Sean HarkinData Scientists: Don’t Feel Like You Have to be Instant Experts on LLMsIt’s an easy temptation, but it’s better to be clear that most data scientists are still mastering this tech.1 min read·Jan 15, 2024----
Sean HarkinData Science for Refuting Bad IdeasOne of the most important uses of statistical methods3 min read·Jan 12, 2024----
Sean HarkinLimiting Learning CapacityThe most essential idea and the most counter-intuitive idea in data science 2 min read·Jan 12, 2024----
Sean HarkinAgile Project Management for Data ScienceYou need it because both data and code bases spring surprises 4 min read·Jan 12, 2024----
Sean HarkinThe Two Cultures in Data ScienceWe get better results if we have both statistics-type backgrounds and engineering-type backgrounds in our teams 2 min read·Jan 12, 2024----
Sean HarkinDon’t Leave Models Static for a Long TimeIt’s common practice in many businesses, but it has serious pitfalls1 min read·Jan 12, 2024----
Sean HarkinUnderstand Your Data Before You Model ItEven though it’s very tempting to race ahead to the most fun part of data science 1 min read·Jan 12, 2024----
Sean HarkinThe Different Kinds of Data Science PredictionAnd how to avoid allergic reactions to the word “prediction”3 min read·Jan 11, 2024----
Sean HarkinSaying No To Building A ModelThe importance of opposing building models when you know they will be bad.2 min read·Jan 10, 2024----
Sean HarkinMachine Learning for Explaining Your DataThe case for combining traditional coefficient inference with explainable ML techniques2 min read·Jan 8, 2024----