Marcus Reeve’s Post

For this solution, Euromonitor International offers a combination of technologies and approaches that include large language models, retrieval-augmented generation, and multi-agent systems. “Unlike other generative AI tools, the information provided by Passport AI is 100% sourced from Euromonitor's trusted content, shared Lamine Lahouasnia, director of generative AI at Euromonitor International. Tim Kitchin, CEO at Euromonitor International, also remarked, "By providing clients with access to Passport AI, we will be equipping them to take data-driven decisions, make the most of opportunities, and achieve greater success, demonstrating our commitment to continuous product improvement.” #productlaunch #artificialintelligence #marketintelligence ------ -[Newsletter] To learn more about the latest news, trends and insights in the technology industry, subscribe to UpTech Media’s newsletter by filling out this form: https://lnkd.in/eKr8Se4y

The AI Forum NZ's Working Group on Generative AI is seeking input from businesses in the Transport sector. They are compiling a living white paper to help organisations and the Government understand how LLMs are being used in New Zealand. To provide an accurate picture, they have created a short questionnaire covering current and planned use cases, potential benefits, constraints, and concerns. If you or someone in your business can contribute, please use this link to complete the questionnaire online. Your insights will be invaluable. https://lnkd.in/g3ZYf7fB #AI #GenerativeAI #Transport #NewZealand #LLMs

Heads up Aotearoa New Zealand #transport sector. Are you considering or using #GenAI / #LargeLanguageModels? You're invited to #contribute ⬇️ to this AI Forum NZ white paper

Thank you, Jacek Korneluk, for sharing my guest post on Spektrumlab's blog! In this post, I explore how structured conversations with large language models can create persistent, intelligent AI agents. Whats next? 1. Workshop that teaches the soft skills required for agent development via dialog with large language models. 2. Open source code base that can be used in the workshop for the exploration via a few assignments that drive you in the right mindset. 3. Use the MIT Licensed code base for your own projects. We are using this approach for solving real business cases. Interested? Read the blog post on Spektrumlab

What does the public in six countries think of generative AI in news? This report looks at how people use generative AI, and what they think about its application in journalism and other areas of work and life across six countries. https://lnkd.in/eSjcctWu

So may we say that as more AI generated language interpretation comes in the same thing may be happening? AI drifting more and more from actual language and inventing a homunculus of its own? This reminded me of the Facebook experiment several years ago when 2 bots created their own language and eventually had to be shut down because the experimenters stopped understanding what they were saying to each other. But it also may mean that in order to stay afloat AI language interpretation models may have to continue to prey on human interpretation output (with interpreters’ knowledge and consent?). And maybe soon it will be difficult even to differentiate between AI and non AI generated data so the models are bound to degrade ?

During its World News Media Congress in Copenhagen, WAN-IFRA, the World Association of News Publishers published its 5th report focused on #AI. The report, “AI in Action,” features 10 practical, global use cases of how news #publishers are working with #ArtificialIntelligence, particularly #generativeAI.

The dream: Your #generativeAI system continues learning, becoming more effective and efficient 😄 The reality: Your #GenAI drifts from time to time, and it’s not clear why 😕 Research shows that large language models (#LLMs) drift sooner than later, sometimes just months. Without an evaluation framework in place, even a minor change in LLM behavior can majorly impact your generative #AI system, from higher operational costs 💰 to hallucinations that violate policy 😵💫 Here at WillowTree, we’ve developed a four-part framework we trust to continually evaluate the performance of our clients’ LLM-based applications: 1️⃣ Choose your LLM evaluation metrics 2️⃣ Create a gold standard dataset 3️⃣ Generate new responses 4️⃣ Compare new responses to the gold standard dataset With this evaluation framework, you’ll better balance the human oversight of your team with the efficiency of your generative AI system. Milton Leal and Christopher Frenchi of our Data & AI Research Team (DART) walk you through it. Get the framework 👉 https://lnkd.in/eE8VTgAx

WAN-IFRA has released its fifth report at the World News Media Congress in Copenhagen, spotlighting AI in the news. The report, titled "AI in Action," showcases 10 practical global use cases of how news publishers are leveraging #ArtificialIntelligence, specifically #generativeAI. Download and explore the report here: https://lnkd.in/e_2Hvsf7 #WNMC24

Generative AI and large language models are going to make their mark on the federal government in 2024. Here are 3 federal generative AI trends for this year. https://lnkd.in/gyjY8j4e