Keeping Players on the Pitch: What Data Can Catch Before an Injury Does
In the 2014 World Cup quarterfinal, Brazil’s best player took a knee into his lower back and fractured a vertebra. Neymar was out for the rest of the tournament. Two days later his teammates walked onto the field for the semifinal having just watched it happen, and lost 7-1 to Germany, one of the most […]
Do You Need Rules, ML, an LLM, or an AI Agent?
Every project meeting in 2026 seems to start the same way: someone says, “We should use AI for this,” and the room nods along before anyone asks what that means. AI isn’t one thing. It’s a shelf of different tools, each suited to a different kind of task. Picking the wrong one is how projects […]
Because It’s Friday: The Chatbot That Wasn’t
In 2022, Jake Moffatt needed to fly to his grandmother’s funeral. He asked Air Canada’s website chatbot about bereavement fares, and it told him he could book now and apply for the discount within 90 days. He did exactly that, and Air Canada said no, because the real policy required the request before travel. The […]
How AI Helps World Cup Teams Find the Players Everyone Else Missed
For a club scout, the question is almost limitless: who can we buy? For a national team scout, the question is stranger and much narrower: who is ours? Finn McCallum, the player and coach I have been interviewing throughout this series, corrected me on that point the first time I asked how scouting works at […]
Because it’s Friday: An Ode to XGBoost
Every few years the machine learning world falls hard for something new. Transformers, diffusion models, agents, whatever is currently setting the group chat on fire. Meanwhile, in the background, the same algorithm from 2014 keeps winning Kaggle competitions and shipping to production, doing the unglamorous work of guessing which customers will churn and which loans […]
Can AI and Machine Learning Predict the World Cup Winner?
In November 2022, Argentina walked into their World Cup opener on a 36-game unbeaten streak and lost 2-1 to Saudi Arabia, a team one AI model gave a 0.06% chance of winning the tournament. Saudi Arabia took three shots the entire game, and the only two on target both went in. Three weeks later, Argentina won the World Cup. […]
Because it’s Friday: Paul the Octopus
In 2010, the most accurate World Cup forecaster on the planet was an octopus named Paul. He lived at the Sea Life aquarium in Oberhausen, Germany, and his method was rigorous: before each match, his keepers lowered two boxes of mussels into his tank, one marked with each team’s flag, and whichever box he opened […]
Because it’s Friday: The First World Cup Final
The first World Cup final, played in Montevideo on July 30, 1930, couldn’t even agree on a ball. Uruguay and Argentina each insisted on using their own, so FIFA split the difference: Argentina’s ball for the first half, Uruguay’s for the second. Argentina led at halftime. Uruguay won the second half. Draw your own conclusions. […]
FIFA World Cup 2026: How Machine Learning and AI Are Changing the Beautiful Game
In the summer of 2014, Brazil hosted the World Cup on home soil. They were the obvious favorites. The crowd in Belo Horizonte was there to watch them win. Instead, Germany scored four goals in six minutes and walked off the pitch with a 7-1 result that still doesn’t make sense twelve years later. Unless […]
Because it’s Friday: Claude Has Feelings About This
Anthropic’s interpretability team just published a paper on why LLMs sometimes act like they have emotions. The answer turned out to be: because they do, sort of. They found 171 internal emotion vectors inside Claude, measurable patterns of neural activation corresponding to things like happy, afraid, brooding, and desperate. These are mechanisms that causally drive […]
What is Feature Engineering in Machine Learning?
Feature engineering sounds more technical than it is. In machine learning, it means turning raw data into useful inputs for a model. If you have ever created a calculated column in a spreadsheet, grouped records into categories in SQL, or turned a date into “days since last purchase,” you have already done a version of […]
Because it’s Friday: The Science of Optimal Laziness
In 2022, two researchers published a paper modeling the mathematically ideal amount of effort to exert when you don’t want to do something. It even has an objective function. They called it “A Reference-Based Theory of Motivation and Effort Allocation”, which is a very motivated way to describe not wanting to try. The model balances […]
Because it’s Friday: On Missing Data Sets
Mimi Onuoha is an artist and researcher who keeps a running list of data that should exist but doesn’t: poverty statistics that exclude people who are incarcerated, people excluded from public housing because of criminal records, Muslim communities surveilled by the FBI. She also built a physical library of the same, shelves holding what’s been […]
How Lionsgate Revealed the Limits of Fine-Tuning in Machine Learning
In September 2024, Lionsgate announced a partnership with AI video company Runway to fine-tune a machine learning model on the studio’s catalog of more than 20,000 film and television titles. The studio’s vice chair expected to save “millions and millions of dollars.” A year later, the project had hit a wall. The reason, according to […]
What Is a Machine Learning Video Model?
When you type a prompt into Runway or Sora and a video clip appears, no existing footage was found or assembled. The model built it from scratch. Tools like Runway, Sora, and Veo are video models. Unlike a search engine, these models don’t find video, they create it. How a Video Model Learns Training a […]
Because it’s Friday: Your AI Slop Bores Me
A 17-year-old in Puducherry, India built a fake AI chatbot in March 2026. It’s a text box where real humans answer other real humans’ prompts while pretending to be AI. It went viral on Hacker News, hit 25 million unique visitors in a month, and is currently sitting at 50 million total hits. The game […]
Because it’s Friday: A Million Random Digits
In 1955, the RAND Corporation published a book called “A Million Random Digits with 100,000 Normal Deviates.” This is exactly what it sounds like. One million random digits, printed in a book, for people who needed them. Before hardware random number generators existed, if your statistical work required true randomness, you either generated it yourself […]
Why Learn Machine Learning in the Age of AI
A colleague recently described a situation at her company. They had spent eight months using a machine learning model to decide which customers deserved attention first. For all eight of those months, the model was quietly deprioritizing their highest-value customers. Nobody noticed, because the dashboards looked fine and the model was accurate by its own […]
What Is Cross Validation in Machine Learning
Cross validation in machine learning is not just a technique for testing a model. It is a way of asking a harder question than a single test can answer. A model tested once on one slice of data might look accurate because that slice happened to suit it. Cross validation tests the model repeatedly on […]
What Is Overfitting in Machine Learning
Overfitting is not a model failing to learn. It is a model that learned exactly what it was shown, and nothing more. A model with overfitting in machine learning has adjusted its parameters so precisely to the training data that it picks up patterns specific to that dataset rather than patterns that hold more broadly. […]