This is from nearly a week ago, but it's currently blowing up on Bluesky.

You click into a server. A blur. A prompt asking you to prove you’re an adult before you can see what’s behind it. If you’ve hit this on Discord recently, you weren’t hacked and you didn’t violate anything. The platform shifted to a teen-by-default model — and now adult content access and certain safety-setting changes require age verification before Discord will budge. This isn’t a minor tweak. It’s a structural reset, already enforced where local laws require it, with a broader global rollout delayed until the second half of 2026 while Discord adds more verification options and documentation, per the company’s support page. This shift reflects a broader push around age laws reshaping how platforms handle minors online.

Job postings and interviews with senior security officials at Anthropic show that the frontier AI lab is building out an extensive monitoring system to keep tabs on activists who oppose the rapid development of artificial intelligence.

In addition to monitoring activists in the vicinity of Anthropic executives and keeping tabs on protests near physical Anthropic assets, the firm is also implementing a “pre-crime” approach, attempting to predict incidents before they happen. In some cases, that also means reporting suspects to police before a crime occurs. Anthropic did not respond to the Prospect’s request for comment.

Anthropic’s plans to surveil dissent are at odds with the firm’s efforts to cast itself as the responsible alternative to OpenAI. At the beginning of the year, the Department of Defense and Anthropic engaged in a high-profile dustup over Anthropic’s refusal to allow the military to use its tools for mass domestic surveillance and autonomous weapons. That tension seems to have eased as Anthropic hires for “national security sales” positions, seeking to restart military contracts. The increase in threat monitoring of domestic opponents fits with a renewed focus on national security.

I was laid off at my then-local paper in August 2014, which I'd known was coming because I was only hired knowing the copy desk was being centralized later that year. There was no plan to come down here.

I got unemployment, but a third of newspaper wages only goes so far.

After several months of attempting to find work locally, I reached out to the director in Texas who'd offered to immediately hire me when centralization hit but offered an insulting pay cut.

By April, I was worn down enough to reach out again. He was willing to negotiate only an 18% pay cut, but that's still a lot better than zero dollars, and the household wasn't functioning.

My ex and I bonded over being equally manipulative back in 2009, so I get the U-Haul and announce I'm moving to Texas.

It worked.

Simply saying I was going to wouldn't have. She had the only vehicle at the time, so "moving in what?" would have been a valid initial question.

But this was not my goal. She'd already said in August that Texas was not an option. And so I stayed in rural Oregon, interviewing for media-adjacent jobs ... school PIO, tech at a TV station ... but the UI benefits run out at a certain point, and if you have one "can you start tomorrow?" up your sleeve, that's the only remaining option.

In terms of manipulative things I've done in life, this likely ranks as a 7/10 short-term, 10/10 long term. But I saw the options as being homeless (again) or an unpleasant move -- I'd explicitly forbidden Texas my entire career (I may have been able to start my career a few months earlier had I been willing to move to San Angelo in 2001), so this was practical, not a goal.

She has to have known at some level that it was a farce. Not the job, but the U-Haul. I'd packed nothing up, and there was no discussion of division of particle board. In no universe am I getting out of there in a day.

However, this brings her to the table with a sharper sense of reality in about an hour. I think that whole charade ran me $30 all-in.

And later that day, I call the director and say I need to push my start date back by a couple of weeks. Then we start planning the move properly.

Perhaps not as properly as hoped. I arrived here by train on Memorial Day 2015, and flooding meant I had to get off the train at an ungodly hour without sleep to get on a bus for the final leg.

Not that the trip was any easier for her. She lost her stuffed animal in a hotel and nearly got arrested in Nevada for her wedding collar (it's apparently a deadly weapon).

What. The. Actual. Fuck?

Donald Trump said every adult citizen in the US will receive a dividend of $5,000 if Republicans retain the House and Senate in the midterm elections, while delivering a historically long convention speech in Dallas in which the president assailed Democrats as radical communists and defended his unpopular war in Iran.

“If the Republicans win, you win with us, and you get $5,000,” Trump told the Republican midterm convention in Dallas on Wednesday, claiming the payment was possible “because of our tremendous strength and success economically”.

Trump made the pledge, almost immediately raised ethical concerns, during a nearly two-hour long speech in which he also urged voters to “pretend that I’m on the ballot” – despite his historic unpopularity causing many vulnerable Republican politicians to avoid the event.

Speaking in Dallas, Trump claimed, without evidence, that his second term had been “the most successful two years, financially and every other way, in the history of the presidency”, but much of his speech also focused on fearmongering about what would happen should Democrats win the House or the Senate in November.

I will say I admire The Guardian for refusing to give him a title. They just work it in somewhere else, lowercase.

The universal adoption of electronic shelf labels in grocery stores across the US could cost tens of thousands of jobs and billions of dollars in lost wages while further driving up grocery costs, according to a report released on Tuesday.

An analysis by the AFL-CIO Tech Institute of the label manufacturers’ own marketing materials found the universal adoption of electronic shelf labels, which are marketed to retailers as a cost-cutting measure, could cost workers between $1.6bn and $6.9bn in lost wages annually and affect between 44,223 and 191,633 jobs.

The policy brief – Priced Out, Pushed Out: Electronic Shelf Labels Raise Prices and Shrink Paychecks – calls for a ban on electronic labels to protect consumers and workers.

Manufacturers have hailed the labels as a way to collect huge amounts of personal data from customers, increase profits and cut labor costs.

The report argues that electronic shelf labels use the same pricing systems that allow dynamic pricing – price changes such as increasing allergy medication prices when pollen counts are high – to surveillance pricing, changing prices based on personal data.

These technologies are being implemented across the US retail industry at a time when food prices have surpassed inflation and wage gains, increasing 33% over the past seven years.

I'm just old enough to remember price tags being on each item at the grocery store. UPCs were being phased in, but both systems were in use for several years as scanners finally became universal.

I don't think anyone is clamoring for a return to having to navigate around stockers putting price tags on each individual can of tuna from a case. The issue here isn't the tautological "new technology removes jobs;" it's that dynamic pricing introduces new points of failure (what if the price changes between grabbing something and checking out? You can't go take a picture of the price tag you saw at the time to prove you're being overcharged) and fundamentally changes the nature of shopping.

And that's before we get into personalized pricing, though loyalty cards have been offering that for years, so the real change here would be: Somehow the screen knows who you are and determines a price in a fraction of a second so you don't see it change.

I just need a loaf of bread, not some dystopian "god only knows what it'll cost" grocery store game.

Reading scores among 15-year-olds in the United States have fallen to the lowest level in about 25 years, according to an international education assessment.

On Tuesday, the Organisation for Economic Co-operation and Development (OECD) released findings from the 2025 Programme for International Student Assessment (PISA), a global metric for educational achievement.

The results showed that US reading scores fell 14 points from 2022, the last time PISA results were released.

The decline was part of a broader international trend, with an OECD average from 23 countries also dropping 16 points.

But the US stood out for the size of its performance gap. Some 25.7 percent of US students were ranked as low performers in reading, while 13.3 percent were flagged as top performers.

“Looking across all 90 participating countries, there was not a single country with significantly wider inequality in reading scores than the US,” Michael Kieffer, a professor of literacy education at New York University, told Al Jazeera. “What is clear is that we are under-serving many students.”

In a victory for Republican campaign committees, the Supreme Court issued an order on Friday that forces broadcast TV stations to cut the prices of election ads purchased by political parties and joint fundraising committees. The September 4 ruling came in response to a petition filed by the National Republican Congressional Committee and National Republican Senatorial Committee.

The Supreme Court order was issued just in time for the start of a 60-day period before the election, in which broadcasters are required by US law to offer ad discounts to individual candidates. Because of the top court’s order, TV stations must also give their lowest ad rates to political parties and joint fundraising committees, which face fewer limits on how much money they can raise and spend.

The legally required discount is known as the “lowest unit charge,” or LUC. A US law that applies to any licensed station that airs election ads requires that the lowest price be charged for “the use of any broadcasting station by any person who is a legally qualified candidate for any public office in connection with his campaign.” The main legal question is whether “use… by” a candidate can refer to ad time purchased on a candidate’s behalf by parties and joint fundraising committees.

The Federal Communications Commission this year issued a public notice ordering broadcast TV stations to give the discounts to parties and joint fundraising committees. While both major parties could take advantage of the discounted ads, the Trump administration has pushed for the change, and four Democratic candidates contested it in court. The Democratic candidates won at the US Court of Appeals for the Fourth Circuit, where a judges’ panel found that the FCC public notice contradicts the plain language of US law that limits the discount to individual candidates.

The Republican campaign committees then filed an emergency motion for a stay that would block the Fourth Circuit ruling, and the Supreme Court granted the stay on Friday. The Supreme Court didn’t rule on the merits of the case but said the Fourth Circuit should not have intervened because the FCC hasn’t made a “final” decision on a challenge filed by the Democratic candidates.

In light of Donald Trump’s attempts to rename a gulf, lakes and even states after “America”, South Park, the animated satirical program that has long skewered those in power, has announced an official name change. For its 29th season, the Comedy Central series will be known as “South America”.

“Inspired by the bravery and patriotism of Apple and Google, we are changing the name of South Park to SOUTH AMERICA,” creators Matt Stone and Trey Parker said in a statement. “We especially want to thank our parent company Paramount – a Skydance Capitulation.” The show’s parent company, Paramount Skydance, owned by Trump ally David Ellison, is poised to become one of the biggest media conglomerates in the US after the Trump administration approved its $111bn merger with Warner Bros Discovery (WBD), drawing much creative industry backlash.

This is a story that could have handled some editing. Just ignore the last section, which you've already read.

Drilling miles into the Earth to tap energy hot enough to melt lead usually breaks equipment before it breaks records.

Texas-based Mazama Energy has announced a drilling performance milestone at its Athena geothermal well in Newberry, Oregon. It is a key step toward commercializing superhot rock geothermal energy.

The energy company said its latest geothermal well, Athena, reached a depth of 10,350 feet in a swift 15 days. Interestingly, it is roughly an 80 percent reduction in drilling time compared to a landmark well drilled at the same site just last year.

The development will help move the clean energy sector a step closer to tapping “superhot rock.” These subsurface formations are hotter than 750°F (399°C)^[after you've already said "hotter than," you can round up to 400C] and hold enough energy to transform green grid economics.

“Our goal was to demonstrate that we could drill faster and be cost-effective as we move toward increasingly hotter and deeper conditions,” said Pat Brand, Project Athena Director and Mazama’s Drilling SME. “We still have work ahead as we drill toward 750+°F temperatures, but the performance we are already seeing gives us confidence.”

On Friday, the Fifth Circuit Court of Appeals affirmed the dismissal of a lawsuit against the city of Jackson, Mississippi, which alleged the city knowingly let residents drink lead-contaminated water.

The lawsuit, first filed in 2022, says the contamination violated Jacksonians’ constitutional right to bodily integrity. Donald Trump-appointed judge Kurt Engelhardt disagreed. “The Constitution does not provide redress for every governmental wrongdoing,” Engelhardt said. “Defendants’ actions allegedly deprived Plaintiffs of clean water and guileless information. These deprivations, while grievous, do not infringe upon any deeply rooted constitutional right.”

That may be true in Mississippi—but on the state level in at least nine US states, and in more than a dozen other countries, clean drinking water is in fact a protected right. In countries including Uruguay and South Africa, a right to clean water is enshrined in the constitution. New York’s state constitution, as of a 2021 amendment, states that “each person shall have a right to clean air and water, and a healthful environment.” So too in Massachusetts, where “the right to clean air and water” has been constitutionally protected since 1972.

But according to the Fifth Circuit ruling, the US constitution is too narrow for such a claim. The Jackson plaintiffs alleged that their city government deliberately withheld information about the water’s toxicity, and exacerbated the problem through a series of boil-water notices—a practice that can eliminate some bacterial contamination, but also concentrates lead in water. Engelhardt stated there is no “constitutional right to truthful information from officials during a public health crisis.”

Picture humanity in a boat being swept down a raging river, praying there is no Niagara Falls ahead. Or imagine standing with the pioneering physicists in 1942 before they triggered the first self-sustaining nuclear fission chain reaction beneath a Chicago stadium.

These were two of the analogies used this week by Prof Robert Trager, an expert in the nascent field of AI governance, to describe the perilous but potential-filled moment the world stands at with accelerating artificial intelligence.

Trager, the director of the Oxford Martin AI Governance Initiative, was speaking in the week OpenAI claimed the technology had crossed the threshold known as AGI – artificial general intelligence – with its newest model, GPT-6 Astra.

That claim came as the company prepared for a potential $850bn (£630bn) stock flotation, so included a dose of marketing spin, but if true it may be significant. The San Francisco company defines AGI as “autonomous systems that outperform humans at most economically valuable work”.

The tasks it claims Astra can automate include designing circuit boards, filling out tax returns, building video games, financial modelling, engineering design and helping assemble legal documents. The threat to some white-collar jobs is implicit.

The UN general assembly will vote on Friday on a resolution for the world to formally phase out the traditional Mercator map of the globe in favour of one that more accurately displays Africa’s size.

Robert Dussey, Togo’s foreign affairs minister, said the resolution was sponsored by Togo and backed by countries including the African Union’s 54 other member states. He did not specify the total number.

“This discussion is not a political discussion [but] a scientific discussion, because there is scientific proof so we all have to accept the reality,” he said.

On Friday before the vote, France’s foreign minister spoke in support of the resolution. “[We] will abandon the Mercator projection in favour of map representations tailored to specific uses, more faithful to actual surface areas and in line with the recommendations of cartographers and scientists,” Jean-Noël Barrot said during a speech at the Sorbonne University in Paris, in a move seen as part of France’s push to increase goodwill among its former colonies.