Kaare Brandt Petersen

The next few months Tina Westergaard Milbak will be visiting Laerdal Copenhagen in a research exchange as part of her Ph.D. studies at the Department of Computer Science at University of Copenhagen.   In her research, Tina is investigating generative AI for speculative design options in product development and we at Laerdal are very eager to discuss our use of AI and curious to see how it all helps in our mission to help saving lives. #laerdalCPHlife #helpingsavelives #ai

108

3 Kommentarer

🤖 New Study on The Implementation of LLM in Design Processes Daniel Nygård Ege, Henrik H. Øvrebø, Vegar Stubberud, Martin Francis Berg, Christer W. Elverum, Martin Steinert, Håvard Vestad (2024) ChatGPT as an inventor: Eliciting the strengths and weaknesses of current large language models against humans in engineering design Link: https://lnkd.in/dZBxWe_s Abstract: This study compares the design practices and performance of ChatGPT 4.0, a large language model (LLM), against graduate engineering students in a 48-hour prototyping hackathon, based on a dataset comprising more than 100 prototypes. The LLM participated by instructing two participants who executed its instructions and provided objective feedback, generated ideas autonomously and made all design decisions without human intervention. The LLM exhibited similar prototyping practices to human participants and finished second among six teams, successfully designing and providing building instructions for functional prototypes. The LLM's concept generation capabilities were particularly strong. However, the LLM prematurely abandoned promising concepts when facing minor difficulties, added unnecessary complexity to designs, and experienced design fixation. Communication between the LLM and participants was challenging due to vague or unclear descriptions, and the LLM had difficulty maintaining continuity and relevance in answers. Based on these findings, six recommendations for implementing an LLM like ChatGPT in the design process are proposed, including leveraging it for ideation, ensuring human oversight for key decisions, implementing iterative feedback loops, prompting it to consider alternatives, and assigning specific and manageable tasks at a subsystem level. Keywords: #ArtificialIntelligence #AI #ProductDevelopment #Prototyping #LargeLanguageModel #EngineeringDesign #ChatGPT #LLM

3

1 Kommentar

How should we tackle the challenge of overly dense knowledge graphs? Sria Louis proposes that we assess the relevance of each edge by measuring Pointwise Mutual Information (PMI) between its occurrence and the target variable.

24

Financial Statement Analysis with Large Language Models Alex Kim Maximilian Muhn Valeri Nikolaev Abstract We investigate whether an LLM can successfully perform financial statement analysis in a way similar to a professional human analyst. We provide standardized and anonymous financial statements to GPT4 and instruct the model to analyze them to determine the direction of future earnings. Even without any narrative or industry-specific information, the LLM outperforms financial analysts in its ability to predict earnings changes. The LLM exhibits a relative advantage over human analysts in situations when the analysts tend to struggle. Furthermore, we find that the prediction accuracy of the LLM is on par with the performance of a narrowly trained state-of-the-art ML model. LLM prediction does not stem from its training memory. Instead, we find that the LLM generates useful narrative insights about a company's future performance. Lastly, our trading strategies based on GPT's predictions yield a higher Sharpe ratio and alphas than strategies based on other models. Taken together, our results suggest that LLMs may take a central role in decision-making. https://lnkd.in/eAP7JVuq

1

#ResearchYouShouldKnow Pauline Mørner and Øivind Ullaland of University of Stavanger in Norway have just completed a very interesting Masters thesis on the subject of formal and informal organization. The title is "Intra-Organizational Functioning and Autonomy Satisfaction in a Matrix Organization: A qualitative study in an IT-consultancy company." The thesis explores "...the interplay between the formal and informal organization within an IT consultancy firm in Norway, which emphasizes autonomy. Using a qualitative case study approach, we found that a strong informal organization significantly impacts adherence to formal structures, potentially leading to conflicting directions. Moreover, deficiencies in the formal coordination structure challenge functioning of the informal organization. At the same time, it is evident that accommodating the psychological need for autonomy within the informal organization is crucial. This underscores the importance of a balanced approach, where both the formal and informal organization are aligned to ensure optimal intra-organizational functioning." I look forward to digging into this thesis in more depth in the coming week. Until then, congrats to Pauline and Oivind for completing your thesis together, and to their thesis advisor Espen Olsen, as well. #networkanalysis #organizationalnetworkanalysis #socialnetworkanalysis #organizationalstructure #matrix #matrixorganization #organizationdesign #organizationalbehavior #informalorganization #autonomy #formalorganization #hierarchy #coordination #IT #consulting

6

1 Kommentar

🎙️ #TheDataDemocracyPodcast Episode 29 is out now! 𝐌𝐞𝐬𝐡𝐞𝐝 𝐔𝐩 𝐓𝐞𝐫𝐦𝐬: 𝐀𝐥𝐢𝐠𝐧𝐢𝐧𝐠 𝐰𝐢𝐭𝐡 𝐃𝐚𝐭𝐚 𝐌𝐞𝐬𝐡, 𝐃𝐞𝐛𝐚𝐭𝐢𝐧𝐠 𝐭𝐡𝐞 𝐃𝐢𝐜𝐭𝐢𝐨𝐧𝐚𝐫𝐲 ✔️📖 In this podcast episode, host Ole Olesen-Bagneux sits down with Jon Cooke, the founder of Dataception and a seasoned pioneer in the realms of data architecture and AI. He takes us through his diverse career journey, starting from his early days in virtual reality and aerospace to spearheading innovations in data strategies and AI at major consultancies and enterprises. Jon delves into the practical challenges of implementing data mesh architectures, sharing his nuanced views on the transition from vision to execution. He critiques common terminologies like 'data products’, proposing clearer communication to bridge the gap between technical teams and business stakeholders. Throughout the conversation, he emphasizes the importance of composable data architectures and the evolving role of generative AI in transforming how businesses handle and leverage unstructured data. As the discussion unfolds, Jon highlights his current work on cutting-edge projects that integrate AI with traditional data processes, reshaping risk assessment and decision-making frameworks. His insights offer a glimpse into the future of data management, where AI and data mesh principles converge to create more adaptive and intelligent enterprise ecosystems. 🎥 Watch this episode on YouTube: https://hubs.ly/Q02t6ZQd0 📻 Listen to this episode on your favorite podcast platforms: https://hubs.ly/Q02t71rb0 #podcast #datademocracy #datamesh #dataproducts #AI

12

2 Kommentarer

Research Council of Finland introducing simpler research plan structure for applications 📣 We will abandon subheadings, and the call-specific instructions will only be shown in the template. The new structure has three main headings: 1) Background and significance, 2) Impact and 3) Implementation. There is also the heading ‘List of references’ at the end of the plan. The basic research plan structure and instructions are easy to adapt to the objectives of each call. We will also move the short data management plan, previously included in the research plan, to the application form. As before, funded projects are required to submit a full data management plan before the accept the funding. Starting in autumn 2024, all Research Council of Finland research plan templates will be available at www.aka.fi/researchplan. The templates are available in English only, as applications submitted to the Research Council are reviewed in panels composed of international experts. The revised research plan structure also better reflects the needs of researchers applying for ERC funding. What, why and how 👇 In the section ‘Background and significance’, applicants are asked to reflect on the research-based starting points of their research and examine their own research against current knowledge in the field. Impact has been given a more prominent role in the research plan. In the ‘Impact’ section, applicants should describe the expected results, effects and impact of the research, both within the scientific community and beyond. The wider impact of the research can be described in relation to the 17 Sustainable Development Goals of the UN. In the third section, applicants shall describe in concrete terms how the research will be carried out: how, by whom and where. They also need to consider how to incorporate responsible science into their research and its implementation. Responsible science includes principles such as open access data and open access publishing, research ethics, equality and nondiscrimination. Learn more ➡ https://lnkd.in/dZj2E3JU #ResearchFunding

69

Artificial Intelligence is on everyone’s lips, and interpreters in particular are now often asked what AI means for their job and their future. We believe: Instead of being afraid, interpreters should use AI for their own good and can benefit massively from it. On our blog we explain how interpreters can make AI tools work for them. ✨ Read the blog post here: https://lnkd.in/dDwkYj_v #ai #artificialintelligence #ChatGPT #prompts #interpretation #interpreters #languageinterpretation #tips

11

🚀ASTOUND: Pioneering Excellence in Conscious AI 🌐In a groundbreaking study published in Nature, key team members from the #ASTOUND project consortium, explored the #theory of mind capabilities in humans and large language models (LLMs) such as GPT-4 and LLaMA2. Central to the ASTOUND project, this research delved into how well these AI models can track and interpret human mental states, which are crucial for social interactions and communication. 📈 Through a comprehensive series of tests measuring various aspects of theory of mind, we compared the performance of LLMs with that of 1,907 human participants. The findings reveal that while AI models can mimic human-like reasoning in several tasks, they also exhibit distinct limitations and biases, such as GPT models' hyper conservative approach to conclusions. This collaborative effort involved designing and conducting experiments, analyzing data, and interpreting results, thereby providing invaluable insights for future AI development. Ultimately, understanding how AI models process social information can guide the creation of more sophisticated and human-like AI systems and opens avenues for further research into mitigating biases and enhancing AI's social reasoning abilities. 📖For an in-depth analysis, read the full article on "Testing Theory of Mind in Large Language Models and Humans" here: https://lnkd.in/gzQmtk2F #ASTOUND #TheoryOfMind #LLM

15

2 Kommentarer

What is the economic impact of genAI in Denmark? Google asked, we answered. Read about it in the post below👇 or reach out to one of my talented economic colleagues Martin Hvidt Thelle, Bodil Emilie Hovmand, Laura Virtanen

12

Save BIG on your ticket to the premier Nordic event for Data Science & Machine Learning! The 9th annual NDSML Summit is packed with cutting-edge insights focusing on Applied and Value Driven AI Strategy, Generative AI, LLMs and SLMs, Machine Learning Infrastructure, Robust AI - Model Refinement and Deployment, Machine Learning Operations (MLOps), Advanced Tools and Frameworks, Data Engineering and DataOps, Computer Vision, Robotics and Edge Computing, AI Infrastructure and Platforms, Ethics and Governance in AI, featuring speakers from industry giants like Tesla, Amazon, Nike Spotify, Meta and many more! The NDSML Summit is the place to be for all professionals and organizations working with the utilization of Data Science, Machine and Artificial Intelligence. Don't delay, secure your spot at the premier event for Nordic Data Science and ML experts! https://hubs.li/Q02BRmxk0 #NDSMLSummit #DataScience #AppliedDataScience #MachineLearning #ML #ArtificialIntelligence #AI #EnterpriseAI #DeepLearning #DataEngineering #MLEngineering #DataPipelines #BigData #Automisation #GenerativeAI #EthicalAI #TrustworthyAI #LLMs #DataOps #MLOps

11

🔍 Exciting News Alert! 🚀 Cutting-edge research in Denmark delves into the transformative potential of Artificial Intelligence (AI) on knowledge production. Two groundbreaking studies spearheaded by experts in computer science and sociology of science at the University of Copenhagen are set to unveil the impact of generative AI on scientific practices. With funding from Villum Fonden and the European Research Council, these projects aim to map AI's influence on science, analyze its current and future impacts, and explore AI's role in shaping scientific knowledge. Led by Professor Roberta Sinatra, the research endeavors to compare human-generated knowledge with AI outputs, shedding light on the unexpected consequences of AI-infused science. Stay tuned for groundbreaking insights into the evolving landscape of scientific discovery! 🌟 #AIResearch #ScienceInnovation #KnowledgeProduction https://lnkd.in/dnmvS7EF Jan Petter Myklebust University World News

1

🌟 Promoting Diversity & Inclusion in Statistics and Data Science 🌟 We all know that diversity is key to driving innovation and expanding perspectives in any field. Yet early-career women in statistics and data science often face unique challenges that can limit their participation and growth. IISA excited to announce a special panel discussion where our diverse group of panelists—spanning academia, industry, and government—will share their personal stories of overcoming gender and cultural barriers in their careers. This session is a must-attend for: 🎓 Senior students 📚 Postdocs 👩‍💼 Early-career professionals interested in academia, industry, or government 👨‍🎓 Recent graduates exploring their next career steps. Join us for an insightful conversation where we’ll explore the challenges, solutions, and inspiring journeys of those who’ve made it despite the obstacles. 🌍 🔗 All are welcome to attend! Please register: https://lnkd.in/e5MiFudN Broti Garai; Shrabanti Chowdhury, Trinetri Ghosh, Arinjita Bhattacharyya

80

4 Kommentarer

Join our acclaimed Superforecasting workshop to improve calibration and boost your forecasting accuracy. Learn how to develop a structured process for forecasting, combat cognitive biases, and more. See our new Substack post on some of the Superforecaster strategies taught in the workshop: https://lnkd.in/ecj98R7n 📈 Top Reasons to Attend Good Judgment’s data science team analyzed training outcomes for our workshop graduates vis-à-vis untrained forecasters. They found immediate and consistent improvements in forecasting accuracy (our extended workshop graduates were 33% more accurate after training compared to untrained forecasters) and a marked reduction in collective bias after training. (Source: https://lnkd.in/eWdawPjb) Reserve your seat now: https://lnkd.in/eycA5uwT Learn more: https://lnkd.in/dj-_W8C

7

Since there is a lot of hype and one-sided media coverage of OpenaAI o1 out there (looking at you, Norsk rikskringkasting (NRK) and Aftenposten) I tried to come up with a more nuanced, informed and realistic perspective on the topic with a blog post and some examples also from other "thinking" models. I believe it is important to look beyond the excitement and we need to consider the complexities and limitations of all these shiny AIs we keep hearing about... https://lnkd.in/dY4reJmM Norwegian version: https://lnkd.in/drinnve5 #AI #MachineLearing

90

7 Kommentarer