Feedback is New Scientist's popular sideways look at the latest science and technology news. You can submit items you believe may amuse readers to Feedback by emailing feedback@newscientist.com Recently, Feedback was delighted to peruse the raciest conference invitation we have ever received. We get a lot of conference invites from organisers labouring under the delusion we are doing something akin to science journalism, and they are mostly a little prosaic: what's new in G-protein signalling, more findings about the biology of molluscs, that kind of thing. Here is the opening line: "From its groundbreaking inception in London to its spectacular evolution in the vibrant heart of China, the Love and Sex with Robots Conference is gearing up for its most thrilling chapters yet: its landmark 12th International edition, scheduled for June 2026."

The number of scientific articles published in the domain of food safety has consistently been increasing over the last few decades. It has therefore become unfeasible for food safety experts to read all relevant literature related to food safety and the occurrence of hazards in the food chain. However, it is important that food safety experts are aware of the newest findings and can access this information in an easy and concise way. In this study, an approach is presented to automate the extraction of chemical hazards from the scientific literature through large language models. The large language model was used out-of-the-box and applied on scientific abstracts; no extra training of the models or a large computing cluster was required. Three different styles of prompting the model were tested to assess which was the most optimal for the task at hand. The prompts were optimized with two validation foods (leafy greens and shellfish) and the final performance of the best prompt was evaluated using three test foods (dairy, maize and salmon). The specific wording of the prompt was found to have a considerable effect on the results. A prompt breaking the task down into smaller steps performed best overall. This prompt reached an average accuracy of 93% and contained many chemical contaminants already included in food monitoring programs, validating the successful retrieval of relevant hazards for the food safety domain. The results showcase how valuable large language models can be for the task of automatic information extraction from the scientific literature.

Part of the motivation for deploying a neural network arises from the belief that algorithms that learn features/representations generalize better than algorithms that do not. We try to give some mathematical ballast to this notion by studying a data model where, at an intuitive level, a learner succeeds if and only if it manages to learn the correct features. The data model itself attempts to capture two key structures observed in natural data such as text or images. First, it is endowed with a latent structure at the patch or word level that is directly tied to a classification task. Second, the data distribution has a long-tail, in the sense that rare and uncommon instances collectively form a significant fraction of the data. We derive non-asymptotic generalization error bounds that quantify, within our framework, the penalty that one must pay for not learning features.

For a century, researchers have tracked genetic traits to find out which cattle produce more and better milk and meat. Now, two University of Florida scientists will use artificial intelligence to analyze millions of bits of genetic data to try to keep cattle cooler and thus, more productive. Raluca Mateescu, a UF/IFAS professor, and Fernanda Rezende, a UF/IFAS assistant professor – both in animal sciences -- gather hundreds of thousands of pieces of information about cattle genetic traits. They plan to use UF's supercomputer, the HiPerGator, to analyze that data. With the information Mateescu and her team get from the HiPerGator, they can give ranchers better recommendations on which animals to keep and breed for improved quantity of beef and dairy.

Co-founder and chief strategy officer David Hunt says their technology allows farmers to see what is happening on their dairy "in high resolution in real time…without anyone needing to go into the barn." Based in California, Canada and Ireland, the company launched their first product in late January. Alus Nutrition focuses on "all things related to feed bunk management," according to portfolio growth lead Tyler Bramble. This includes when feed is delivered to cows or when the cows have cleaned out the feed and need more. Cainthus' smart cameras monitor cows, while their software interprets what the cameras see.

Deep tech is the newest catchphrase in the Indian startup ecosystem. A bunch of homegrown companies are using new-age technologies like artificial intelligence, machine learning, data analytics, cloud, and the internet-of-things (IoT) to solve real-world problems, and essentially, alter the way humans lead daily lives. On Day One of TechSparks 2019, YourStory's flagship annual conference, a panel of founders, investors, and technical heads gathered to take stock of the evolution of the local deep tech startups ecosystem. Swapan Rajdev, Co-Founder and CTO, Haptik (maker of AI chatbots, recently acquired by Reliance Jio) elaborated on how the growth of AI has spurred new jobs and roles. Gone are the days when Indian companies failed to make a mark in hardware.

Jin Kawaguchiya gave up a career in finance to help revive Japan's ailing dairy industry, one robot at a time. In a country that relies increasingly on imported foods like cheese and butter, Japan's milk output tumbled over two decades, touching a 30-year low in 2014. Costs rose faster than prices as the economy stagnated, eroding profit, and aging farmers quit the business because they could not find enough young people willing to take on the hard labor of tending to cows every day. But technology is altering that dynamic. On Hokkaido, Japan's top dairy-producing region, Kawaguchiya transformed the 20-cow farm he inherited from his father-in-law 16 years ago into Asia's largest automated milking factory.

The University of Nottingham is developing an artificially-intelligent sensor system to clean food manufacturing equipment more precisely, which could save 100m a year for the UK industry alone. This revolutionary AI-driven monitoring system could lead to greater production capacity and therefore cheaper food prices for consumers. Food and drink production is the largest manufacturing sector in Britain and the highest industrial user of water at approximately 430 million litres a day. As current technologies cannot accurately determine exactly how dirty food and drink processing equipment is inside, cleaning can last up to five hours a day - to minimise food safety risks. Cleaning accounts for 30 per cent of energy and water use and leads to excessive productivity down time and over-use of chemicals, at huge cost to manufacturers and the environment.

Cows and robots go together. Throw in facial recognition software, and it's the perfect trifecta. This is because cows are happier when they are not around people, since they identify humans as predators. Using facial recognition software to count a herd, or signal when a cow is sick or injured or not eating, is another way to keep humans out of the pastures as much as possible and keep cows happier and more productive. "No prey animal never wants to see a predator. The less they see the happier they are. A cow doesn't know what a robot is, so they aren't scared of it," said David Hunt, co-founder of Cainthus, a company digitizing agricultural practices, speaking at an Alltech conference in Lexington, Ky.