After more than 2,000 hours of government-imposed connectivity blackouts, there were signs on Tuesday that Iran’s internet is coming back—at least at very low levels.

Iran’s more than 90 million citizens have been without internet for the overwhelming majority of 2026, between the current blackout that began on February 28, when Israel and the United States attacked the country, and a previous internet shutdown enforced after widespread protests in January. The reconnection appears to have been ordered by officials in Iran’s government—but could only be temporary.

Though some Iranian networks appeared to be connecting to the global internet on Tuesday, researchers cautioned that the level of access was far below even the partial restoration that Tehran allowed at the end of January and throughout February—and it was drastically below Iran’s typical baseline of global internet connectivity from December 2025. Internet monitoring experts at Kentik, NetBlocks, and Cloudflare began documenting the partial restoration of connectivity in Iran beginning in the early afternoon local time on Tuesday.

“We do see some traffic coming from Iran,” says Amir Rashidi, a cybersecurity expert with the internet freedom organization Miaan Group. “Some providers have come back online, but it is still too early to say exactly what will happen. After the January protests, some providers were also reconnected, but around 50 percent of the country’s traffic remained down.”

Doug Madory, the director of internet analysis at Kentik, says, “We’re not seeing much change for the mobile networks.” Instead, he says, some fixed-line providers have appeared to be restoring their services, with the Telecommunication Company of Iran’s fiber-optic service around Tehran showing the “biggest gain.”

At the start of January, the Iranian regime entirely shut down internet connectivity as the state killed thousands of protesters who took to the streets demanding improvements to economic conditions in the country. The government then entirely cut connectivity again at the end of February when the United States and Israel went to war in Iran—leaving millions of Iranians unable to contact their families, damaging the local economy, and prohibiting news and video footage about the war from getting into and out of the country. The limited reconnection of internet services on Tuesday comes as the US government continues to negotiate with Iran about a permanent end to the war.

Over the last decade, the Iranian regime has undertaken a massive project to control connectivity and censor content in the country while also building out a national intranet meant to essentially replace the global internet. This includes homegrown, surveillance-heavy tech such as search engines, messaging apps, and ride-hailing platforms. In practice, though, the regime’s digital mechanisms for control are often wielded as brute-force tools rather than precision instruments. It is unclear whether this is the result of technical limitations, political instability, or both.

Iran’s Supreme National Security Council seemingly ordered the current internet shutdown at the end of February as the war with the US started. A different group formed by current Iranian president Masoud Pezeshkian—known as the Special Headquarters for Organizing and Governing the Country’s Cyberspace—reportedly ordered connectivity restoration on Monday, though the move drew a legal challenge in Iran’s High Court. Nevertheless, the Iranian communications minister said that the reconnection would move forward per the president’s order, and that the process is underway to restore connectivity within 24 hours.

“What we are seeing now is an increase in traffic from Iran, but we need to wait and see the outcome of the power struggle,” Miaan Group’s Rashidi says. “Challenging the president’s order in court, given Iran’s political culture, was in a way a humiliation of Pezeshkian. So we should wait and see how this power struggle plays out.”

Rows of plain sheds sit on former cattle land near Rockwood in central Texas. During daylight hours they look ordinary enough to store tools or hay. Once the sun drops low, the roofs slide open together and expose long lines of telescopes mounted on solid bases. Owners scattered around the globe then take control from their laptops or phones. They point the instruments, collect images, and gather data on distant galaxies or nebulae without ever leaving home.

Light pollution has a horrible habit of obscuring weak stars and sky details from most people who live in or near a city, especially when streetlights and large buildings light up the night, making the vista a complete wash-out. On the other side, rural areas such as Rockwood remain dark enough to reveal some crisp details that city dwellers cannot perceive. It’s no surprise that amateur astronomers and astrophotographers want to bring their equipment to these ranches so their telescopes may sit permanently beneath those bright sky, eliminating the need to make the trek to some isolated location or even fight the weather on short notice.

Sale

Celestron StarSense Explorer LT 114AZ App-Enabled Telescope – 114mm Newtonian Reflector with Smartphone…

• SMARTPHONE-POWERED SKY TOUR: No experience needed! Just dock your phone, launch the StarSense Explorer app, and follow the on-screen arrows to locate…
• PATENTED STARSENSE TECHNOLOGY: Unlike other astronomy apps, StarSense Explorer uses sky recognition technology to turn your phone into a celestial…
• TONIGHT’S BEST TARGETS, INSTANTLY: The app generates a curated list of the top objects to see based on your time and location. See planets, bright…

Bray Falls, an amateur astronomer himself, founded Starfront Observatories around 18 months ago and transformed what had previously been open ground into one of the largest operations of its sort. Now he has eleven buildings with over 500 telescopes taking in the view from above, and high-speed fiber internet feeds those large image files directly back to the viewers. The telescopes are all positioned on clever robotic arms that track celestial objects as the Earth revolves, while weather sensors monitor the skies and automatically close the roofs when it rains or a storm is forming. It all works so effortlessly that consumers may plan their sessions from anywhere with an internet connection.

They have customers from all around, including Europe, Asia, North America, and the Middle East. One man who lives near Detroit sends his telescopes since his neighborhood cityscape ruins his attempts at stargazing at home. He live streams to his fans and enjoys the sharp results that a black sky provides, and he isn’t the only one. Monthly rates start around $99 for a basic setup, and if you own some high-end gear worth $10,000 or more, you pay for space and some continuous assistance, and the best thing is that the telescopes stay fixed on their piers, ensuring consistent focus and alignment night after night.

The facilities handle all of the day-to-day operations, such as opening the roofs at dusk, monitoring the conditions, and generally keeping things in order. The crew, sometimes known as telescope ranchers, ensure that the mounts, cameras, and software are all in working order so that remote control is possible. This position is particularly appealing to young astrophotographers who enjoy merging their technological skills with good old-fashioned astronomy. Word has spread, and videos and reports have appeared all over the place, including one by tech journalist Ashlee Vance, all of which show the roofs sliding back simultaneously to display the entire array.

Barnes & Noble CEO James Daunt recently sat down with NBC News, and he said something that has been percolating in my mind. When asked about AI-written books, Daunt said, “Yes, I have actually no problem selling any book, as long as it doesn’t masquerade or pretend to be something that it isn’t, and that it has an essential quality to it, and that the customer, the reader, wants it.”

On the surface, that sounds perfectly sensible. As long as readers can clearly see the label, they can make a choice. But if you take a moment to think about it, there are important questions that this approach leaves unanswered. 

Is “just label it” really good enough?

Barnes & Noble is one of the most powerful retailers in the publishing world. When the largest retail bookseller in the United States signals that AI-written books are welcome on its shelves, it sends a message to publishers, agents, and authors alike that this is a legitimate product category. 

Think about what a real book represents. A writer spent months, sometimes years, researching, writing, revising, creating a concoction, and then pouring it onto a page. Not only that, everything a writer puts on the page is colored with the lens they formed with their life experiences. That’s what makes books human, and why we sometimes read books covering the same topic from different writers. 

AI, on the other hand, takes everything it learned from the generation of human experience, strips the humanity, and serves the slop. Yes, the book might have the best grammar, the best plot structure, and even a good story. But will it have the human touch that makes a book special? I think not. At best, it can pretend, using the knowledge it stole from the great books written by human authors. 

The moment a major retailer shrugs and says AI books are fine as long as they’re labeled, it starts chipping away at the understanding that a book is a human endeavor. Also, who decides what constitutes an AI-written book and what the label looks like? Is it enough if the label is hidden obscurely on some page, where no one can find it unless they are looking for it? 

Even if they have a clear label, so what? Will you let a thief enter your home, as long as they wear a label saying they are one? It’s ridiculous. And make no mistake; any AI-written book, no matter how good it is, is a thief parading in costume, which has stolen the stories from human-written books, without consent. 

The human cost of letting AI books in our bookshops

Every bookshop has a limited space. If we allow AI books to enter our bookshops, it doesn’t create a space out of a vacuum. Every AI book taking a shelf space is replacing one written by a human. And without a proper system in place, which Barnes & Noble doesn’t seem to have, it would be hard for a reader to differentiate between a human and an AI-written book. 

Daunt even acknowledged that Barnes & Noble might already be selling AI-written books without knowing it. “We have 300,000 titles across all of our stores. Do we think that some of those may be AI? The chances are that they are, but we’re not really conscious of them,” he said in the NBC News interview. That is not the reassuring admission he thinks it is.

What you see is what you buy. If thousands of readers walk into the store and see AI books prominently placed, some of them are bound to pick one up. It will make money for some mega corporation or AI-bro who has started treating books as his new side business. That’s a sale that could have gone to an author who actually deserved it. 

I am not saying all human-written books are great. I have written some bad ones myself. But even if a book is bad or just not your taste, you know someone put real effort into it, so the hit on the purse doesn’t sting that much. 

Think how you will feel if your books were written by a prompt? Also, since AI can generate books at a much faster rate than we can write them, if we open the doors to these books, the market will be flooded. The e-book market is already filled with AI slop; we don’t want our bookstores to look the same. 

This is not happening in a vacuum

It would be one thing if Barnes & Noble were making this call in isolation. But this is part of a much larger and deeply troubling pattern.

Vox Media and The Atlantic both signed deals with OpenAI, allowing the company to train its models on their entire content archives. The New York Times signed its first AI content licensing agreement with Amazon. USA Today, Condé Nast, and Hearst have also signed multi-year licensing deals with Amazon. 

AI licensing deals are now becoming a big source of revenue for publishers. So publishers are getting paid, and that money is making these deals feel justified. As for the writers whose work is being used to train these models? Most of them are seeing nothing.

The pattern is clear here. First, media companies license their content to AI. Then AI uses that content to generate new content. Then retailers agree to sell that AI-generated content. This will repeat until all human writers are fired and all of us are left with a steaming pile of AI slop in our hands, wondering how we got here. 

Books are one of the last places where human creativity has not been fully colonized by AI. Opening that door, even with a label slapped on it, is a precedent the industry will struggle to walk back. Some doors should remain closed, no matter how lucrative the prize behind them seems to be.

Looking for the most recent Connections answers? Click here for today’s Connections hints, as well as our daily answers and hints for The New York Times Mini Crossword, Wordle, Connections: Sports Edition and Strands puzzles.

Wow, today’s NYT Connections puzzle has a really intriguing purple category. Hint: Look for words that use the same exact letters. Read on for clues and today’s Connections answers.

The Times has a Connections Bot, like the one for Wordle. Go there after you play to receive a numeric score and to have the program analyze your answers. Players who are registered with the Times Games section can now nerd out by following their progress, including the number of puzzles completed, win rate, number of times they nabbed a perfect score and their win streak.

Read more: Hints, Tips and Strategies to Help You Win at NYT Connections Every Time

Hints for today’s Connections groups

Here are four hints for the groupings in today’s Connections puzzle, ranked from the easiest yellow group to the tough (and sometimes bizarre) purple group.

Yellow group hint: You won.

Green group hint: Let’s get back to the topic.

Blue group hint: Ghostbusters is a classic.

Purple group hint: Mix up the letters.

Answers for today’s Connections groups

Yellow group: Championship awards.

Green group: Matter at hand.

Blue group: ’80s comedies.

Purple group: Anagrams.

Read more: Wordle Cheat Sheet: Here Are the Most Popular Letters Used in English Words

What are today’s Connections answers?

The yellow words in today’s Connections

The theme is championship awards. The four answers are cup, medal, pennant and ring.

The green words in today’s Connections

The theme is matter at hand. The four answers are concern, focus, point and subject.

The blue words in today’s Connections

The theme is ’80s comedies. The four answers are Airplane, Big, Clue and Twins.

The purple words in today’s Connections

The theme is anagrams. The four answers are enlist, listen, silent and tinsel.

Toughest Connections puzzles

We’ve made a note of some of the toughest Connections puzzles so far. Maybe they’ll help you see patterns in future puzzles.

#5: Included “things you can set,” such as mood, record, table and volleyball.

#4: Included “one in a dozen,” such as egg, juror, month and rose.

#3: Included “streets on screen,” such as Elm, Fear, Jump and Sesame.

#2: Included “power ___” such as nap, plant, Ranger and trip.

#1: Included “things that can run,” such as candidate, faucet, mascara and nose.

What will you miss most about Hacks? Ava (Hannah Einbinder) is furthering the queer agenda? Deborah’s (Jean Smart) sharp tongue? Or the fact that Meg Stalter won’t be as ever-present in our minds unless you’re chronically on Instagram?

Regardless, Hacks season 5 is going down in a blaze of glory. Last week, we saw Deborah on a ‘non-press tour press tour’ — and by that I mean that she appeared everywhere she could without explicitly promoting her Madison Square Gardens show.

Offbeat

Autonomous shuttle’s second passenger trip ends with rear-end collision and a tow truck

A new self-driving bus service in the Swedish city of Gothenburg got off to a rough start this week when one of its vehicles was hit by a tram on its second passenger-carrying trip.

The autonomous bus, running on route 169 between Gothenburg Central Station and Liseberg, opened to passengers on May 25. It was struck from behind shortly after setting off on its second run, resulting in damage to both vehicles and the bus enduring the ignominy of being towed away.

According to reports, the bus braked and was rear-ended by the tram. A spokesperson for Västtrafik, the bus operator, told The Register: “A small number of passengers were on board. No one was seriously injured, and that is the most important outcome. Both the bus and the tram sustained minor damage.”

The autonomous bus project began in 2024, and trials were scheduled to conclude by 2027 [PDF]. At present, a driver is still required, although the controls are not touched during normal operation.

However, someone else driving into it is a whole different challenge. On the rear of the Karsan bus was the warning “Keep your distance! The bus can brake sharply!” but that did not appear to deter the tram. As a rule of thumb, trams tend to have the right of way since they cannot swerve around a suddenly slowing vehicle.

Self-driving public transport remains a challenge. Elon Musk’s Cybercab is beginning to trundle onto the streets after a 2024 announcement, and Waymo’s taxis have been rolling around cities such as San Francisco for a few years now, although it yanked thousands of vehicles off the road over fears they might drive headlong into floods on high-speed roads. 

The UK’s first registered autonomous bus route was unveiled in 2023. However, operators reportedly decided to end the service in 2025 due to a lack of passengers. The service ran over a 14-mile journey in Scotland and also required drivers on board.

As for Gothenburg, the Västtrafik spokesperson told El Reg: “We are now conducting a thorough analysis of the incident together with relevant parties to better understand what happened as soon as possible.”

One of the goals of the project was to “gather and share knowledge about how autonomous mobility works in practice.” To which the answer, at least initially, appears to be: “Great! Until a tram enters the chat.” ®

• Apple’s AI-powered health coach has apparently been delayed
• It’s now not expected until “later in the iOS 27 update cycle”
• The news arrives in a new report from a reputable source

Apple’s WWDC event is just around the corner, which means we can expect a host of software announcements from the tech giant on June 8. One thing fans of the best Apple Watches might have been excited for is Apple’s long-rumored AI health coach (AI) — but a new report has just put a dampener on those expectations.

According to Bloomberg journalist Mark Gurman, Apple’s AI health coach — apparently dubbed ‘Project Mulberry’ inside Apple — might not be revealed at WWDC after all. That will come as a blow to anyone looking forward to this feature, but it’s not all doom and gloom.