Day: November 14, 2023

Early in the morning on October 21, 2016, Scott Shapiro got out of bed, opened his Dell laptop to read the day’s news, and found that the internet was broken.

Not his internet, though at first it struck Shapiro that way as he checked and double-checked his computer’s Wi-Fi connection and his router. The internet.

The New York Times website was offline, as was Twitter. So too were the websites of The Guardian, The Wall Street Journal, CNN, the BBC, and Fox News. (And WIRED.) When Twitter intermittently sputtered back online, users cataloged an alarming, untold number of other digital services that were also victims of the outage. Amazon, Spotify, Reddit, PayPal, Airbnb, Slack, SoundCloud, HBO, and Netflix were all, to varying degrees, crippled for most of the East Coast of the United States and other patches of the country.

Shapiro, a very online professor at Yale Law School who was teaching a new class on cyber conflict that year, found the blackout deeply disorienting and isolating. A presidential election unlike any other in US history loomed in just under three weeks. “October surprises” seemed to be piling up: Earlier that month, US intelligence agencies had jointly announced that hacker breaches of the Democratic National Committee and Hillary Clinton’s presidential campaign had in fact been carried out by the Russian government. Meanwhile, Julian Assange’s WikiLeaks had been publishing the leaked emails from those hacks, pounding out a drumbeat of scandalous headlines. Spooked cybersecurity analysts feared that a more climactic cyberattack might strike on Election Day itself, throwing the country into chaos.

Those anxieties had been acutely primed just a month earlier by a blog post written by the famed cryptographer and security guru Bruce Schneier. It was titled “Someone Is Learning How to Take Down the Internet.”

“Over the past year or two, someone has been probing the defenses of the companies that run critical pieces of the internet,” Schneier, one of the most highly respected voices in the cybersecurity community, had warned. He described how an unknown force appeared to be repeatedly barraging this key infrastructure with relentless waves of malicious traffic at a scale that had never been seen before. “These probes take the form of precisely calibrated attacks designed to determine exactly how well these companies can defend themselves, and what would be required to take them down. We don’t know who is doing this, but it feels like a large nation-state. China or Russia would be my first guesses.”

Finally, a robot to tell you which jacket you should wear to the function. Google DeepMind, the search giant’s AI-centric brain trust, just announced a new weather forecasting model that beats traditional systems more than 90% of the time. Named GraphCast, the machine learning model promises 10-day predictions that are better, faster, and more energy-efficient than the tools that run your weather app today.

“We believe this marks a turning point in weather forecasting,” Google’s researchers wrote in a study published Tuesday.

In general, the current model for forecasts is called “numerical weather prediction (NWP).” NWP plugs current weather conditions into enormous models that simulate upcoming changes based on the principles of fluid dynamics, thermodynamics, and other atmospheric sciences. It’s complicated, expensive, and calls for tons of computing power.

Instead of running simulations about how molecules will fly around and slam into each other, GraphCast breaks with tradition by placing a heavier emphasis on historical data. In other words, it’s a machine-learning model that makes predictions based on what happened in the past. There’s a lot of fancy computer science involved, but in general, it’s a lot simpler in terms of the level and number of computations it requires.

GraphCast starts with the current state of Earth’s weather, and data about the weather six hours ago. Then, it makes a prediction about what the weather will look like six hours from now. GraphCast then feeds those predictions back into the model, performs the same calculation, and spits out longer-term forecasts.

The Google team compared GraphCasts results to the current model that’s used for medium-range weather prediction, called HRES. According to the study, GraphCast “significantly” outperformed HRES on 90% of the targets used in the test.

GraphCast also had surprising success predicting extreme weather events including tropical cyclones and freak temperature changes, even though it wasn’t specifically trained to handle them.

The study authors say their work is meant to work alongside the standard systems meteorologists rely on. “Our approach should not be regarded as a replacement for traditional weather forecasting methods,” the study authors wrote. “Rather, our work should be interpreted as evidence that [machine learning weather prediction] is able to meet the challenges of real-world forecasting problems and has potential to complement and improve the current best methods.”