Data Science Bulletin

In the nineteenth episode of the Data Science Bulletin podcast, hosts discuss two articles focusing on the influence of large language models and AI on the future of various fields."Something Big is Happening" This article analyzes the potential impact of AI on our lives, highlighting both the current enthusiasm and the possible overestimation of the capabilities of today's language models."Global Intelligence Boom 2028" This piece offers an optimistic look at the economic effects of AI, speculating on deflation and the creation of new job opportunities—though the authors critique some of its assumptions as being unrealistic.Short writeup on DSB website

In this episode we discuss the research paper 'Large Language Model Reasoning Failures,' which was published in Transactions on Machine Learning Research. We focus on a categorization framework for identifying situations where large language models fail at reasoning, pointing out that the goal of the paper is not to evaluate whether these models actually think, but rather to understand their limitations compared to human reasoning. The episode also highlights that these errors can be similar to cognitive biases in humans and discusses possible approaches to overcoming them.Blog post with detailed description.

In the 17th episode of the Data Science Bulletin podcast, hosts delve into a recent paper on "Decepticon," which examines the vulnerability of AI agents to dark patterns on websites, revealing that larger models are surprisingly more susceptible. Additionally, they dissect the recent hype around "Moldbook," a social network for AI agents, discussing its security flaws and the societal and technological implications of autonomous agents engaged in online interactions.Blog post with sources

In the 16th episode of the Data Science Bulletin podcast, hosts Martin and Jakub delve into the intricacies of AI strategies, focusing primarily on the national AI strategy of the Czech Republic and its implementation plan. They critically analyze the document's ambitious goals, such as aspiring to be in the top 10 AI countries, and the lack of concrete steps or prioritization to achieve these aims, highlighting the excessive generality and bureaucratic hurdles. We also discuss the role of state interventions versus market-driven approaches in fostering AI development and express skepticism about the strategy's effectiveness in its current form.blog post with sources and summary

In this episode of the Data Science Bulletin podcast, hosts Martin and Jakub discuss the latest developments in the world of AI, focusing on the release of the GPT-5 model. They share their personal experiences and disappointments with the model's performance and the implications of its cost-cutting measures. Additionally, they explore the broader impact of AI's integration into everyday life and discuss the concept of open-source politics and its potential influence on governance.Blog post and sources

In the 14th episode of the Data Science Bulletin podcast, hosts Martin Münch and Jakub Kramata discuss the latest advancements in AI regulations and infrastructure, focusing on the America's AI Action Plan announced by Donald Trump. They explore the plan's emphasis on reducing bureaucracy, enhancing AI infrastructure, and fostering international AI diplomacy, all while weighing its potential impacts against European approaches. The episode also touches on the security challenges faced by tech companies, highlighted by recent data breaches in popular applications.Blog post and sources

In this episode of the Data Science Bulletin Podcast, hosts Martin Münch and Jakub Kramata delve into recent AI reports, including the 2025 AI Index Report, AI Engineering Report, and a boxed State of AI Report. They discuss the implications of AI adoption across industries, the evolving roles of data scientists versus AI engineers, and the increasing relevance of human-in-the-loop systems. The hosts also highlight the importance of infrastructure and evaluation frameworks in the successful integration of AI technologies.DSB Blog

In the 12th episode of the Data Science Bulletin podcast, hosts Martin Münch and Jakub Kramata discuss two intriguing topics. They delve into the recent news of Meta reportedly offering $100 million signing bonuses to researchers, with some from OpenAI already making the jump, although there are debates about the details and offers. The episode also explores Revolut's upcoming AI-driven banking assistant and reflects on Revolut's unique place in the fintech landscape, emphasizing the company's focus on customer experience and innovative banking solutions.Blog post

In this episode of Data Science Bulletin, hosts Martin and Jakub, delve into the current state of reasoning models in AI, with a focus on a recent research paper from Apple entitled "The Illusion of Thinking." The discussion highlights the paper's critique of current benchmarking methods for AI models, emphasizing issues like data contamination and questioning the true reasoning capabilities of these models. The hosts also explore the broader implications of these findings and how they fit into the ongoing debate about the generalizability and limitations of AI reasoning.Fo more: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-11-illusion-of-thinking/

In the 10th episode of the Data Science Bulletin podcast, Martin and Jakub celebrate reaching this milestone by discussing their journey and roles as team leaders in the data science field at ČSOB. They reflect on the challenges and learning curves faced as they transitioned from individual contributors to team leaders, emphasizing the importance of delegation and mentorship. Additionally, they touch on topics from the recent Trendy Práce conference, exploring the implications of AI on the future of work and education.Related blog summary: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-10/

In this episode, we are joined by Ondra Draganov, a research mathematician at INRIA in France, who shares his 15-year academic path from Prague to Vienna and now to France. Ondra delves into his work in topological data analysis, discussing its application and relevance to various fields like biomedicine, while also exploring its conceptual complexity. The conversation wraps up with a look at the potential of generative AI in research and Ondra's personal approach to coding and utilizing libraries.More: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-9-exploring-topological-data-analysis-with-ondra-draganov/

This episode we talked about a trip to Darmstadt where PyCon and Pydata DE happend. More about here: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-8-pycon-de/

This episode we dedicated to the material AI-2027. Prediction circulating social networks how the world will look like in 2 years. Accompanying blog post: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-7-ai-2027/

In this episode, we dove headfirst into the hotly debated question: Are programmers on the road to extinction, or will coding roles simply evolve? Here’s a concise look at the highlights, insights, and spirited discussions from the show.Details: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-6-future-of-coding/

Conversation about data science and AI. This episode we went through Klarna story and AI consultanciesSummary and links: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-5/

In the 4th episode we went through topics of 2024 that we found interesting to mention and comment. Summary and links: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-4/

We went through the question "how to get into data science" in the third episode recorder just before Xmas 2024. Audio is in Czech, but AI generated episode summary is in English. Tips on online academies, courses, books and what to do next. Summary and sources: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-3/

DSB podcast #2 [CZ] by Data Science Bulletin

We are bringing you new DSB experience – podcast. Audio is in Czech language and we will try to accompany every podcast epizode with insights in english – mostly AI generated from the original content (thanks to Whisper, LLMs and more).More info about epizode: https://datasciencebulletin.com/dsb-podcast/dsb-podcast-1/Feedback welcomed to hosts directly or to [email protected].