FARI - AI for the Common Good Institute on LinkedIn: FARI - FARI is an independent, not-for-profit Artificial Intelligence… (2024)

"Progress without wisdom is not real progress". Wisdom versus intelligence in the deployment of AI? A central theme of last year's FARI Conference. How should the Brussels-Capital Region work towards a functional and inclusive AI policy? The time is now.Some recommendations:1. Foster citizen engagement for an inclusive and beneficial AI ecosystem2. Establish a unified framework for developing truly inclusive AI technologies3. Implement Robust Safeguards and Guidelines for the Responsible use of AI Technologies in Critical Applications4. Maintain criticality and transparency in AI development and deploymentRead the report for the full summary.

2

Research and Innovation initiatives,TITAN,AI4media,AI4Trust, andvera.ai- in cooperation with the European Commission - are pleased to invite you to a policy and innovation conference in Brussels on 29 June 2023, focused on exploring the issues and challenges around generative AI and the tackling of disinformation.

7

I am excited to moderate the panel discussion at today's 5th Annual AI Conference in Sandton. The panel topic is "Realizing Generative AI’s full competitive potential" https://lnkd.in/dWsdq8Nb Machine Learning Africa

9

"Eventually, [machines] could be our natural evolution. But to play safe let’s start mimicking first human intelligence and giving room to less biological and unsupervised processes within controlled frameworks. By doing so we would crucially ease the convergence between the regulators and the industry needs."Read more about it in our new paper:Advances in AI:When Applied Science is not Science Appliedhttps://lnkd.in/dtMHQWjF

5

“Putting our knowledge on the table, together with the engineers so that we can discuss cause-and-effect relationships in an intuitive language, [we saw] so much success”Karim Said B., Research Scientist from Bosch, speaking at our Causal AI Conference last year, discussing the uphill battle for true impact in a diverse engineering enterprise.Karim identified the genuine, tangible impact of causal AI, particularly in diverse engineering enterprises where decades of specialized know-how coexist with widespread applications of AI research.Watch the talk here: https://lnkd.in/erS3RhJE

21

Overview page of this submission has some of the most concrete and promising ideas for the governance of advanced AI models I've seen anywhere. Check it out.

4

JPA Health’s EVP of Data Science, Colin Baughman, will be attending ISMPP (International Society for Medical Publication Professionals)’s 2024 European Meeting tomorrow in London. Presenting on the panel “Making an Impact: What Does Success Look Like in Publications?,” he will present “Predicting Publication Impact with AI,” a new methodology leveraging generative AI to predict publication impact prior to release using a unique combination of topic relevance & novelty, stakeholder knowledge assessment, author productivity, digital affinity, and journal impact. Learn more: https://lnkd.in/eT-hgGHC#ISMPPEurope2024 #ArtificalIntelligence

16

1 Comment

What a blast we had at the 'Meet the Future of AI' conference last week. Thank you to all who spoke and/or attended. Lots to digest but here's a quick snapshot of the day via our blog 👇 #disinformation #AI #generativeAI #misinformation #publictrust #fakenews #EC #researchandinnovation

9

The Computing Community Consortium, in collaboration with INFORMS and ACM SIGAI, will be hosting the final workshop of a three-part series titled,Artificial Intelligence (AI)/Operations Research (OR), in late February or early March 2024, to set a course for fundamental research that needs the partnership of both disciplines. Organized by Yu Ding (Georgia Institute of Technology), Pascal Van Hentenryck (Georgia Institute of Technology), Sven Koenig (University of Southern California), Ramayya Krishnan (Carnegie Mellon University), Radhika Kulkarni (INFORMS), and Phebe Vayanos (University of Southern California), the workshop will focus on drafting a strategic plan for increasing the AI/OR partnership and on outlining real-world opportunities for collaboration, based on discussions from the previous two workshops.For this workshop, organizers are requesting proposals from the AI and OR communities for compelling and inspiring grand challenge problems of a theoretical or applied nature that: require the collaboration of AI and OR researchers and will result in basic research on the integration of AI and OR techniques that align with societal needs and national priorities with potential for real-world impact. The organizers would like proposals to describe a challenge problem that will result in the integration of AI and OR methods and articulate the AI and OR methods needed, dataset availability (if applicable), method innovation, evaluation criteria, path to applications, and the potential to result in impact on societal needs and national priorities. Some questions to consider are:A.What are challenges that you believe will be faced to tackle this problem, and what are the mechanisms needed to address them?B.How can policies/mechanisms incentivize collaborations between the two communities?C.Why do you think the two communities need to collaborate to address this challenge?Challenge problems can be submitted using thisform. https://lnkd.in/gpjbNz_u Submissions can be written in paragraph or bullet point style formats. Your submission should convincingly explain why it satisfies points 1, 2, and 3 above and how it addresses questions A, B, and C. Some high level ideas of challenges include leveraging optimization to mitigate bias in large language models, or using robust optimization to enhance privacy of generative AI. The deadline for submission isSeptember 15, 2023.Selected submitters will be invited to participate in the third workshop to help shape the discussion around collaboration between AI and OR communities and researchers and to help break barriers in such collaborations.

20

Diffusion models are going federated! Phoenix is a model from University of Washington Bothell built by the lab of Afra Mashhadi that allows diffusion models (similar to those by Stability AI, and other Gen AI companies) to be trained on federated data. Phoenix is implemented on Flower! 🌼 This is an important step in the evolution of Generative AI that we are excited to see taken. 🚀 Paper available here: https://buff.ly/44BRBHL

70

1 Comment