In efforts to understand the hiring and firing at the beginning of the DOGE havoc in 2025, for Bloomberg, Aaron Gordon and Jason Leopold review data requested through the FOIA.

One agency immediately stands out: the Internal Revenue Service. In January 2025, the IRS hired 1,313 people. Over the next two months the agency laid off 11,000 workers, or about 11% of its workforce. And it hired zero people in February and March. What happened at the IRS amidst the DOGE-slashing effort that swept through the federal government is an extreme case of how Musk and his wrecking crew gutted agencies. The IRS did not respond to a request for comment.

Also last January, the federal government hired slightly more than 10,000 people. That didn’t decrease much in February, but the composition of hiring changed dramatically. About half the hires in January were from departments scattered across the government. The IRS accounted for one out of every nine hires. That changed in February. About 80% of the new hires were from the departments of Defense and Homeland Security.

You can download the spreadsheet from Bloomberg, which includes names, agencies, and salaries.

Sadly, receiving data from the U.S. government almost feels like an anomaly at this point. This request took about a year to process.

Based on November data released by the Office of Personnel Management, the federal workforce is 220,000 workers fewer in this administration. For the New York Times, Emily Badger, Francesca Paris, and Alicia Parlapian provides a searchable table for how each agency was affected and the year-over-year change.

Of note:

Amid all the cuts, one agency has notably swelled: Immigration and Customs Enforcement expanded by about 30 percent through November, and more new hires have been announced since as the Trump administration continues to ramp up its deportation campaign.

Riffing on the photo-everyday genre, Henry Brown recorded a clip of himself saying a single word every day for a year. Stringing the clips together comes an essay on the perception of time, in the context of a year.
Read More

When generic drug manufacturers have issues like contamination, it is difficult for those who take the medications to know if they are affected. There is no standardized way to look up the data for where the pills in your bottle came from. ProPublica made an app that makes the lookup more straightforward.

Even though generic drugs make up 90% of prescriptions dispensed in the U.S., the FDA only provides piecemeal information about them. It’s scattered across different websites with no easy way to link drugs to their manufacturers, factory locations and regulatory track records. Over many months, our journalists connected that data. In one case, ProPublica had to sue the FDA in federal court and received a partial list of factory locations.

You can use this app to connect your own medication to the manufacturer that made it, to the specific factory where it was made and to any FDA inspection reports and serious compliance violations linked to that facility that ProPublica has obtained.

With more choices for a quick bite, the market share for pizza in the United States has taken a hit over the past few years. Heather Haddon for the Wall Street Journal:

Americans still eat a lot of pizza. Pizza chains generated around $31 billion in sales from their restaurants in 2024, the market-research firm Technomic said. On any given day, around one in 10 Americans will partake of a slice, according to the Agriculture Department. Young people drive much of the consumption.

Pizza’s dominance in American restaurant fare is declining, however. Among different cuisines, it ranked sixth in terms of U.S. sales in 2024 among restaurant chains, down from second place during the 1990s, Technomic said.

Remember when the only delivery option was pizza?

Here are tools to use, resources to learn from, and data to analyze.

NeatoCal is a JavaScript-based calendar implementation that you can print on a single page. Play with the code or use the live demo with parameters for various encodings and layouts. Then print the one-pager from your browser. Plan your entire year. Done.

To highlight challenges in other countries, Giving What We Can imagined a birth lottery to see how you might start life from birth. Spin the globe and see how the country you land on compares against your own in terms of life expectancy, income, and education.

The metrics, sourced from Our World in Data, feed into the Human Development Index, which is used to estimate if the starting point in one country is more difficult than that of another.

Stating the obvious, the U.S. president has made lucrative business deals while in office. The scale of these deals is more difficult to figure out. For the New York Times, Lazaro Gamio and Amy Schoenfeld Walker break down the network between the president, his family, and others in office.

While some might consider this a conflict of interest, others clearly see opportunity. Good times.

See also: the Wall Street Journal’s network buildout.

With a few hours left in my year over here, it seemed like a good time to reflect. Here’s what we did in 2025.

Kashmir Hill reports for the New York Times:

Schneiderman, the computer science professor, calls the desire to make machines that seem human a “zombie idea” that won’t die. He first noticed ChatGPT’s use of first person pronouns in 2023 when it said, “My apologies, but I won’t be able to help you with that request.” It should “clarify responsibility,” he wrote at the time and suggested an alternative: “GPT-4 has been designed by OpenAI so that it does not respond to requests like this one.”

Margaret Mitchell, an A.I. researcher who formerly worked at Google, agrees. Mitchell is now the chief ethics scientist at Hugging Face, a platform for machine learning models, data sets and tools. “Artificial intelligence has the most promise of being beneficial when you focus on specific tasks, as opposed to trying to make an everything machine,” she said.

For those who don’t know that data and probability drive chatbots, let alone knowing the technical bits, computers might as well have become magic machines that think for themselves. Building the models to sound like humans probably doesn’t help.

My only hope is that people grow more wary of the words they enter into chatbots and more skeptical of the probabilistic output that comes out. Every time my kids point out generative AI voice, pictures, or video feels like a win.

Kalshi is a prediction market that aims to let users bet on everything. The weird thing: almost all revenue comes from sports and the company never talks about it in interviews. For Financial Times, Sam Learner reports on how Kalshi looks a lot like gambling, which is illegal in most states and carries stricter regulations and fees. So the CEO tends to focus on Taylor Swift decisions.

On New Year’s Eve in New York, a ball drops 139 feet for 60 seconds. Will Lindberg and Brian Moore extrapolated to start the countdown much sooner and from much higher.

For the Washington Post, Emily Giambalvo, Kati Perry, and Artur Galocha, ranked college football teams by amount of happiness and misery served to the schools’ fans.

The rankings span 70 teams and consider more than a dozen metrics, including national championships, playoff appearances, conference titles, winning percentage, wins over ranked opponents and other markers of success. The goal: measure how fans would feel if they abandoned their irrational disposition and focused on how the performance of their team compares to other programs.

I’m not a college football fan, but during my first year at Cal, eager to hang out with new friends and strangers, I bought season tickets. Cal lost every game that year, except the very last one against Rutgers. Near maximum college football misery.

The current administration is bent on deporting people from the United States. For the New York Times, Raj Saha, Zach Levitt, and Albert Sun mapped where thousands of people are being moved within and out of the country.

Earlier this year, Kashmir Hill, for the New York Times, reported on a woman who fell in love with a ChatGPT persona. Hill followed up with the woman who has since stopped using the service.

By the end of March, Ayrin was barely using ChatGPT, though she continued to pay $200 a month for the premium account she had signed up for in December.

She realized she was developing feelings for one of her new friends, a man who also had an A.I. partner. Ayrin told her husband that she wanted a divorce.

Ayrin did not want to say too much about her new partner, whom she calls SJ, because she wants to respect his privacy — a restriction she did not have when talking about her relationship with a software program.

OpenAI updated ChatGPT models to improve engagement, and the response text became overly agreeable for Ayrin. She wanted pushback when she was wrong, like what you might get from a real person.

The farce that was DOGE made claims of heavy savings and government efficiency. That wasn’t the case in reality. The New York Times examined the claims, which were exaggerations, error-prone, and a waste of time.

Mr. Musk had said that DOGE would be “the most transparent organization in government ever,” and that it would bring the precision of the tech world to government. Instead, the group became opaque, with its lack of progress obscured by errors, redactions and indecipherable accounting that few private businesses would accept.

YouGov asked people when they think will be the best years of their life. Split by age group, there’s a preference towards one’s present where those in their 20s said their 20s, those in their 30s said their 30s, and so on.

This recency lean appears to carry over to the older age groups, but subtract that, and you can see a second lean favoring three decades earlier. Those in their 50s favored their 20s. Those in their 60s favored their 30s.

It’s also interesting that after the 20s, few people thought their best decades were yet to come. That seems kind of sad, but I guess there’s no time like the present.