Adam Workman’s Post

Oxford Semantic Technologies, a University of Oxford spinout specialising in knowledge graph and reasoning technology, has been acquired by Samsung Electronics.   Read ➡️ https://lnkd.in/djww8ixH     Founded in 2017 by three University of Oxford professors — Ian Horrocks, Boris Motik and Bernardo Cuenca Grau, OST supports accurate AI applications through sophisticated inference techniques.    OST's innovative approach, Knowledge Representation and Reasoning (KRR), allows AI to make logical and explainable decisions by combining data with expert knowledge, unlike machine learning, which relies on finding patterns in large datasets.    Mairi Gibbs, CEO at Oxford University Innovation, said: "Oxford Semantic Technologies represents a significant advancement in the field of AI, showcasing how academic research can be transformed into real-world applications that drive industry innovation.      “Their acquisition by Samsung Electronics underscores the commercial value and potential of their technology in multiple sectors. We see this catalysing further innovation and partnership in Oxford’s Deep Tech ecosystem between our academic and research community, industry and investors.”    ➡️ https://lnkd.in/djww8ixH   #innovation #artificialintelligence  #knowledgegraphs #oxfordsemantictechnologies #AI #reasoningtechnology  

🤖 Machine learning models have been shown to be prone to memorize their training data and this can cause problems if the training data contains personal data or other sensitive information. To avoid this problem, Professors Antti Honkela and Samuel Kaski are training an AI assistant for privacy-preserving machine learning models. They are collaborating with researchers in RIKEN Center for Computational Science and Tokyo Institute of Technology in Japan. This is the 6th blog in series that represents collaboration projects between Finland-Japan and Finland-Colorado that use the LUMI supercomputer to address global challenges and top-level research topics in different fields. 👉 Read the blog post on our website: https://lnkd.in/dZPCq7PG #MachineLearning #HPC #InternationalCollaboration #LUMIsupercomputer #AI

🚀 **Exciting News from the University of Oxford!** 🎓 The University of Oxford has just unveiled a groundbreaking AI-powered research platform called "Oxford Research Analytics." 🤖💡 This innovative tool leverages advanced machine learning algorithms to sift through vast datasets, uncovering patterns, trends, and correlations that are crucial for informed research decisions. 🔍 **What's in It for Researchers?** - **Enhanced Efficiency:** Speed up your research processes significantly. - **Improved Accuracy:** Reduce errors and increase reliability in findings. - **Faster Discoveries:** Accelerate the path to new knowledge and insights across various fields. This platform is set to transform how research is conducted, making it smarter, faster, and more efficient. 💼📈 Stay tuned for more exciting updates and breakthroughs! 👉 Follow [Your Page or Name] for more well-curated tech and education insights! Read more about this innovative platform here: [University World News Article](https://lnkd.in/eRcWsr7J) 🌐 #UniversityOfOxford #AI #ResearchInnovation #MachineLearning #TechNews #HigherEducation #DataScience #Innovation

In a groundbreaking study that could redefine the path of AI development, researchers from the University of Copenhagen (Københavns Universitet) have mathematically proven that creating fully stable machine learning algorithms for complex problems is an unattainable goal. This revelation exposes the Achilles heel of AI – its vulnerability to the unpredictability of real-world scenarios. Imagine an automated vehicle thrown off course by a mere sticker on a road sign; this kind of sensitivity challenges the stability of even the most advanced AI systems. Professor Amir Yehudayoff and his team are not just pointing out the limitations; they are setting the stage for a deeper understanding of AI's weaknesses. Their research is not about halting progress but enhancing the robustness of AI systems. By developing a language to discuss these weaknesses, they aim to guide future testing protocols and algorithm development. This study is a crucial step in bridging the gap between AI's current capabilities and its potential to handle the complexities of real life. This research, approved for publication at a leading conference on theoretical computer science, underscores a vital lesson: AI, despite its rapid advancement, still cannot replicate the adaptability of human intelligence. It prompts a reflection on the future of AI development: How can we harness the power of AI while remaining cognizant of its fundamental limitations, and what new strategies can we develop to create more resilient and adaptable algorithms? Read the full story on SciTech Daily: https://lnkd.in/gz3eHBve #AI #ethics #Innovation #challenge #Research ---- If you enjoyed this content, be sure to download my new app for a unique experience beyond your traditional newsletter - you can have real-time insights, recommendations and conversations with my digital twin via text, audio or video in 28 languages! Go to app.thedigitalspeaker.com and sign up to take our connection to the next level!

Stanford University just released its AI Index Report 2024—its most comprehensive to date—covering technical advancements in AI, public perceptions of the technology, and the geopolitical dynamics surrounding its development. Some of the report's key 🔑 takeaways include: • AI beats humans on some tasks, but not on all. 🤖 AI has surpassed human performance on several benchmarks, including some in image classification, visual reasoning, and English understanding. Yet it trails behind on more complex tasks like competition-level mathematics, visual commonsense reasoning, and planning. • Industry continues to dominate frontier Al research. 💪 In 2023, industry produced 51 notable machine learning models, while academia contributed only 15. There were also 21 notable models resulting from industry-academia collaborations in 2023, a new high. • The United States leads China, the EU, and the U.K. as the leading source of top Al models. 🇺🇸 In 2023, 61 notable Al models originated from U.S.-based institutions, far outpacing the European Union's 21 and China's 15. Read the report here: 👉 https://buff.ly/43Y9T6Q It is well worth your time! #AI

Respondents wanted! Please participate in the following survey, aiding my Master Thesis for the MSc Business Information Management at RSM, EUR. Directly to the survey: https://lnkd.in/et6SFkWr Survey via SurveySwap: https://lnkd.in/eiAVeXF7 The survey is mobile-friendly, but a laptop may provide more comfort. Collaborative AI: humans and AI working together, successfully. The possibilities with AI are endless, and we have only just started benefiting from its use in our day-to-day lives. Unfortunately, the collaboration between humans and AI is not always beneficial. There is an academic gap in identifying and predicting exactly when collaborative AI is going to be succesful and when it is not. With this thesis research, I hope to narrow the gap a little more by investigating the underlying mechanisms that may affect the relationship between humans and AI. Thank you for your participation!

Participants of the CAS ETH in AI and Software Development learn to apply key computer science concepts in specific industry settings with three study modules: 💡 Programming with Python: learn to code with regular one-on-one meetings for support. 💡Software Engineering Fundamentals: an end-to-end introduction to the core techniques used for building quality software. 💡Building ML/AI applications: fundamental training in areas of machine learning that enable AI applications with an introduction to LLMs and NLP as an industry relevant example. The next application period is open from 1-30 November 2024. #lifelonginspiration #continuingeducation #AI #machinelearning #development

🚀🌟 Join Us for the Next Spotlight Seminar on AI! 🌟🚀 The Italian Association for Artificial Intelligence, in collaboration with the AISB (Society for the Study of Artificial Intelligence and Simulation of Behaviour, UK), is excited to present our next seminar in the Spotlight Seminars on #AI initiative! 🔍🤖 📅 Date: June 20 2024 🕔 Time: 5:00 PM (CET) 📽️ Live Streaming: https://lnkd.in/dgKJ93yY Title: Symbolic Reasoning for Large Language Models Speaker: Guy Van den Broeck, UCLA 🌟 About the Seminar: 🌟 As the world looks to #ArtificialIntelligence to solve complex societal challenges, it's crucial to understand how these systems can develop reasoning capabilities similar to human experts. In this seminar, we delve into the intriguing domain of neurosymbolic reasoning for #LargeLanguageModels. Guy Van den Broeck will explore whether #AI can learn to reason from data and how to design next-generation AI techniques that seamlessly integrate learning and reasoning using probabilistic circuits. 👨🏫 About the Speaker: 👨🏫 Guy Van den Broeck is an Associate Professor and Samueli Fellow at UCLA's #ComputerScience Department, leading the StarAI lab. His notable achievements include awards from prestigious conferences like #AAAI, #UAI, #KR, and #OOPSLA, and honors such as the #NSF CAREER award, a Sloan Fellowship, and the IJCAI-19 Computers and Thought Award. Join us for an insightful exploration of the future of AI and its reasoning capabilities! Don't miss out! ✨ #AISeminar #AIResearch #NeurosymbolicReasoning #MachineLearning #TechTalk #UCLA #AICommunity #AIInnovation #SpotlightSeminars

📰 [YG's Tech News!] 🔍 New Research Finds Removing Unwanted Data Can Hurt AI Model Performance! De-learning techniques, designed to remove specific undesirable information from generative AI models, often lead to significant drops in model performance. Researchers from the University of Washington, Princeton University, the University of Chicago, the University of Southern California, and Google found that current de-learning methods are not ready for real-world applications. Researcher Shiqi Wang emphasizes that effective methods for making models “forget” specific data without significantly reducing their utility are currently lacking. Further research is needed to address this technical challenge. More Content👉https://lnkd.in/gq9zwCrE #ArtificialIntelligence #DataRemoval #DeLearningTechniques #AIperformance #TechnicalResearch #GenerativeAI #DataPrivacy #CopyrightIssues #ModelTraining #TechInnovation #FutureResearch #technology #technews

With the new Certificate of Advanced Studies in AI and Software Development, ETH Zurich prepares leaders for the future of Artificial Intelligence, machine learning and other digital technologies. https://lnkd.in/e8Yi43Ts ETH Zurich | School for Continuing Education #AI #MachineLearning #CAS #ArtificialIntelligence