Because it’s Friday: The Music Video Where Physics Wrote the Song
The universe arranges itself into patterns, completely invisible, right in front of us. Every so often someone finds a way to make it visible. There is a music video on YouTube that is also a physics demonstration, and it is one of the most beautiful things on the internet. It’s called Cymatics, by a New […]
Because it’s Friday: Wrong in Real Time
A nonprofit research group called METR (Model Evaluation & Threat Research) ran a randomized controlled trial earlier this year to measure what AI coding tools actually do to developer productivity. Sixteen experienced developers, working on real open-source projects they’d contributed to for years, completed 246 tasks – some with AI tools, some without, randomly assigned. […]
Because it’s Friday: The Mozzarella Phase
A group of Italian physicists got tired of their cacio e pepe turning clumpy when cooking for big groups, so they did what any reasonable scientist would do: they mapped the phase diagram of the cheese sauce. The paper, published in Physics of Fluids and titled “Phase behavior of Cacio e Pepe sauce,” won the […]
What Regularization Does (and Why Your Model Needs It)
Overfitting is not a mistake the model makes. It is what happens when a model does exactly what it is told. Regularization in machine learning is not just a way to prevent overfitting. It changes what the model is allowed to learn in the first place. Instead of only minimizing error on the training data, […]
Because it’s Friday: Spurious Scholar
Tyler Vigen’s Spurious Correlations has been a beloved corner of the internet since 2014, pairing line charts of things like Nicolas Cage film appearances and swimming pool drownings to illustrate, with great affection, why correlation is not causation. The site now has over 25,000 variables to mine for accidental patterns, which is already more than […]
How I Vibe Coded an AI Assistant for My Blog in One Weekend
Last weekend I built an AI assistant for this blog in about a day, using roughly five dollars of cloud services and a handful of open-source tools. The idea came after Liz Saunders of Type A Circle ran an office hours session on vibe coding. She walked through the concepts clearly and showed what was […]
How Machine Learning Works: It’s Not Learning
A machine learning model doesn’t work the way you probably think it does. When most people hear the word learning, they picture something like understanding: a system that studies examples, figures out what they mean, and applies that knowledge to new situations. What actually happens is narrower than that. A model makes predictions, measures how […]
Because it’s Friday: The Animal Fart Database
Somewhere between a Twitter joke and a peer-reviewed research tool lives #DoesItFart, a crowdsourced spreadsheet maintained by biologists cataloging which animals pass gas and which don’t. It started in 2017 when zoologist Dani Rabaiotti couldn’t answer her teenage brother’s question about snakes, tweeted at a snake expert, and then a third scientist decided this should […]
Machine Learning for Non-Technical Professionals
Why Your “Silly Questions” Are Product Requirements We have a technical writer on our team. She’s sharp, curious, and exactly the kind of non-technical professional who needs to understand machine learning to do her job well. Yesterday after a team meeting, she sent me a Slack message. My boss had explained that we needed to […]
A Friendly Introduction to Principal Component Analysis
Most datasets don’t have two or three measurements per observation. They have dozens, sometimes hundreds. A patient record might include blood pressure, cholesterol, glucose levels, BMI, age, and dozens of lab results. A manufacturing sensor array might log temperature, vibration, pressure, humidity, and electrical output from every machine on the floor, every few seconds. The […]
Machine Learning in Healthcare: AI Scribes and Heart Disease Screening
How ambient AI scribes turn doctor-patient conversations into clinical notes, and how a University of Michigan model detects hidden heart disease from a 10-second EKG.
Machine Learning in the NBA: Player Tracking and Injury Prevention
Every NBA arena has cameras mounted in the catwalks that track every player and the ball at 25 frames per second. Machine learning in the NBA turns that tracking data into pregame scouting reports, real-time broadcast graphics, and the injury prevention models that decide whether a player sits out practice. NBA Player Tracking Technology From […]
Strike Zone Prediction Model: How Neural Networks Analyze MLB Statcast Data
Baseball has always been a numbers game. Batting averages, earned run averages, and RBIs have been part of the sport for over a century. But since 2015, MLB Statcast and machine learning have pushed that much further. Statcast now generates up to seven terabytes of data per game, and ML models turn that raw data […]
How Formula 1 Teams Use Machine Learning to Win Races
In Formula 1, milliseconds decide races. The cars generate massive amounts of data, around 400 GB per race from sensors tracking aerodynamics, fuel usage, tire wear, and driver behavior. Add in data from simulators and previous races, and no human team can analyze it all in real time. That’s where machine learning comes in. And […]
Machine Learning in Sports: How AI Predicts NFL Tackles
This is the first post in the “ML in the Wild” series, where we look at specific machine learning applications in the real world. We’re starting with one football fans might recognize: the tackle probability statistics that appear during NFL broadcasts. What AWS Built The NFL partnered with Amazon Web Services (AWS) to build a […]
Bias Variance Tradeoff: Why Your ML Model Can’t Have It All
For years, every time I fixed one problem with my models, I created another one. Make the model more sophisticated to capture complex patterns? It performs terribly on new data. Simplify it to work better on new examples? Now it’s missing obvious relationships in the training data. If that sounds familiar, you’re experiencing the bias […]
Training vs Testing Data: ML Models Must Prove Themselves
Does this sound familiar? You’re tutoring a student for an upcoming math test. You help them solve dozens of practice problems over several days, and by the end, they’re getting every problem right. You feel confident they’ve mastered the material. The same problem shows up when splitting training and testing data in machine learning, and […]
Inference vs Prediction in Machine Learning: What’s the Difference?
Some machine learning applications can make accurate recommendations but can’t explain the reasoning. Others provide clear explanations but aren’t quite as precise. That tradeoff between inference and prediction in machine learning shows up early in any project: do you need the most accurate answer possible, or do you need to understand why that answer makes […]
Classification vs Regression: Predicting What vs. How Much
In our previous post, we explored supervised vs. unsupervised learning. Now we’re diving into another fundamental choice you’ll face in every supervised learning project: are you trying to predict what category something belongs to, or are you trying to predict a specific number? This might sound like a subtle distinction, but it completely changes how […]
Supervised vs Unsupervised Learning in Machine Learning: What the Names Don’t Tell You
Remember learning to ride a bike? Some of us had a parent running alongside, holding the seat and shouting guidance: “Pedal faster! Look ahead! Balance!” Others figured it out through trial and error, falling, getting back up, and gradually finding that magic balance point with no one giving instructions. These two approaches perfectly capture the […]