Get your iPhone or iPad to automatically download the next two episodes of the TV show you’re currently watching and delete them later.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Montana’s plan to become an experimental medical hub just pushed forward 

As of this week in Montana, biotech companies whose drugs have been through preliminary testing—sometimes in as few as 10 healthy people—can pay $12,500 to apply to a newly established review board for approval.

Once its treatment is rubber-stamped, the company can sell it via experimental treatment clinics, the first of which is likely to be up and running around the end of this year.   

Montana’s latest right-to-try legislation is unique. Access to drugs is theoretically available to anyone who gives informed consent and can pay. For some, especially people in the longevity community, that’s a hopeful and exciting prospect. But to others, it’s unethical and dangerous. 

Read our story to learn about where this may all be headed. 

—Jessica Hamzelou

Montana’s new “right to try” law can’t come soon enough for some 

Kris DeVault is desperate.

His son, Brody, born in March 2023, has something called creatine transporter deficiency—a rare condition in which the brain and muscles lack the energy they need to develop.

There are no cures for Brody’s condition. But DeVault has learned of a company developing a drug that might help. That drug is still in the early stages of development and has only been tested in animals and a small number of healthy adults. Doctors can’t prescribe it.

DeVault knows the drug might not work. But he’s doing all he can to access it regardless. Read our story about DeVault’s efforts. 

—Jessica Hamzelou

This story is from The Checkup, our weekly biotech newsletter. Sign up to receive it in your inbox every Thursday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 Anthropic says its models hacked external organisations during testing
It was prompted to conduct a review after similar issues at OpenAI. (TechCrunch) 
+ Inside OpenAI’s hack of Hugging Face. (New Yorker $)
+ OpenAI called the Hugging Face attack unprecedented. But we’ve been here before. (MIT Technology Review)

2 Europe is bracing itself for an even more fiery future
Countries like France and Spain can expect regimes akin to those in California. (Nature $)
+ Even the UK is experiencing fires now, too. (Guardian)
+ El Niño is partly to blame for this year’s grueling heatwaves. (Wired $)
+ District cooling could help beat extreme heat in cities. (New Scientist $)

3 Drone warfare is making the skies more dangerous
A fatal plane accident in the US this May, caused by a military GPS jamming exercise, may signal what’s to come. (Wired$)
+ Zelensky asked Trump to secure Musk’s permission for the use of Starlink to guide drone strikes inside Russia this week. (The Atlantic $)

4 As Big Tech’s AI spending grows, so do the jitters
Amazon, Google, Meta and Microsoft are set to invest $1.5 trillion into AI infrastructure. But who will pay for it? (NYT $)
+ Are investors really getting cold feet about the AI boom? (FT $)
+ An AI-focused hedge fund just imploded. (CNBC)

5 How China has changed the sovereign AI playbook
Kimi 3 from Moonshot makes forking out for expensive US models seem foolish. (Rest of World)
+ The US is winning on the cutting edge. But China dominates cheap models. (CNBC)
+ Could AI end up too cheap to control? (Vox $)

6 Google’s Gemini can now control a humanoid robot
It was previously restricted to upper body movements. Now, it can manage the entire range of motion. (Bloomberg $)
+ These gig workers are training humanoid robots at home. (MIT Technology Review)

7 Amazon says it’s found “catastrophically expensive” AI cost overruns
It’s far from alone—and that’s why the “tokenmaxxing” trend disappeared so quickly. (FT $)

8 How a remote indigenous community built a super fast fiber network
It was a complex, expensive endeavor. But it now outperforms Starlink. (IEEE Spectrum) 
+ Stratospheric internet could finally start taking off this year. (MIT Technology Review)

9 Meet Wikipedia’s most diligent editor
Steven Pruitt is unremarkable in some ways—yet truly amazing in others. (New Yorker $)

10 You can now report AI slop with a button on LinkedIn
The people have spoken. And they are tired of reading bland nonsense. (404 Media)

Quote of the day

“It’s a shock, thinking that just an hour earlier I’d been in my room and everything… and now, there’s nothing left.”

—18-year-old Raphael Fohanno tells Reuters his reaction to wildfires destroying his parents’ house in the French town of Biscarrosse.

One More Thing

Congress used to evaluate emerging technologies. Let’s do it again.

The US Office of Technology Assessment, an independent office created by Congress in the early 1970s, produced some 750 reports during its 23-year history, assessing technologies as varied as electronic surveillance, genetic engineering, and remote sensing from outer space. 

The office functioned like a debunking arm. It sussed out the snake oil. Its reports saw through the alluring gleam of overhyped technologies. 

Since its unceremonious defunding in 1995, perennial calls have gone out: Rouse the office from the dead! 

Read our story about why, with advances in robotics, big data, and AI systems, these calls have taken on a new level of urgency. 

—Peter Andrey Smith

We can still have nice things

A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ Fencing on roller skates is the self-defense you didn’t know you needed.
+  Meet Pumpkin, a chubby cat that’s in no mood to lose weight.
+ Discover what the animals near you are probably doing right now at Nature This Week.
+ Woodworker extraordinaire Matt Thompson has installed a model railroad that runs along his fence. 

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

A fundamental flaw leaves LLMs strikingly vulnerable to attack 

It is impossible to make large language models fully secure against hacks because of a fundamental flaw in how they work, a team of researchers argue in a paper presented at a top AI conference earlier this month.  

The flaw concerns how LLMs identify who or what is giving them instructions. By taking advantage of it, the researchers were able to make popular LLMs spit out information they had been trained not to provide, such as how to synthesize cocaine and how to sabotage a commercial aircraft’s navigation system.  

Read our story about the flaw the researchers found, its implications—and why it may never be fixed.

—Will Douglas Heaven

How an overlooked geothermal plant got a second chance 

In June 2024, a small company called Zanskar purchased a geothermal power plant in New Mexico that was failing fast. The water coming from the underground reservoir was getting colder by the day, making the plant uneconomical to run.

Now, two years later, that plant is running at full capacity again, thanks to a new well. With the help of advanced modeling and modern drilling technology, the company was able to identify a better well site, drill down thousands of feet, and revive the entire operation.

As the world looks for more sources of emissions-free electricity that are available 24-7, Lightning Dock shows there’s still hidden potential deep beneath our feet. Read the full story.

—Casey Crownhart

This story is from The Spark, our weekly climate tech newsletter. Sign up to receive it in your inbox every Wednesday.

MIT Technology Review Narrated: South Korea’s hottest new bachelors are chip workers 

Baek, a 35-year-old manager at the South Korean semiconductor titan SK Hynix, was enrolled in a matchmaking company a year ago. In a move typical of anxious South Korean parents, his mother signed him up, hoping to find a good wife for her son.

Lately, says Baek, he and his coworkers are having better luck finding dates—perhaps because of the dazzling bonuses they just got. Flush with eye-popping profits from the AI chip boom, SK Hynix agreed to pay 10% of operating profits to employees, which translates to an extra $476,000 per employee this year. 

Read our story about how chip workers like Baek have become the most sought-after bachelors and bachelorettes in South Korea. 

—Michelle Kim

This is our latest story to be turned into an MIT Technology Review Narrated podcast, which we publish each week on Spotify and Apple Podcasts. Just navigate to MIT Technology Review Narrated on either platform, and follow us to get all our new content as it’s released.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 The US government has just banned Roombas
It’s not just Chinese humanoids—robot vacuum cleaners are included in a new FCC ban too. (The Verge $)
+ Who wins and who loses from the ban on foreign robots? (Ars Technica)
+ China says it will retaliate. (CNBC)

2 The ongoing fires in Europe are breaking modern records
And there’s likely even more destruction coming. (NYT $)
+ Thousands of people have just been evacuated on the Greek island of Crete. (BBC)

3 Google DeepMind has dismantled its Nobel-winning AlphaFold team
It reflects a broader industry move away from specialist tools towards more general, AI-powered agents for science. (FT$)
+ Google I/O showed how the path for AI-driven science is shifting. (MIT Technology Review)

4 Data centers are easy to build. To run? Not so much
For that, you need time and money to invest in grid infrastructure, like new transmission lines. (404 Media $)
+ The power line that could reshape New York’s grid is hitting snags. (MIT Technology Review)
+ AI companies are hiring thousands of electricians and carpenters to get data centers up and running. (NYT $)

5 DoorDash plans to launch a drone delivery program
It’ll be a while before we get to make use of it, though. (TechCrunch)
+ The US may be heading toward a drone-filled future. (MIT Technology Review)

6 AI is accelerating global digital inequality
Money, infrastructure and talent are pooling in a relatively small number of places. (IEEE Spectrum)

7 Quantum computers promise mathematical superpowers 
And it feels like we’re inching closer to a commercial machine. (The Economist $)
+ PsiQuantum has a plan to make a massive quantum computer out of light. (MIT Technology Review)

8 Anxious Chinese students are using AI for university admissions
It’s common to pay private coaches to help navigate this high-stakes decision, but AI companies now offer the service for free. (Rest of World)
+ How DeepSeek became a fortune teller for China’s youth. (MIT Technology Review)

9 Boomers keep giving their grandkids AI-generated slop books
And it’s driving millennial parents mad. (Wired $)

10 Minecraft is helping children to redesign their cities
It just goes to show how creativity can still flourish, even amid war. (NYT $)

Quote of the day

“I want to see the blood, sweat, and tears that went into it.”

—Ray Slater Berry, founder of marketing agency dslx, tells Wired why he’s drawn to text written by humans. 

One More Thing

Europe’s drone-filled vision for the future of war 

Europe has started testing an invisible automated intelligence network, known as a “digital targeting web,” conceived under the name Project ASGARD. Its purpose is to connect everything that looks for targets—“sensors,” in military lingo—and everything that fires on them (“shooters”) to a single, shared wireless electronic brain.  

Eighty years after total war last transformed the continent, the system signals a brutal new calculus of European defense. “The Russians are knocking on the door,” says Sven Weizenegger, the head of the German military’s Cyber Innovation Hub. 

Read our story about how strategists and policymakers are counting on increasingly automated battlefield gadgetry to keep them from bursting through. 

—Arthur Holland Michel

We can still have nice things

A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ Love this story about the studio accident that birthed the 1980s ‘gated reverb’ drum sound.
+ Wewere.online turns solitary web browsing into a multiplayer experience by rendering live cursor trails of other visitors.
+ Imagine how relaxed you’d get staying in this coastal cabin in Tasmania. 
+ Why do cats always land on their feet? It’s thanks to their amazingly flexible spines.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

Samsung’s chip workers are jumping ship to rival SK Hynix  

Lee, an engineer at Samsung’s semiconductor division, used to work late. But lately, he’s been clocking out on time and heading straight home to work on his job application for the chipmaker’s South Korean rival SK Hynix.  

His colleagues are doing the same. They feel demoralized by the $476,000 bonus that SK Hynix is set to pay its employees, flush with record profits from making the high-bandwidth memory (HBM) chips that power Nvidia’s AI accelerators. The figure dwarfs what chip workers at Samsung are set to receive and is sparking an exodus.

Read our story about why this fierce talent war could help determine who dominates the next generation of AI chips. 

—Michelle Kim

The AI Hype Index: Unsexy AI  

Separating AI reality from hyped-up fiction isn’t always easy. That’s why we’ve created the AI Hype Index—a simple, at-a-glance summary of what’s shaping the industry right now. 

The latest edition includes Meta’s creepy glasses, doom-mongering about AI and jobs, and (to some, oddly sexy?) robotic hands. See where it all landed on this month’s index.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 OpenAI’s rogue agent compromised another customer
An unnamed client of a company that provides AI infrastructure, called Modal Labs. (Reuters $)
+ OpenAI called the Hugging Face attack unprecedented. But we’ve been here before. (MIT Technology Review)

2 Hugging Face has a nonconsensual deepfakes problem
Researchers tested nine of the top image editing models it hosts, and found seven will ‘nudify’ images of women. (Wired $)
+ The shock of seeing your body used in deepfake porn. (MIT Technology Review)

3 Investors are only getting more anxious about AI’s earnings potential
The numbers don’t add up, and maybe they never will. (The Verge $)
+ A global tech stock sell-off intensified yesterday. (FT $)
+ Ratings agency Fitch says an AI market correction is now one of the biggest risks facing the global economy. (The Next Web)
+ New real-world data from Google shows most tasks at most jobs are unaffected by AI. (Ars Technica)

4 Moonshot wants advanced Nvidia chips to train its next model
The Chinese AI firm already got access to some of them to train its latest release, called Kimi 3. (The Information $)
+ Taiwan detained an Nvidia employee as part of a probe into the alleged smuggling of AI chips into China.(Bloomberg $)
+ The Trump Administration has banned Chinese humanoid robots. (Reuters $)

5 Minnesota has been blocked from banning prediction market platforms
It’s the latest skirmish in a fight between states and the Trump Administration over this issue. (AP)
+ Here’s how prediction markets incentivize weather data sabotage. (MIT Technology Review)

6 Apple is launching a device leasing program
As the global memory shortage worsens, the idea is that this will make your iPhone feel more affordable. (Quartz $)
+ Apple just topped a $5 trillion valuation for the first time. (FT $)

7 Some Substackers are complaining about its new AI detection tool
Presumably the ones who’d prefer to secretly continue using AI to write. (404 Media $)

8 Ebay had to pay nearly $50 million to settle a harassment suit
Their employees threatened and sent bizarre deliveries—including cockroaches—for a pair of bloggers for years. (Bloomberg $)

9 A New York school just rowed back on plans to deploy a robot teacher
Sanity prevails, for once. (NPR)

10 This startup wants to build a better microwave beep
I’d like to request a sax solo for mine, please. (Gizmodo)

Quote of the day

“The government hopes to set a precedent that no one has the right to privacy.” 

—Samuel Tunick, an Atlanta resident, tells 404 Media why he thinks he’s facing charges after wiping his phone during a customs search.

One More Thing

AI coding is now everywhere. But not everyone is convinced. 

Depending who you ask, AI-powered coding is either giving software developers an unprecedented productivity boost or churning out masses of poorly designed code that saps their attention and sets software projects up for serious long term-maintenance problems.  

The problem is right now, it’s not easy to know which is true. Coding has been touted as the killer app for large language models, but after speaking to more than 30 developers, technology executives, analysts, and researchers, MIT Technology Review found that the picture is not as straightforward as it might seem.

Here’s what we found out. 

—Edd Gent

We can still have nice things

A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ Check out Moscow’s stunning subway stations.
+ This website lets you create your own little ‘fridge magnet’ style poems. 
+ Look how beautiful sand can be when it’s put under the microscope. Who knew?!
+ Can’t get enough of this weird pastel dreamcore aesthetic.

This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology.

OpenAI called the Hugging Face attack unprecedented. But we’ve been here before. 

—Will Douglas Heaven, senior AI editor

Reading OpenAI’s account last week of how some of its models broke their containment and hacked into the computer systems of Hugging Face, another AI company, was the first time I got genuine chills about what large language models are now able to do. But this is a case of human hubris, not rogue AI.

I am not an alarmist. In fact, I have been pushing back against AI scare stories for years. Even so, this incident crossed a line. I think it’s the clearest illustration yet of how the people building and testing this technology do not fully understand what they’re doing. 

Here’s why OpenAI could—and should—have seen this coming.

This story is from The Algorithm, our weekly AI newsletter. Sign up to receive it in your inbox every Monday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 There’s a growing global AI stock sell-off underway 
Chip and memory stocks are bearing the brunt so far. (FT $)
+ It was partly sparked by a report that a Chinese company has started making a key piece of chip equipment for the first time. (The Information $)
+ Like their US rivals, Chinese AI firms are struggling to find a path to profitability. (NYT $)
+ What even is the AI bubble? (MIT Technology Review)

2 Some people’s chats with Claude were open to anyone online
OpenAI had a near-identical issue with ChatGPT last year. (BBC)
+ Is a secure AI assistant even possible? (MIT Technology Review)

3 France just witnessed its first ever “fire cloud”
It’s a scary sign of the intensity of the blazes burning there. (Wired $)

4 Meta is screwing up its smart glasses rollout
Privacy issues keep cropping up, and its response is invariably too little, and too late. (The Verge $)
+ They even have a nickname: “pervert glasses.” (Vox $)

5 Efforts are underway to measure Spotify’s AI slop problem
People are desperate for it to start labelling AI-generated music. (404 Media)
+ We’re in a weird era regarding AI’s role in literature. (The Atlantic $)

6 People are renting out their faces to AI in China
Often for microdramas, which are big business and increasingly rely on AI. (Rest of World)
+ How Chinese short dramas became AI content machines. (MIT Technology Review)

7 Microsoft is having a torrid year
Rivals building better AI tools are threatening its business on multiple fronts. (Business Insider $)

8 How bots took over the internet
They now outnumber humans, in terms of overall web traffic. (New Yorker $)

9 Robotaxis are hitting London’s streets this summer 
Just for testing for now, though there are plans to launch rides to the public next year. (Engadget)

10 Why video games are obsessed with Backrooms
These spaces, which players aren’t supposed to access, can provoke a strange, liminal sort of discomfort. (Guardian)

Quote of the day

“It’s going to be a wild ride.”

—Christine Peterson, co-founder of the grantmaking group Foresight Institute, tells Wired charities anticipate a huge philanthropy windfall from Anthropic and OpenAI IPOs.

One More Thing

On the ground in Ukraine’s largest Starlink repair shop 

Starlink is critical to Ukraine’s ability to continue in the fight against Russia, but Donald Trump’s fickle foreign policy and reports suggesting Elon Musk might remove Ukraine’s access to the services have cast the technology’s future in the country into doubt. 

For now Starlink access largely comes down to the unofficial community of users and engineers, including the expert “Dr. Starlink”—famous for his creative ways of customizing the systems—who have kept Ukraine in the fight, both on and off the front line. 

He gave MIT Technology Review exclusive access to his unofficial Starlink repair workshop in the city of Lviv. Read the full story.

—Charlie Metcalfe

We can still have nice things

A place for comfort, fun, and distraction to brighten up your day. (Got any ideas? Drop me a line.)

+ As these retirees have discovered, dancing is incredibly good for you.
+ Do you feel lucky, punk? Life feels a lot easier if you do!
+ 100% would visit this cat every day if I lived in Istanbul. 
+ Watch this to understand how the movie Sinners had such impressive visual effects.