Content digest #27

Hyperight Read

From Transformers to ChatGPT: Navigating through key technical innovations in the past few years best describes the journey from Transformers to ChatGPT. ChatGPT and similar chatbot experiences are quintessential examples of how the sum of individual innovations can be greater than its parts. Each innovation enhances the machine’s ability to understand and process language, finally bringing us to a seamless product like ChatGPT. In this article, we’ll cover the following things – The evolutionary narrative of ChatGPT. How one key technical innovation in NLP laid the groundwork for the next.  The magic ingredient behind ChatGPT. What each technical innovation offers to language understanding and processing.

3 Power-Ups: RAG’s Impact on Large Language Models: In natural language processing, the constant pursuit of more efficient and precise text generation methods persists. Large language models, or LLMs, have dazzled us with their prowess in text generation and language translation. However, a pressing question arises: can we trust them to always provide accurate and informative responses? Meet retrieval-augmented generation (RAG), a game-changing technique that tackles this challenge head-on by seamlessly integrating real-world knowledge with LLM capabilities! In this article, we cover the impact of RAG on large language models. We examine how it enhances comprehension, improves response coherence, and ultimately fosters more reliable applications.

4 Pillars to Effective Training of Large Language Models: Large language models – advanced artificial intelligence systems trained on vast amounts of data to understand and generate human-like text. How exactly do these models carry out such functions? In this modern age of sophisticated artificial intelligence, the training of large language models (LLMs) is instrumental. It shapes the course of future communication, automation, and information processing. Today, the need to ensure not only the effectiveness, but also reliability and fairness of LLMs, is non-negotiable. Their performance hinges on precise pre-training, absorbing vast data to understand human language. For this reason, rigorous examination and improvement are essential to mitigate biases.

Measure Outcome, Not Output: Navigating Data and Analytics Metrics in Today’s Business Landscape: In the fast-paced world of data and analytics, where every decision counts and every insight shapes the trajectory of a company, the way we measure success becomes paramount. Recently, I had the opportunity to convene with 18 Chief Data Officers and other senior leaders in data and analytics from various sectors, such as IT, finance, retail, manufacturing, services, and public sector across Europe. Through our discussions, several key insights emerged, shedding light on the intricacies of measurement frameworks, Key Performance Indicators (KPIs), and Objectives and Key Results (OKRs).

tpo33 Events 2024 Reflection: A New Immersive Knowledge-Sharing Experience! In a world saturated with information where technological trends often dictate our paths, a new way of sharing knowledge arises: tpo33 events! At tpo33 events, learning transcends the ordinary, and solutions are born from collaborative minds. These gatherings are designed for practitioners seeking more than just surface-level insights. As four tpo33 events, limited to 33 participants, unfolded, they offered profound discussions, problem-solving endeavors, and invaluable networking opportunities!

Beyond the Basics: 5 Fine-Tuning Stages for Precision in Machine Learning: When it comes to machine learning, mastering the fundamentals is just the beginning. Delving deeper, beyond the basics, lies fine-tuning – a process for achieving precision and enhancing model performance. Pre-trained models, such as ChatGPT, have transformed natural language processing (NLP) and intelligent chatbots. Fine-tuning adjusts these models to different contexts by refining parameters with task-specific data. In this article, we cover the stages of the fine-tuning process for machine learning models, with a focus on precision enhancement in various applications. We delve into methods for customizing pre-trained models like ChatGPT for specific tasks, ensuring peak performance.

Video Interviews

Robocops and Superhumans: Ethical Dilemmas of Frontier Technology – Interview with Uthman Ali, Senior Product Analyst at BP: In this interview, we speak with Uthman Ali, AI Ethics Expert and Senior Product Analyst at BP. Uthman founded the company’s first Digital Ethics Center of Excellence and advised AI startups on regulatory navigation. Uthman is a speaker on AI ethics, sits on the advisory board of Oxethica, and has published influential papers with organizations like the IEEE, World Economic Forum, and MIT. This interview explores AI, robotics, and human augmentation advancements, addressing the balance between hype and realistic expectations for the next 5 to 10 years. We discuss ethical risks, societal preparation, equitable access, and the role of human workers. Moreover, Uthman will be presenting as a keynote at the upcoming Data Innovation Summit 2024. Get ready for his presentation on “Robocops and Superhumans: Navigating the Ethical Maze of Frontier Technology!”

Empowering Customer Conversations with SagaAI’s Intelligent Insights and Recommendations – Interview with Anca Iordanescu, VP Engineering, Store of the Future at Ingka Digital, IKEA: In this interview, we speak with Anca Iordanescu, VP Engineering, Store of the future at Ingka Digital, IKEA. Anca is an engineering leader known for her global team-building prowess in consumer and enterprise tech sectors. She advocates for women and girls in technology, speaking at forums on leadership and digital transformation. Recognized as a top Woman in Tech and a Digital Transformation Leader in 2022, she leads Sweden’s “Women in Tech” community. This interview delves into AI’s impact on retail and personalization’s link with AI in enhancing shopping experiences. It addresses generative AI, ethical concerns, and advice for AI-driven solutions. Moreover, Anca will be presenting at the Data Innovation Summit 2024 happening in less than 2 weeks! Her presentation is titled “Unlocking Ikea SagaAI- Empowering customer conversations with SagaAI’s Intelligent Insights and Recommendations.”

Semantic Layers: Your Strategic Advantage for AI-driven Insights – Interview with Ernesto Ongaro, dbt Labs: In this interview, we speak with Ernesto Ongaro, Team Lead and Solutions Architecture at dbt Labs. In this interview, we talk about semantic layers and why they offer a strategic advantage in the age of LLMs. Get ready to learn from cutting-edge research and understand how the semantic layer can accelerate your organization’s data-driven decision-making! Moreover, Drew Banin, Co-Founder at dbt Labs, will be presenting on “Setting the Right Foundation for Innovation” at the Data Innovation Summit in less than 2 weeks! We are also proud to say that dbt Labs are one of our partners at this edition of the summit.

News, Articles, and Reports

AI Copilot: Launching Innovation Rockets, But Beware of the Darkness Ahead: Imagine a world where the software that powers your favorite apps, secures your online transactions, and keeps your digital life could be outsmarted and taken over by a cleverly disguised piece of code. This isn’t a plot from the latest cyber-thriller; it’s actually been a reality for years now. How this will change – in a positive or negative direction – as artificial intelligence (AI) takes on a larger role in software development is one of the big uncertainties related to this brave new world.

Adobe’s ‘Ethical’ Firefly AI Was Trained on Midjourney Images: When Adobe Inc. released its Firefly image-generating software last year, the company said the artificial intelligence model was trained mainly on Adobe Stock, its database of hundreds of millions of licensed images. Firefly, Adobe said, was a “commercially safe” alternative to competitors like Midjourney, which learned by scraping pictures from across the internet.

Microsoft to invest $1.5bn in Abu Dhabi AI group G42: Microsoft has agreed to invest $1.5bn in Abu Dhabi artificial intelligence group G42, its latest big bet on the technology that underscores deepening collaboration between the US and United Arab Emirates. The agreement gives Microsoft a minority stake in G42, and its vice-chair and president Brad Smith will have a seat on its board. It comes after G42 severed its links to Chinese hardware suppliers, which had been the subject of scrutiny by US lawmakers.

Workers Who Regularly Use Gen AI Would Like More Support: Although U.S. employees who regularly use generative artificial intelligence tools in their daily work are confident about the technology, they also would like additional support from their leaders. “We will continue to find new and innovative ways to shorten our labor gaps with intelligent customized AI interactions and processes, but we will need access to skills in our teams who can build them,” said Steve Watt, chief information officer at Hyland.

Chair Rodgers in Fox News: “It is past time for us to put people in control of their data” Energy and Commerce Chairwoman Cathy McMorris Rodgers, R-Wash., said it is her ‘biggest fear’ as a parent for her children’s sensitive data to be exploited online, and she is aiming to quell those fears for herself and millions of Americans with a new federal data privacy standard. “‘There’s very dangerous activity online, and data collection is foundationally how these algorithms are developed and then ultimately used to target children, but also people in general,’ Rodgers told Fox News Digital in an interview. “‘This is really my biggest fear as a parent, having three school-aged children, other friends who have had kids that have been targeted online. We know that these algorithms have been targeting children with dangerous content, targeted advertising that leads to dangerous life-threatening behaviors.’

H&R Block answers tax questions using gen AI: H&R Block is leveraging generative AI to help customers as they prepare their taxes using the company’s products before today’s deadline. Tax preparation company H&R Block is no stranger to AI and machine learning (ML), having leveraged the technologies across its business for years. But now it’s diving headfirst into gen AI as it sees the potential to transform nearly every aspect of its business, from customer-facing applications to internal functions like engineering, marketing, and legal.

AI After Work Podcast

AIAW Podcast E122 – AI Powered Product Development – Fredrik Stockman: Tune into Episode 122 of the AIAW Podcast! In this episode, we cover the exciting world of AI-powered product development with Fredrik Stockman, CEO & Co-Founder of Version Lens. Explore Version Lens, transforming Product Managers’ roles with the world’s premier AI co-pilot for product development. Learn from Fredrik’s wealth of experience as a serial entrepreneur, and explore how Version Lens is transforming the industry by automating mundane tasks, unlocking new insights from existing data, and enabling product teams to concentrate on innovation rather than overhead. Don’t miss this insightful episode where technology meets creativity in the realm of product management!

AIAW Podcast E123 – The Power Grid of Tomorrow – Kiryl Zhdanovich & Jonatan Raber: Tune in to yet another AIAW Podcast! Episode 123 features a discussion with Kiryl Zhdanovich, seasoned Machine Learning and Data Engineer at Fever, and Jonatan Raber, Founder and Chief Product Officer at Fever. With impressive backgrounds with industry giants like Ikea and Spotify, they dive into ‘The Power Grid of Tomorrow.’ This episode explores how the future of power technology is shifting its focus from traditional hardware to ground-breaking advances in software engineering, computing, and modeling. Discover Fever’s system. It makes autonomous millisecond decisions, optimizing and balancing power production and consumption for a future fueled by clean, abundant, and resilient energy. Don’t miss this insightful conversation that illuminates the next generation of energy solutions! AI

AIAW Podcast E124 – All about #DBRX AI Model – Hagay Lupesko: Tune in to Episode 124 of the AIAW Podcast! In this episode, we cover artificial intelligence with Hagay Lupesko, one of the contributors behind the #DBRX AI model. With a rich history in software and ML engineering at giants like GE, Meta, and Amazon, Hagay now shines as VP Engineering at MozaicML, a recent addition to the Databricks family. This episode covers a spectrum of fascinating topics: Hagay’s journey before MozaicML and Databricks, the story behind MosaicML’s acquisition, and the features of DBRX. We explore debates between open-source and closed-source in AI, GenAI’s enterprise applications, and the technical intricacies behind DBRX’s performance. Get Hagay’s insights on the future of LLMs and his vision of the AGI era!

AIAW Podcast E125 – Liza-Maria Norlin: In the latest AIAW Podcast episode 125, we host Liza-Maria Norlin, Party Secretary of the Christian Democrats in Sweden and author of “The Courage to Lead through Values!” This episode covers a discussion with Liza-Maria, starting with her recent election as a Member of Parliament and key issues she aims to address. We explore her background and discuss how AI aligns with Christian Democratic values. It also dissects the pros and cons of the EU AI Act. Liza-Maria shares her perspective on what it means to be a Party Secretary, and the emerging landscape of GovTech Sweden. A highlight of our conversation is how AI supports value-based leadership, drawing insights from her book. We also tackle pressing topics like AI’s role in enhancing public value and ethical considerations in AI development. Also Sweden’s position in global AI governance, and speculate on the potential futures of AGI.

Hyperight Attend

Data Innovation Summit 2024: What You Can’t Afford to Miss!

Are you ready to dive into the future of data, analytics, and artificial intelligence? Buckle up as in less than 10 days, the Data Innovation Summit 2024 kicks off!

Set against the backdrop of the Nordic landscape, this hybrid event unites the global community of data enthusiasts, innovators, and industry leaders. The Data Innovation Summit is the largest and most influential annual gathering in the Nordics. As such, it convenes the brightest minds in applied data innovation, machine learning, AI, and beyond. This summit boasts nine stages, 250+ sessions, and more than 3000+ delegates from leading companies. It engages in world-class case studies, connects with industry experts, and explores innovative workshops.

The initial 3000 tickets planned for this edition are now SOLD OUT! In response to overwhelming demand and with our valued attendees in mind, we’re thrilled to announce the release of 300 extra tickets this year! These are the last remaining extra tickets for this edition. If you’re thinking about attending, now is your final opportunity to secure your spot!