signals’ Post

Panel II: focusing on #Competition_Dynamics in #Generative_AI, at the #GenAIVirtualWorldsWorkshop organised by DG Competition of the European Commission. https://lnkd.in/dAsreKBQ The discussion chaired by Carlota Reyners Fontana is with Dominique Costesec, Google, Blanche Savary de Beauregard, Mistral AI, Thomas Senderovitz, Novo Nordisk, Prof. Dr.-Ing. Stefan Wagenpfeil. This panel includes discussions on what the main sources of market power are in generative AI; and the competition dynamics currently unfolding in relation to generative AI. The role of partnerships and collaboration between large players and AI start-ups and discuss the best approaches to enabling a healthy competitive environment for AI innovation and deployment, delving into interoperability, open- and closed-source access, and other aspects. Key questions are what data is used for #training_models: public or private data? Who owns the data resulting from the training on #personal_data and consolidated with other data? Who can port them? Does the difficulty in identifying who owns the elaborated and consolidated data nullify the efforts to ease competition through personal data portability? How can AI aims to net zero, given its high-energy consumption? What does it mean open source AI? What is explainability in AI decision process? What are the implications of non explainability for transparency? Inspiring conclusions in the Keynote speech by Margrethe Vestager (Executive Vice President, European Commission) on the relevance to work together on AI ecosystems to keep markets open and contestable, collaborating on lesson learned with competition authorities across countries. AI is a new technology but what can drive it, as for example abusive power, are behaviours we know well about and we have the tools to deal with it. #data_portability #network_effects #new_digital_Divide #digital_transformations #ARU_CISE Anglia Ruskin University Hughes Hall, University of Cambridge International Telecommunication Union

What will it take to scale the next wave of AI innovation? In his latest piece for the World Economic Forum Annual Meeting, Jack Hidary emphasizes the critical role of collaboration across tech, academia, government, and finance to unlock the potential of Large Quantitative Models (LQMs). These models can drive real-world impact, revolutionizing industries like healthcare, finance, and environmental sustainability. But realizing this vision requires collective effort. Read Jack’s insights on how to build an ecosystem that ensures everyone benefits from AI’s transformative power: https://bit.ly/4fNFGvL

ICYMI - here's a replay: New breakthroughs in artificial intelligence could usher in a generational technology shift that transforms businesses, economies and the workforce. In a Washington Post Live program, Revolution CEO Steve Case, the European Commission’s top digital policymaker Roberto Viola and Center for Democracy and Technology CEO Alexandra Reeve Givens discussed how companies can navigate the new era of AI innovation and the balance between innovation and risk. In a segment presented by AWS, head of the AWS Generative AI Innovation Center Sri Elaprolu discussed how organizations can deliver meaningful value for their employees and customers with generative AI – with trust at the forefront.

Reflections from my talk at Berlin: “The Transition to Compound AI Applications” I recently had the privilege of speaking at the EdgeAI Summit in Berlin on a topic that is pivotal to the future of AI: the transition from single Model to Compound AI applications. This shift is essential for addressing complex business problems and generating substantial value through AI and Generative AI. The focus is not on choosing between open or closed AI models. Instead, it's about the system design and architecture, which revolves around leveraging high-quality data. This approach enables us to build AI applications that effectively invoke a chain of models to automate repetitive tasks. Moreover, the increasing importance of LLM Ops/ MLOps. These operations are critical for scaling AI applications and ensuring their smooth operation in production environments, providing reliable and robust solutions. It was a pleasure to meet with the team at our beautiful Berlin office. Together, we are driving innovation and leading the way in this exciting field. Moritz Winkel David Goebel Feras Shamasna #Data.ON #EON #Innovation #MLOPS #Data #GenAI #Welovetechnology.

Earlier this week, I attended the FARI - AI for the Common Good Institute conference in Brussels. As always, here’s a brief recap for those who couldn’t make it.   Cecile Huet opened the event with insights into the EU's vision for fostering AI-driven innovation. Beyond the significant investments in computing power and European data spaces, what stood out to me was the emphasis on innovation-friendly policies, particularly for startups. This focus on reducing barriers to creating groundbreaking AI solutions feels essential for driving the meaningful progress we need in the EU.   Martin Kotthaus, in his opening remarks, provided a thoughtful perspective on AI’s dual nature—its immense opportunities and inherent risks. In a field often dominated by polarized views, his balanced approach was refreshing. Striking the right balance between embracing opportunities and mitigating risks is a conversation we need to have more often.   Mireille Hildebrandt, FBA delivered a keynote that critically examined the concept of "AI for the common good." Rather than focusing solely on AI itself, she explored how we define notions of “common” and “good.” Her perspective opened the door to discussions beyond GDP growth—emphasizing the importance of using AI to strengthen governance and societal well-being. In other words: the language we use and the frameworks we adopt when discussing AI are crucial.   Paul Duan from Bayes Impact shared his experience implementing AI projects in collaboration with NGOs and the public sector. He talked about the concept of “death by a thousand pilots,” which resonated strongly with both the audience and other speakers. While pilots are essential, AI solutions eventully need to move beyond testing phases and into the real-world implementation.   Anna Colom, PhD from datatank.org tackled the narratives we build around AI. She urged us to move away from solely deterministic or techno-solutionist perspectives, which often shape how we approach and evaluate AI. A particularly striking quote from her talk was: “Lived experience is an essential type of expertise.” This deeply resonated with me, as it aligns with my work at Best Goed studio. Hearing it articulated so well was inspiring.   While there were many other engaging speakers at the conference, I hope this recap offers a glimpse into some of the key ideas and discussions.

I sat with Walt Maciborski for an interview that went live on CBS Austin's Tech This Out segment to share about the Austin AI Alliance! IBM's a founding member, building on our long history of economic impact in Austin. The mission of the group aligns to our strategy of building open and transparent AI and infrastructure. Some takeaways to reflect on... 1. People First: This has been rewarding because the issues we're working on are people problems - regardless of whether there's AI or other tech. 2. Right Time, Right Place: Austin has a VERY rare mixture of creatives, technologists, educators, creators, doers, makers, and just great people that are the "all y'all" that is the extended AI stakeholder community. 3. Rising to Our Potential: Tapping into Austin's people will show how much we can do -- on the world stage. 4. Don't let progress pass you by: Start dabbling, even if "let's see what happens if..." experiments. That will help you build an understanding of the tech. 5. Everyone's learning & You should too: The pace of change is incredible. "Agentic systems" didn't really exist ~6 months ago. "The RAG Pattern" didn't exist ~18 months ago. AI assistants as we know them didn't exist 24 months ago. Be curious and don't be intimidated.

Learn more about the mission and vision of the Austin AI Alliance in this edition of CBS Austin's "Tech This Out!" A rising tide lifts all! #AI #Innovation #Collaboration

Let this sink in…..2025 “You.com, the AI-powered search and productivity platform, announced today it has raised $50 million in Series B funding, marking a significant shift in the enterprise AI landscape. Georgian led the round, with participation from tech giants Nvidia and Salesforce Ventures, in addition to Day One Ventures, bringing You.com’s total funding to $99 million. This investment highlights the growing demand for AI solutions that can demonstrably enhance workplace productivity. Richard Socher, former chief scientist at Salesforce and a renowned figure in natural language processing, founded You.com in 2021. The company is pioneering a new category in the AI space: “productivity engines.” These tools aim to revolutionize how knowledge workers interact with information and complete complex tasks.” “You.com’s growth trajectory is impressive. The company reports serving over 1 billion queries since its launch and claims a staggering 500% revenue growth since January. This rapid expansion comes at a time when enterprises are grappling with “AI sprawl” — the proliferation of disparate AI tools and subscriptions across organizations. To address this challenge, You.com offers access to multiple leading AI models through a single platform, enhanced with live web access. The company also places a strong emphasis on accuracy, a critical factor for enterprise adoption.” #ai #genai

As Aurora explores AI technology, it's crucial that every resident and city employee understands the building blocks of these systems. In my latest piece, I break down our community's core AI principles and show how they could improve city services for everyone - from streamlining infrastructure maintenance to ensuring equitable service delivery. Read about: * The Three Rs of AI in practical terms * How our CLEAT model protects community interests * Real possibilities for improving city services * Ways to ensure AI benefits reach all neighborhoods This isn't about complex technology - it's about building a better Aurora together.

Taking part, remotely, to the really exciting Workshop of 28 June 2024 on virtual worlds and generative AI, #GenAIVirtualWorldsWorkshop organised by DG Competition of the European Commission. https://lnkd.in/dAsreKBQ Already in Panel I, on #Competition_Dynamics in #Virtual_Worlds, the discussion with Fabiana Di Porto Mariana de la Roche Wills Sebastien Massart chaired by Cani Fernandez covers how #competition_policy can better address virtual worlds data ecosystems, and how it needs to revise outdated notions of market definition, bottlenecks and dominance. Also, critical is the role of interoperability and standards for virtual ecosystems. This brings new meaning to how to address #data_portability and how it changes once data are elaborated. What are the new barriers to entry? #data? #clowds? #skills? #affordability? #new_digital_Divide #digital_transformations #ARU_CISE