\\articles
Notes and thoughts on data science, machine learning, LLMs, and software engineering.
Are Days-Long Dashboards Becoming History? Claude and the Future of Data Analysts
•6 min readA new kind of BI that connects straight to the data warehouse and writes live SQL. What will it solve, and where will it fall short?
#data-visualization#ai#data#claude#data-analysisThe “Numbers Don’t Lie” Lie: Making the Wrong Decision with the Right Data
•4 min readCode that runs without errors is not the same as a correct result. How one silent JOIN mistake can push management into a costly wrong decision.
#data-analyst#data-visualization#data#data-analysis#data-scienceThe Hidden Bill Behind the AI Boom: Are We Mortgaging the Future While We Speed Up?
•3 min readAI assistants write hundreds of lines in seconds. Research on thousands of repositories shows what that speed costs once the code reaches production.
#coding#technical-debt#code#aiDatabase Performance: A Technical Detail or a Strategic Advantage?
•6 min readQuery speed is not just an IT metric. How indexing, query design and cloud cost management turn milliseconds into revenue, customer trust and lower bills.
#strategy#data#database#sql#optimizationWhy Do AI Models “Give Up” on Hard Questions?
•2 min readApple's “The Illusion of Thinking” shows that reasoning models think less, not more, once problems get hard enough. What that says about AI “thinking”.
#claude#ai#llm#apple#chatgptCode Hallucinations, Security Vulnerabilities and the Erosion of Technical Skill in AI-Assisted Software
•17 min readBringing LLMs into software development adds speed, but also hallucinated code and packages, new supply-chain attacks, falling code quality and a loss of engineering skill.
#coding#technology#information-technology#ai#security