the same.
It’s large producers like Avara that stand to gain the most from AI. They’re the ones that will first be able to afford to install the necessary hardware, pay people to maintain it, and train staff to use it. They’re also the ones where marginal cost savings per farm make the biggest difference for the whole company. What’s more, the largest producers have generally got as big as they have by subjecting animals to cramped, barren conditions that maximize ‘yield’ per dollar at the expense of all else; unfortunately, these are the conditions that also best lend themselves to AI applications. AI systems function best when working in a predictable environment, and a cage where animals’ movements are heavily restricted (and they can never venture far from AI-assisted sensors) is much more predictable than a farm where animals are given freedom to interact, play, and roam outside.
This all has major implications for animal welfare. In our next edition, we’ll explore those implications in more detail and what advocates can do about it.
📚 Resources
The Hive Community Slack has several channels dedicated to discussion of AI and animals, including #c-ai-discussion for broad discussions and #s-ai-coalition for project collaboration.
If you want to dig deeper, the aiforanimals.org website has a list of relevant articles, papers, and other materials giving an overview of the AI and animals space.
🌏 Opportunities
🔔 Deadline Extended! The AI for Animals Program Lead role is now open until August 5th! Full-time and part-time applicants are welcome. Learn more and apply here!
🚨 Updates
The first AI for Animals San Francisco meetup had a fantastic turnout of 85 attendees (43% of RSVPs). Participants came from a wide variety of backgrounds, ranging from nonprofit animal advocacy to big tech companies like Microsoft, Anthropic, Nvidia, and Meta. Following a short presentation about the history and purpose of AI for Animals, attendees shared their ideas and projects through structured speed networking. Attendees reported finding this unique networking opportunity valuable, with many noting this was the first event they had ever encountered at the intersection of AI and animals.
We plan to make these meetups monthly in future!
The Animal Law Commission was formed in 2022 to expand animal law internationally through the UIA, an international legal organization spanning over 100 countries. On July 19–20, the Commission held its first international animal law conference, Business of Animal Law: The Global Impact of Animals and the Law, which included presentations on AI and animals. Yolanda Eisenstein, president of the Commission, explored the opportunity to use AI-assisted natural language processing to decode animals’ communications, and AI-assisted image recognition to track endangered species, but also advocated for an “ethos of restraint” in such interactions. In other sessions, Ian McDougall, former LexisNexis general counsel, argued that fears about AI replacing animal lawyers are overblown, as human supervision will still be crucial into the near future; while Jamie McLaughlin, the Commission’s VP, presented on AI and agriculture, highlighting greenwashing risks and the need for traceability.
🗞️ News & Research
🗣️ Understanding animals
Why talking to animals could soon become a reality (BBC Science Focus)
Veterinarian Jess French suggests AI could soon allow humans to better understand animal communication by analyzing sensory data like sounds and smells. This could revolutionize veterinary care and improve our interactions with animals, enabling us to detect health issues in species that typically hide symptoms, such as rabbits, and to understand complex signals from animals like snakes and fish. French also highlights potential ethical considerations, as AI might reveal distressing messages from animals about their treatment and environments.
Introducing BirdAVES: Self-Supervised Audio Foundation Model for Birds (Earth Species Project)
The Earth Species Project has released BirdAVES, a new AI model designed to analyze bird vocalizations, achieving over 20% improvement in performance compared to previous models. This model uses a large dataset of bird recordings and self-supervised learning to better understand bird communication and aid in conservation efforts. BirdAVES is available on code-sharing platform GitHub for use in bioacoustic research and applications.
AI tool 'SuperAnimal' can accurately analyze animal behavior (Earth.com)
The École Polytechnique Fédérale de Lausanne (EPFL) has developed SuperAnimal, an advanced AI tool that accurately analyzes animal behavior across various species without human intervention. This open-source model tracks animal movements, offering significant benefits for veterinary medicine, conservation, neuroscience, and agriculture by improving the accuracy and efficiency of behavioral analysis. SuperAnimal aims to revolutionize animal research and conservation efforts by providing a more objective and scalable method for studying animal behavior.
What Science Can Tell us About Animal Intelligence with Brandon Keim (Sentient)
In this transcript from the Sentient podcast, Brandon Keim discusses the role of AI in enhancing the study of animal intelligence, emphasizing its usefulness beyond analyzing vocalizations to include body language and gestures, which are crucial for understanding non-verbal animal communication. He highlights research by primatologist Catherine Hobaiter, who uses AI to study chimpanzee gestures. Keim cautions against AI hype, noting that AI should complement the extensive knowledge already gained from traditional animal communication research.
[Podcast] If we could talk to the animals (BBC Sounds)
A podcast with Peter Gabriel, who discusses AI-assisted human-animal communication and endorses the Coller Dolittle prize, a multi-year challenge recognizing significant scientific research that supports the aim of Interspecies Communication. The chair of the prize, Prof Yossi Yovel, provides further details.
🐔 Chicken farming
How artificial intelligence could transform broiler genetics (Watt Poultry)
Advances in AI and data analytics are enhancing broiler genetics by enabling more precise measurement and selection of traits such as growth rate and feed efficiency. AI could also enable objective assessments of traditionally subjective measures, like gait quality and leg health, through the use of cameras. AI also has potential applications in improving hatchability by monitoring and analyzing mating behaviors, which could lead to increased hatch rates and overall efficiency in broiler production.
AI transforms poultry farming in China's mountainous village (People’s Daily Online)
In Tiantaishan Village, China, Xu Qiyong's chicken farm has implemented an AI-based smart breeding system developed by Tencent Cloud and Shenzhen University students. This technology allows consumers to monitor their chickens' well-being via a mobile app and has significantly improved farm management by preventing disease outbreaks and protecting chickens from predators. The system includes pedometers to detect early signs of disease and surveillance sensors for real-time monitoring.
AI-powered lasers could keep HPAI away from poultry farms (Watt Poultry)
AI-powered lasers, developed by iChase.io, can repel wild birds carrying highly pathogenic avian influenza (HPAI) from commercial poultry farms. This technology enhances biosecurity by using AI to identify and track foreign birds with cameras, then deploying random laser patterns to keep them away, preventing the birds from learning predictable patterns. The use of AI and lasers is particularly beneficial given that wild birds are a major source of HPAI, which has impacted over 80 million farmed birds in the U.S.
🐮 Cow farming
Sainsbury’s first retailer to milk the benefits of new AI “vet tech” (Sainsbury’s)
Sainsbury’s has become the first retailer to implement AI veterinary technology on dairy farms to enhance animal welfare and efficiency. Partnering with Vet Vision AI, the technology uses cameras to monitor cows 24/7, providing real-time data on their health and behavior, and offering suggestions to improve their comfort and well-being. As well as detecting early signs of disease, the technology tracks positive behavioral patterns, helping to improve cows’ welfare, which in turn increases productivity.
406 Bovine Facial Recognition Driving Adoption of Artificial Intelligence Applications in Animal Agriculture (PR Newswire)
406 Bovine’s technology uses a database of over six million animal images to train AI algorithms for reliable identification of individual animals, achieving over 99% accuracy across common cattle breeds. This AI solution helps ranchers maintain permanent animal identification, streamline data collection, and make informed management decisions, replacing traditional methods like ear tags and DNA identification.
How CattleEye is improving the lives of farmers and their livestock through pioneering AI (Amazon)
CattleEye, an AI-driven startup, aims to transform animal farming by leveraging AWS AI solutions to enhance animal welfare and farm efficiency. The platform uses deep learning video analytics to detect early signs of lameness in cows, which can significantly impact dairy productivity. By monitoring cows' behavior 24/7, the system not only identifies health issues early but also provides actionable insights to supposedly improve animal comfort and sustainability.
Dairy farming 2.0: Milking the potential of artificial intelligence (AgProud)
Applications of AI include early detection of lameness, mastitis, and optimal breeding times, potentially bringing benefits to animal health and farm productivity. Farmers should consider data privacy, the steep learning curve, and potential sustainability issues when adopting these technologies.
🐷 Pig farming
Deep learning pose detection model for sow locomotion (Guimarães de Paula et al., Scientific Reports)
This paper explores the development of a deep learning model to detect and analyze the locomotion of sows. Using a repository of images and videos of sows with varying locomotion scores, the study aims to create an automated, non-invasive system to identify key points on the sow's body. This model, based on the LEAP architecture, accurately tracks skeletal key points and aims to facilitate the early detection of lameness in sows, ultimately improving animal welfare through precision livestock farming.
Precision livestock farming and technology in pig husbandry (Siegford, Advances in Pig Welfare)
This study examines how PLF technologies can improve pig welfare by enabling continuous, real-time monitoring and management of individual animals through advanced sensors and machine learning. It highlights the need for integrating various PLF technologies into holistic systems that are practical for farmers and address social and ethical concerns, such as human-animal interactions and data privacy. The research underscores the potential for PLF to enhance transparency and traceability in pork production, aligning with public demand for responsible animal care.
🐟 Aquaculture
Using AI to streamline gender-sorting processes in salmon aquaculture (Global Seafood Alliance)
AI is being used to automate the gender-sorting process in salmon aquaculture, improving the accuracy and efficiency of sorting fish by sex. Companies like Aquaticode and Econexus are developing AI-based systems that can handle thousands of fish per hour, potentially enhancing fish quality assessment and production optimization. These systems could also offer additional benefits such as early disease detection and improved animal welfare by reducing stress through accurate and early interventions.
Training AI to detect harmful algal blooms (The Fish Site)
An underwater device using AI, the Imaging FlowCytobot (IFCB), has been operational for over a year at a Scottish salmon farm. It scans water samples for harmful algal blooms (HABs), using lasers and cameras to detect and identify phytoplankton species, providing near real-time alerts to fish farm operators. The device, which has photographed over 38 million images, helps understand seasonal trends and mitigates risks posed by toxic phytoplankton to aquaculture.
Wittaya Aqua’s data-driven AI helps seafood farmers increase aquaculture production (TechCrunch)
Wittaya Aqua's data-driven AI platform aims to help seafood farmers increase profitability, sustainability, and efficiency by consolidating data from across the seafood supply chain. It uses AI and machine learning to forecast animal growth, recommend optimal feed types and quantities, and provide insights based on real-time data. The platform has expanded into Asia, leveraging historical data and environmental factors to improve decision-making for farmers.
[Video] AI-Driven Vertical Shrimp Farm in Singapore Achieves Cost Parity with Coastal Ponds (OneIndia News)
A fully automated indoor shrimp farm is set to go to market this year, featuring 60 tanks managed by a robotic gantry system. The farm uses an AI-controlled computer system and underwater cameras to continuously monitor and optimize shrimp conditions, reducing the time needed to reach market size by half compared to conventional methods. The farm also uses a zero-discharge recirculate system with seaweed filtration, allowing it to operate without antibiotics and maintain water quality.
🐑 Animal farming: General
AI reaches into the meat locker (The Western Producer)
Mode40, a Canadian tech company, is developing an AI-driven system for regulating meat processing temperatures, aiming to enhance food safety and improve meat quality. The system uses AI to monitor and adjust the cooling process of carcasses in real-time by collecting data on temperature, humidity, and wind speed. This technology, funded partly by the Canadian Agri-Food Automation and Intelligence Network, aims to make the meat processing industry more efficient and profitable, with commercialization expected by early 2026.
AI ranch hands and data-driven ‘plant factories’ offer a glimpse into the future of Canadian farming (The Globe and Mail)
AI is increasingly being integrated into Canadian farming, enhancing efficiency and productivity. For example, BETSY, an AI ranch hand developed by OneCup, uses cameras and data to help farmers identify animals, monitor health, and detect distress, significantly improving livestock management.
Feed Industry faces hurdles in AI adoption (All About Feed)
The European feed industry is gradually adopting AI to enhance efficiency in operations such as pelleting processes, which involve compressing and forming feed into pellets. Despite AI's potential to improve monitoring through sensors and cameras, boost efficiency, and ensure regulatory compliance, its adoption is slow due to high costs, data privacy concerns, and conservative practices. Companies like Bühler are using AI to optimize the pelleting process, aiming for consistent pellet quality and increased throughput, though industry-wide adoption remains limited.
Livestock’s problem with precision tech (Manitoba Cooperator)
Precision technology in livestock farming can lead to healthier animals, better efficiency, and a reduced environmental footprint. However, economic risks, lack of knowledge, and the absence of commercially viable technology are major barriers to adoption. Researchers emphasize the need for cost-effective, simple solutions and stress the importance of data ownership for farmers, while current market offerings and complex technology implementations lag behind the demand for practical, real-world applications.
To thrive, ag industry must increase transparency (The Detroit News)
The agricultural industry faces a challenge in gaining consumer trust, with only 24% of U.S. adults and 17% of Generation Z expressing high trust in food production information. Technologies like DNA TraceBack enhance transparency by using DNA-based technology to trace food from farm to table, providing verifiable information for producers, processors, and retailers. Precision livestock farming further supports transparency by using sensors and software to monitor and report detailed data on animal health and resource use.
Review: Exploring the use of precision livestock farming for small ruminant welfare management (Morgan-Davies et al., Animal)
This paper explores how precision livestock farming (PLF) technologies can enhance the welfare management of small ruminants like sheep and goats. The study identifies over 80 main welfare issues and categorizes welfare indicators into four broad areas: weight loss or change in body state, behavioral change, milk yield and quality, and environmental indicators. It highlights 24 potential PLF and innovative technologies that could monitor these welfare indicators and improve animal welfare management across diverse farming systems.
The livestock farming digital transformation: implementation of new and emerging technologies using artificial intelligence (Fuentes et al., Cambridge University Press)
This paper highlights the use of biometric techniques, such as remote sensing and computer vision, to non-invasively monitor and assess the health and welfare of farmed animals. Specific applications include real-time monitoring of core body temperatures using implantable biosensors, tracking heart and respiratory rates with infrared thermography, and using deep learning models to predict stress responses and disease outbreaks. These advancements aim to enhance productivity and animal welfare, with a focus on improving accuracy and reducing the stress of traditional monitoring methods.
🦘 Wild animals
Fighting fire from space: How AI could save millions of Australian animals from deadly wildfires (Discover Wildlife)
Scientists at the University of South Australia have developed AI-enabled software for cube satellites (cubesats) to detect wildfires in under an hour, potentially saving millions of animals. During the 2019-20 Australian wildfires, an estimated 143 million mammals, 2.46 billion reptiles, 181 million birds, and 51 million frogs were impacted. Early detection using this technology could significantly reduce such devastating effects on wildlife, improving their survival rates during future fire events.
How artificial intelligence can help prevent illegal wildlife trade (Phys.org)
AI can help prevent illegal wildlife trade by using bioacoustic technology to identify endangered bird species based on their vocalizations. Researchers have developed a neural network model that can recognize specific bird species from their calls with over 90% accuracy, which is crucial for identifying species in wildlife markets where visual identification can be challenging. This technology can be deployed in a smartphone app, allowing law enforcement officers, conservationists, and even citizen scientists to quickly and accurately identify birds in real time, helping to enforce wildlife protection laws and combat illegal trafficking.
AI is rapidly identifying new species. Can we trust the results? (Live Science)
AI is revolutionizing how scientists identify new animal species and monitor ecosystems by analyzing data from smartphones, camera traps, and automated monitoring systems. AI models, trained on images and DNA data, can accurately recognize species and flag potential new species for further study by biologists. However, there are doubts about the effectiveness of AI due to its reliance on existing data, which has significant gaps, and the lower accuracy when identifying species from low-resolution images alone.
Tadoba adds 3 villages to AI-based system to curb man-animal conflict (Times of India)
The Tadoba-Andhari Tiger Reserve (TATR) has expanded its AI-based system to reduce human-animal conflicts by including three more villages. This system uses AI to monitor wildlife behavior and send early warnings about tigers, bears, and leopards approaching villages, thereby reducing attacks on humans and farmed animals. The initiative also includes thermal drones and AI cameras for enhanced real-time surveillance and protection.
🍔 Alternative proteins
The scientific future of pet food: Leveraging alternative proteins (Pet Food Processing)
Bond Pet Foods uses AI and precision fermentation to create sustainable, animal-free proteins for pet food. They bioengineer microorganisms to produce proteins identical to those in animal meat, ensuring nutritional completeness and sustainability. This approach aims to reduce reliance on traditional animal proteins and offers a consistent and eco-friendly alternative for pet nutrition.
MOA Foodtech Raises Funds to Scale AI-Driven Production of Highly Nutritious Microbial Proteins (Vegconomist)
MOA Foodtech, a Spanish company specializing in alternative proteins, is expanding its AI-driven production of microbial proteins. Using precision fermentation and agrifood side streams, MOA creates sustainable proteins with excellent nutritional qualities for applications such as plant-based meat, cheese, sauces, and bread. The investment will enhance MOA's AI platform, Albatros, streamline new ingredient creation, and support a global market launch.
Leveraging AI in cultivated seafood research (University of Waterloo)
Two researchers from the University of Waterloo have received over $700,000 in grants to scale their research on cultivated fish. Using AI and computational modeling, their research aims to improve understanding of fish cell growth, gene regulation, and cell differentiation, with the goal of producing sustainable cultivated seafood.
Proxy Foods CEO on AI’s potential in alt-protein market, creating ‘Canva for the food scientist' (Food Navigator USA)
At the Future Food-Tech Alternative Proteins event, Proxy Foods' CEO Panos Kostopoulos discusses using AI to overcome challenges in the alt-protein market. Proxy Foods' AI technology helps food developers by providing insights into flavor, nutrition, cost, and regulatory information. The goal is to create a user-friendly tool for food scientists to streamline and innovate product formulations efficiently.
Shaping the Future of Food with AI and Novel Ingredients (Vegconomist)
At the IFT FIRST 2024 conference, AI was highlighted for its transformative role in the food industry. It is being used to optimize recipes by analyzing data to enhance taste, texture, and nutrition, and to streamline supply chains by identifying cost-saving opportunities and reducing waste. AI can also support market analysis to predict consumer trends, accelerating product development and innovation cycles.
🧑💻 Speciesism in AI models
AI Animal Welfare: Creating an Animal-Friendly Model (AI Safety Fundamentals)
The winner of the 'AI strategy' prize on AI Safety Fundamentals’ AI Alignment (March 2024) course was a project titled ‘AI Animal Welfare: Creating an Animal-Friendly Model’. Building on the insights from "The Case for Animal-Friendly AI," this project seeks to replicate and extend their evaluation methods by developing an AI model that better considers animal welfare. The objective is to gain hands-on experience with large language models, including their prompting, evaluation, and fine-tuning, while also raising awareness about the intersection of animal advocacy and artificial intelligence.
Multilingual Trolley Problems for Language Models (Jin et al., Arxiv)
This paper evaluates the moral decision-making of large language models (LLMs) across 100+ languages using adapted trolley problem scenarios. One aspect of the study involves decisions affecting humans versus animals, exploring the models' preferences for sparing humans over companion animals such as dogs and cats. The findings indicate that LLMs generally prefer to save humans over animals, reflecting human-like moral tendencies, but the degree of this preference varies across different languages and cultures.
🤖 Digital minds
AI Welfare Debate Week (EA Forum)
July 1–7 was ‘AI Welfare Debate Week’ on the EA Forum, with users debating the statement ‘AI Welfare should be an EA priority’. Proponents argued that ensuring AI welfare could create vast value by safeguarding the well-being of potentially trillions of artificial minds, far outnumbering biological ones. Critics argued that this diverts resources and attention from more pressing issues like AI safety and alignment. There was disagreement on whether to act now to shape the field or wait until we better understand AI. The discussion also touched on practical challenges, like how to determine what constitutes welfare for an AI, and whether early interventions might inadvertently increase risks due to our limited understanding.
What Do People Think about Sentient AI? (Anthis et al., Arxiv)
The Artificial Intelligence, Morality, and Sentience (AIMS) survey is presented as the first nationally representative survey on sentient AI. Analyzing one wave of data from 2021 and two from 2023 (total sample size = 3,500), the survey found that mind perception and moral concern for AI well-being were higher than expected in 2021 and had significantly increased by 2023; 71% agree that sentient AI deserve respect, and 38% support legal rights for AI. Additionally, there is growing opposition to advanced AI technologies, with 63% favoring a ban on smarter-than-human AI and 69% supporting a ban on sentient AI. The expected timelines for these technologies are short, with a median forecast of five years for sentient AI and two years for artificial general intelligence. This was also covered by New Scientist."CyberOctopus":
New AI explores, remembers, seeks novelty, overcomes obstacles (University of Illinois)
Researchers at the University of Illinois Urbana-Champaign have developed an AI system, dubbed "CyberOctopus," which incorporates associative learning rules inspired by sea slug neural circuits and enhanced episodic memory modeled on octopuses. This AI can navigate new environments, seek rewards, map landmarks, and overcome obstacles by using a ‘Feature Association Matrix’ that simulates the hippocampus, allowing it to learn and adapt in real time.
🐬 …and more
Report on the Use of AI in Animal Advocacy (Open Paws)
Open Paws surveyed 194 participants from 142 animal advocacy organizations about AI use. 72.2% of respondents had positive personal attitudes towards AI, and 50.5% currently use AI for writing or editing content. while 60.3% cited lack of technical expertise as the main barrier to adoption. The main challenges cited include lack of technical expertise, concerns about AI accuracy, and limited budgets. To address these issues, Open Paws recommends providing technical training, developing custom AI models tailored for animal advocacy, and establishing funding assistance programs to support AI integration in the sector.
From MilkingBots to RoboDolphins: How AI changes human-animal relations and enables alienation towards animals (Bossert et al., Humanities and Social Sciences Communications)
Disruptive technologies, including AI and robotics, have significant potential to alter human-animal relations. This paper aims to both descriptively analyze the impact of these technologies on human-animal interactions and normatively introduce a non-anthropocentric perspective that acknowledges the moral significance of these relations. The study focuses on how automation in agriculture and the replacement of biological animals with AI-driven entities in various contexts, such as zoos and laboratories, influence these interactions and highlights the ethical considerations involved.
Can AI replace animal testing? (Polytechnique)
Researchers are exploring artificial intelligence as a potential alternative to animal testing in science, motivated by ethical concerns and the limitations of using animals to study human biology. Some new approaches include "digital twin" systems (computer simulations that mimic human tissues or organs), organoids (miniature 3D cell cultures grown to resemble simplified versions of organs), and AI algorithms that predict how genes interact to determine cell functions. However, researchers disagree about how much to value animals’ wellbeing and how well AI can truly replicate complex biological systems, making it difficult to completely eliminate animal testing currently.
📨 That’s it for this edition - as always, please feel free to get in touch contact@aiforanimals.org with any ideas and feedback!