It's just regression

Kill frequentist PROBABILITY in 3 words 😼

Undefined limiting proportion (The limiting proportion of successes as the number of trials tends to infinity has to be left vaguely defined as defining it precisely would render the definition of frequentist probability circular.)

Kill the whole of data science in 3 words 😬

It's just statistics (Better answers in the replies!)

Kill the practice of treating P values as measures of something that is beyond what they are simply by definition in 3 words 🧐

Uncontemplated null distributions (For example, does using P values to measure things like evidence, surprise or compatibility make sense when null distributions are uniform, truncated or multi-modal distributions with one of the modes in the tails?)

Kill all the hype around big data in 3 words 🤖

Routinely collected diarrhoea

"Mathematistry" - George Box (with the meaning of mathematics going beyond the bounds of common sense. Ironic that George Box coined this word because he was a supporter of objective Bayes methods!)

Kill standard P value adjustments for multiple testing in 3 words 🤨

Global effect obsession (when we want to know about the specific effects assessed by individual tests. Tricky unresolved problem in my opinion. Working on it! Anyone wants to collaborate?)

Kill all the hype around causal inference in 3 words 😵‍💫

Reinventing the wheel (or repackaging the old)

Kill type I error rate control in sequential analysis in 3 words 🧑‍🎄