Storyboard When Averages Lie Why single-point predictions mislead and how full probability distributions can improve decisions—illustrated with a Parisian real estate case study. The Random Logic of Gemini 1.x A humorous look at Gemini Advanced's seemingly random response patterns in Europe, illustrated through pseudo-code that suggests a simplified decision-making process. This post contrasts Gemini's occasional brilliance with its inconsistent performance compared to other leading AI models. The Crystal Ball Fallacy Why confusing perfect prediction with certainty can lead to disaster — a quant trader's costly lesson about models and probability. Data-Driven or Data-Derailed? The misuse of predictive modeling intended for business purposes can lead to outcomes ranging from humorous to harmful—organizations need to carefully evaluate whether their implementations match their intended goals. When AI Draws Your Life An AI-generated depiction of my life included unexpected personal details—like my bicycle and a nod to Japan—prompting reflection on how technology interprets identity. From Fermat's Margin Notes to AI Apocalypse Tracing the unlikely journey from Pierre de Fermat’s margin notes to the cutting-edge world of AI, this post humorously explores how the humble pursuit of minimizing functions gave rise to technologies that spark debates about humanity’s future. The 42 Conspiracy Explore how Douglas Adams' iconic number '42' infiltrated global software development, leaving an indelible mark on random number generation (and possibly the fabric of reality). Cultural Hurdles of Innovation via Data Leveraging Drawing parallels between early aviation failures and contemporary digital transformation efforts, this essay explores why traditional companies often struggle with innovation and data initiatives despite significant investments. Like early flight pioneers who failed by mimicking birds without understanding aerodynamics, many firms pursue surface-level digital adoption without embracing the fundamental organizational changes necessary for success. Introduction to LLMs A non-technical overview of Large Language Models, exploring their potential, limitations, and customization for specific challenges. The Future With Generative AI All is well... Marvin Has Got a New Depiction DALL.E was asked to depict Marvin, the Paranoid Android from The Hitchhiker's Guide to the Galaxy. Generative Models and ChatGPT Slide deck prepared to support a couple of rather informal talks on generative models and ChatGPT. The targeted audience is ranging from folks with little technical literacy to mid-level data scientists. iQHorse—Towards Beating the Market Brief introduction to the online magazine iQHorse. Ideation: First Step Towards Innovation Slides prepared for a 'brownbag' talk on ideation. Are Decisions From a Single Point Wise? Slides prepared for a talk given at zeroG, a subsidiary of the Lufthansa Group specialised in data science. The targeted audience was rather broad, spanning from data scientists and business analysts, to senior executives. Competitive Dynamics Behind the Disruptor’s Dilemma The introduction of a potentially disruptive technology into a mature multisided ecosystem often faces a seemingly dilemmatic challenge: persuading a critical mass of incumbents to support the technology that might eventually disrupt them. We analyse this apparently perplexing phenomenon under the lens of a global coordination game of incomplete public and private information; and more precisely an adaptation of a game of regime change. Japan—A Macroeconomics Perspective Five macroeconomics issues about Japan are discussed. (i) Long-term Strengths and Weaknesses of the Macroeconomy; (ii) Trade Position and Competitive Advantage; (iii) Policy Options to Increase Net Exports; (iv) Impact of the 2007–2008 Financial Crisis and the World Economic Slowdown and (v) Possible Impacts on the Economy if Greece Decides to Withdraw from the Eurozone. Customer Centricity in the Airline Industry Traditional airlines in Europe are facing an increasingly fierce competition. In the meantime, the information era surfaces at an inordinate pace, bringing new technological means, new paradigms and new mindsets. The availability of data and the booming of social media have changed the game. In this personal essay, I share my naïve view on how the combo big data/machine learning could be harnessed in order for traditional airlines to regain a competitive edge. In particular, this combo provides a solid ground for transitioning from product centricity to customer centricity.
