{"id":10499,"date":"2025-03-31T10:00:00","date_gmt":"2025-03-31T08:00:00","guid":{"rendered":"https:\/\/haimagazine.com\/uncategorized\/how-its-done-in-science-4\/"},"modified":"2025-06-26T15:40:06","modified_gmt":"2025-06-26T13:40:06","slug":"how-its-done-in-science-4","status":"publish","type":"post","link":"https:\/\/haimagazine.com\/en\/hai-magazine-4\/how-its-done-in-science-4\/","title":{"rendered":"\ud83d\udd12 How it&#8217;s done in science (4)"},"content":{"rendered":"<h4 class=\"wp-block-heading\"><strong>Change exactly what I want<\/strong><\/h4><p class=\"wp-block-paragraph\">Marek \u015amieja is researching new methods of controlling the data generation process in artificial intelligence models. His project focuses on developing deep conditional generative models that allow to precisely shape the outcomes created by algorithms like GANs and VAE. These models are used in many fields, from explaining AI decisions, through generating counterfactual examples, to creating chemical molecules with specific properties or realistically completing missing parts of images.   <\/p><p class=\"wp-block-paragraph\">One of the key challenges in this area is the high computational complexity and the need to adjust models to new data, which often requires resources comparable to those available to global tech companies. In order to address that, \u015amieja\u2019s team is working on plugin networks that will be able to modify data based on existing models, thus reducing computational costs.  <\/p><p class=\"wp-block-paragraph\">This project has broad practical implications \u2013 not only in the field of machine learning, but also in areas such as pharmacology, computer graphics, and AI interpretability. Research results can help you create more transparent artificial intelligence systems, support designing new drugs, and improve generative technologies in various industries. Additionally, to facilitate adapting their solutions, researchers plan to share implementations of their algorithms on open-source platforms.  <\/p><h4 class=\"wp-block-heading\"><strong>On the trail of bacteria at an accelerated pace<\/strong><\/h4><p class=\"wp-block-paragraph\">A team of scientists from the Jagiellonian University, led by Monika Brzychczy-W\u0142och and Bartosz Zieli\u0144ski, has developed a groundbreaking Ai-driven method for fast microbiological diagnostics. By using deep neural networks, the system can instantly recognize bacteria and fungi based on microscopic images \u2013 the whole process takes only a minute. This is a huge change compared to traditional methods like phenotypic cultures or molecular tests, which require expensive equipment and take many      <\/p><p class=\"wp-block-paragraph\">hours, even days. The new technology not only speeds up diagnosis, but also increases its accuracy, which is crucial in quickly treating infections, especially in cases like sepsis, where time plays a major role. Thanks to this method, doctors can implement the appropriate treatment more quickly, which can save lives. This solution also has potential applications beyond medicine, for example in the pharmaceutical industry, food safety monitoring, and scientific research, making it an incredibly promising innovation in the field of microbiological diagnostics.   <\/p><h4 class=\"wp-block-heading\"><strong>Scanning the darkness<\/strong><\/h4><p class=\"wp-block-paragraph\">Monday morning, 7 am. Queue of cases to review \u2013 273 instances. Jakub knows he won&#8217;t manage to analyze all of them before the next ones join the list. Behind each case, there may be a child who needs help. The system works only during team working hours, but criminals never sleep. Time is your biggest enemy.     <\/p><p class=\"wp-block-paragraph\">Every day around the world, millions of new photos and videos depicting the sexual exploitation of children appear. In 2023, online platforms reported almost 55 million such materials. Fighting this crime is a constant race against time where every second can determine the victim\u2019s fate. The problem is the scale \u2013 analyzing reports requires a huge effort from people who have to manually review hundreds of thousands of files. The estimations show that even 20% of them are content that has already been identified but still circulates online.     <\/p><p class=\"wp-block-paragraph\">Artificial intelligence may be the solution. Since 2020, the Dy\u017curnet.pl team, operating within NASK, has been developing AI tools to support the classification of CSAM (Child Sexual Abuse Material). Computerized vision systems can automatically identify visual content and classify it as illegal. This way, analysts can focus on new cases, while the algorithm will take care of detecting duplicates. This allows to reduce the response time and reach children in need faster.     <\/p><p class=\"wp-block-paragraph\">The next step will be developing multimodal models that combine image analysis with natural language processing. Thanks to this system, not only will it detect CSAM materials, but it will also provide information relevant to the investigation. In the future, AI is supposed to work 24\/7, analyzing reports as they come in.  <\/p><p class=\"wp-block-paragraph\">When Jakub logs into the system on Monday morning, AI will suggest which reports require immediate attention, which will allow him to focus on what&#8217;s most important \u2013 protecting children from further harm. <\/p><h4 class=\"wp-block-heading\"><strong>AI, law, and surveillance: who controls whom?<\/strong><\/h4><p class=\"wp-block-paragraph\">Marcin Rojszczak from the Warsaw University of Technology is conducting research on the use of artificial intelligence in untargeted electronic surveillance and its compliance with the European human rights protection laws.  Untargeted surveillance, also known as mass surveillance, involves collecting a wide range of data about a large number of people, without pointing out specific suspects beforehand. Actions like this are aimed at detecting potential threats to national security, but they are controversial due to the extent of interference with citizens&#8217; privacy.<\/p><p class=\"wp-block-paragraph\">The project\u2019s goal is to develop legal frameworks for AI-based systems in this type of surveillance, so their operation is in line with the principles of human rights protection in the European Union. Research includes analysis of regulations and practices in countries like Germany, France, Belgium, United Kingdom and United States. It also takes into account regulations and rulings of European legal institutions such as the Court of Justice of the European Union and the European Court of Human Rights.<\/p><p class=\"wp-block-paragraph\">This project combines legal and technical methods \u2013 in addition to analyzing regulations, it also includes experimental research on the operation of AI technology in surveillance. It fits into the current debate on algorithmic surveillance, that is, using artificial intelligence to process huge amounts of data. The results of these studies can help improve the protection of citizens&#8217; privacy and contribute to creating new legal regulations at the national and EU levels.<\/p>","protected":false},"excerpt":{"rendered":"<p>Quarterly overview of AI projects in science.<\/p>\n","protected":false},"author":5,"featured_media":9748,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"_acf_changed":false,"rank_math_lock_modified_date":false,"footnotes":""},"categories":[783,786,673,781,674],"tags":[],"popular":[],"difficulty-level":[36],"ppma_author":[343,492],"class_list":["post-10499","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-ai-industry","category-education-science","category-hai-magazine-4","category-hai-premium","category-issue-4","difficulty-level-easy"],"acf":[],"authors":[{"term_id":343,"user_id":5,"is_guest":0,"slug":"inez-okulska","display_name":"dr Inez Okulska","avatar_url":"https:\/\/secure.gravatar.com\/avatar\/479f0f5551a6bf974825e84cfe39166b785e5cd476e583be6a22279c2c379917?s=96&d=mm&r=g","first_name":"dr Inez","last_name":"Okulska","user_url":"","job_title":"","description":"Redaktor naczelna hAI Magazine, badaczka i wsp\u00f3\u0142autorka modeli AI (StyloMetrix, PLLuM), wyk\u0142adowczyni, Top100 Woman in AI in PL"},{"term_id":492,"user_id":184,"is_guest":0,"slug":"anita-ciesielska","display_name":"dr hab. Anita Ciesielska","avatar_url":{"url":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/04\/Yellow-and-Black-Simple-Professional-LinkedIn-Profile-Picture.png","url2x":"https:\/\/haimagazine.com\/wp-content\/uploads\/2025\/04\/Yellow-and-Black-Simple-Professional-LinkedIn-Profile-Picture.png"},"first_name":"","last_name":"","user_url":"","job_title":"","description":"Ekspertka AI w nauce i edukacji | Nauczycielka akademicka | Badaczka | Wydzia\u0142 Biologii i Ochrony \u015arodowiska | Uniwersytet \u0141\u00f3dzki"}],"_links":{"self":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/10499","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/users\/5"}],"replies":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/comments?post=10499"}],"version-history":[{"count":1,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/10499\/revisions"}],"predecessor-version":[{"id":10500,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/posts\/10499\/revisions\/10500"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media\/9748"}],"wp:attachment":[{"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/media?parent=10499"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/categories?post=10499"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/tags?post=10499"},{"taxonomy":"popular","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/popular?post=10499"},{"taxonomy":"difficulty-level","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/difficulty-level?post=10499"},{"taxonomy":"author","embeddable":true,"href":"https:\/\/haimagazine.com\/en\/wp-json\/wp\/v2\/ppma_author?post=10499"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}