Content digest #25

Hyperight Read

Source: Midjourney

Evaluation Techniques for Large Language Models – Interview with Rajiv Shah, Hugging Face: In an interview, we sat down with Rajiv Shah, a Machine Learning Engineer with ten years of experience in data science. Rajiv provides a glimpse into Hugging Face’s role in revolutionizing the open-source AI community. As a key figure at Hugging Face, Rajiv focuses on addressing complex challenges using open-source AI. Rajiv will also present on “Evaluation Techniques for Large Language Models” at the upcoming Data Innovation Summit 2024. Additionally, he discusses practical applications of AI in enterprise teams and emphasizes the role of interpretability. The interview concludes with insights into upcoming AI trends and an exploration of Hugging Face’s impact.

The Data Speaker’s Blueprint: Turning Analytics into Applause: During their career as a Data Scientist, Alessandro Romano attended many conferences and meetups. When they started their job, many struggled to understand what data science was and how to leverage cloud computing solutions. That’s why connecting with the data community through meetups and other local events is essential. ‘’At some point, someone invited me to present a project I was working on. It was then I realized how much I enjoy sharing and explaining my work! Since then, I’ve had the pleasure of speaking at various conferences, including Data Innovation Summit, among others.’’ How do you find the right story for a conference? Aren’t you scared of making mistakes while presenting something technical? These questions made Alessandro realize it’s time to share what they’ve learned over the years and, hopefully, inspire new data experts to share their knowledge.

Digital Twins and Synthetic Voices Can Tell Stories 24/7 – Interview with Tove Mylläri, Yle – Finnish Broadcasting Company: In this interview, we had the chance to speak with Tove Mylläri, AI Innovation Lead at Yle, Finland’s national public broadcasting company. Tove is a long-standing technology, and innovation professional, a keynote speaker and the host of Yle’s AI Demo Event. She speaks about AI literacy and the upbringing of AI-native children, a topic she’s passionate about. She elaborates on the importance of AI literacy in society, especially for children, and how it impacts their upbringing and future skills development. Furthermore, Tove will be presenting more insights on “Digital twins and synthetic voices can tell stories 24/7” at the 9th annual Data Innovation Summit in Stockholm this April.

7 Practical Applications of RAG Models and their Impact on Society: As Retrieval-Augmented Generation (RAG) models keep on advancing, their impact spreads across the NLP community. These models are fascinating, sitting at the intersection of natural language processing (NLP) and information retrieval. RAG refines information synthesis and not only leverages context and relevance, but also promotes richer and contextually aware outputs. RAG serves as an AI framework, providing relevant data as context for generative AI (GenAI) models. But how does it achieve this? This article covers the core of the RAG approach.

Show me the money! – The ROI of marketing analytics – Interview with Gabor Harsanyi, Ericsson AB: In this interview, we had the privilege to talk with Gabor Harsanyi, Head of Connected Data and Analytics at Ericsson AB. With over two decades of expertise in strategy, business intelligence, marketing analytics, customer insights, and data science, his domain knowledge spans across ICT and financial verticals. In this interview, Gabor talks about the transition from data-driven to data-informed approach, the challenges in demonstrating marketing value, and the Importance of data quality and democratization. As a featured speaker on the Business and Data Analytics Stage, Gabor will delve into “Show me the money! – The ROI of marketing analytics” at the upcoming Data Innovation Summit 2024.

News, Articles, and Reports

Welcome to another edition of the Monthly Hyperight Newsletter, where we share knowledge, inspiration, insights, and much more! OpenAI focuses on AI agents to automate tasks; DeepMind's AlphaGeometry excels in geometry problem-solving; Google's Exphormer advances graph data processing; Dify.AI offers tools to create intelligent agents with diverse language models.
Source: Midjourney

OpenAI Announces Development of AI Agents: OpenAI is transitioning its focus towards the development of AI agents: software programs designed to automate complex tasks by controlling a user’s device. These agents are intended to perform actions such as transferring data, filling out reports, and entering information into software, mimicking human interactions with various applications. OpenAI is working on two types of agents: one that handles device-based tasks and another for web-based tasks.

More Agents Is All You Need: Simply via a sampling-and-voting method, the performance of large language models (LLMs) scales with the number of agents instantiated. Also, this method is orthogonal to existing complicated methods to further enhance LLMs, while the degree of enhancement is correlated to the task difficulty. Industry experts have conducted experiments on a wide range of LLM benchmarks to verify the presence of our finding, and to study the properties that can facilitate its occurrence. Read more here.

Dify.AI Unveils AI Agent: Creating GPTs and Assistants with Various LLMs: Humans solve complex problems by understanding information, making choices, and using tools. LLMs excel at processing language, offering potential to aid in real-world problem-solving. Previously, OpenAI introduced GPTs and Assistants API, sparking creativity. Dify now expands possibilities by allowing access to various LLMs and tools to create intelligent AI Agents.

Google Launches Gemini Ultra: Google Bard is no more. Almost exactly a year after first introducing its efforts to challenge OpenAI’s ChatGPT, the company is retiring the name and rebranding Bard as Gemini. More importantly, it is also now launching Gemini Ultra, its most capable LLM yet. However, Gemini Ultra will be a paid experience. Google is making it available through a new $20 Google One tier that includes 2TB of storage and the rest of Google One’s feature set, as well as access to Gemini in Google Workspace apps. With that, Google will also sunset the Duet AI brand, which was mostly restricted to its AI features in Workspace, and move that to Gemini as well.

Google’s Exphormer for Graph-Structured Data: Exphormer, Google’s latest stride in machine learning, addresses the scalability challenges of graph transformers, essential for interpreting complex, graph-structured data in various sectors. By combining expander graphs with sparse attention frameworks, Exphormer efficiently processes large-scale graph data, overcoming traditional computational limitations. It demonstrates exceptional performance across different datasets, marking a significant advancement in the application of graph neural networks.

DeepMind’s AlphaGeometry AI System for Geometry: AlphaGeometry, an AI system from DeepMind, demonstrates capability in solving Olympiad-level geometry problems. It integrates a neural language model with a symbolic deduction engine, enabling it to perform complex mathematical reasoning. Achieving near-human performance, it successfully tackled 25 out of 30 International Mathematical Olympiad problems, signaling a significant leap in AI’s mathematical reasoning abilities and marking a step towards advanced, general AI systems.

AI After Work Podcast

AIAW Podcast
Photo by Hyperight AB® / All rights reserved.

AIAW Podcast E115 – Talent in AI and AI in Recruitment – Alexandra Davis and Stefan Wendin: Get ready for an insightful journey into AI and HR with our remarkable guests in this episode of the Weekly AIAW Podcast! This time, we dive into the most crucial aspect of AI – people. Together with us, we had the pleasure of welcoming Stefan Wendin, a Strategic Advisor from You Don’t Know What You Know (YDKWYK), and Alexandra M. Davis, Country Manager & Senior Recruiter at Sperton. Stefan’s visionary insights into technology integration and Alexandra’s wealth of HR expertise provide valuable perspectives on the transformative power of AI in HR.

AIAW Podcast E116 – RAG models to autonomous AI agents – Jesper Fredriksson: Get ready for this episode of the AIAW Podcast featuring Jesper Fredriksson, AI Engineer at Volvo Cars! This episode delves into the fascinating world of Retrieval-Augmented Generation (RAG) models and autonomous AI agents. Join us as Jesper unravels the complexities and potential of these advanced AI systems in transforming industries. A must-listen for anyone interested in AI technology!

AIAW Podcast E117 – AI-driven software testing – Vilhelm von Ehrenheim: Join us on this episode of the AIAW Podcast for an enlightening conversation with Vilhelm von Ehrenheim, Co-Founder and Chief AI Officer of QA.tech! Vilhelm brings over a decade of experience in applying data science and machine learning to develop autonomous agents that test and interact with web pages. Before co-founding QA.tech, he was the driving force behind Motherbrain Intelligence at EQT, leading data science initiatives for one of the world’s leading private equity firms. In this episode, we explore Vilhelm’s journey, delving into his expertise in data engineering, natural language processing, large language models, and more.

Hyperight Attend

Data Innovation Summit
Photo by Hyperight AB® / All rights reserved.

Data Innovation Summit 2024: Kistamässan, Stockholm, 24 – 25th of April 2024

The 2024 edition of the Data Innovation Summit brings together the most innovative minds, enterprise practitioners, technology providers, start-up innovators, and academics, working with Applied Data Innovation, Data Science, Big Data, ML, AI, GenAI, LLMs, Data Management, Data Engineering, Architecture, Databases, responsible AI, Industrial Analytics, in one place to discuss ways to accelerate AI-driven Transformation throughout companies, industries, and public organizations.

The ninth annual edition will cover a range of topics spread across 9 stages aligned over a common theme: ‘’From data to value in the age of applied and generative AI’’. By registering, practitioners will engage with over 300 speakers, benefit from 200+ TIP sessions, and connect with professionals across the globe, accelerating AI-driven transformation in their organizations.

  • Join the largest annual data and AI gathering with over 3,000 attendees for an unforgettable experience!
  • Explore 250 presentations, 7 workshop rooms, and more than 200 demos for a deeper understanding of the data-to-insight-to-action process.
  • Witness the latest technology from more than 100 exhibitors, some making their Nordic debut, offering a glimpse into the future of data and AI.
  • Attend in-person or online and become part of a global movement connecting the data, analytics, and AI community.
  • Immerse yourself in vibrant activities, parties, and challenges for a balanced and enjoyable event.
  • Unlock the future of data innovation, and reserve your seat NOW at the event of the year!

Don’t miss out, secure your spot at datainnovationsummit.com!

tpo33 events provide a unique solution, surpassing traditional learning formats to offer targeted, solution-oriented peer-to-peer networking. Designed for decision-makers and practitioners, these events optimize time investment, facilitating collaboration among a select community of like-minded peers for effective problem-solving.

CAIO (Chief AI Officer) tpo33 is tailor-made for every practitioner or organization working with or interested in learning how to maximize the ROI of their AI initiatives and function. The event is created for Chief AI Officers, CxO’s, or senior managers who are leading the AI initiative and strategy in their organization.

FinOps – tpo33 event: ‎‎‎19th of March 2024, Stockholm‎‎

FinOps tpo33 is an ideal event for every FinOps and cloud efficiency professional or organization seeking to navigate the complexities of cloud spending. By delving into analyzing, optimizing, and automating financial operations, participants will discover effective strategies to not only reduce costs but also to unlock new growth opportunities, drive innovation, and achieve a competitive edge in the rapidly evolving digital landscape.

CBIO – tpo33 event: 21st of March 2024, Oslo

CBIO (Chief BI Officer) tpo33 is tailor made for every practitioner or organization working with or interested in learning how to maximize the ROI of their BI initiatives and function. The event is created for Chief BI and Analytics Officers, CxO´s, or senior managers who are leading the BI and Analytics initiative and strategy in their organization.

Until next time,

Hyperight Team

Add comment