The Download: computing’s next leap and permafrost’s climate threat

This is today’s edition of The Download our weekday newsletter that offers a day-to-day dosage of what’s going on worldwide of innovation.

How far can we press computing?

On a December night in 1959, the physicist Richard Feynman provided a lecture called “There’s Plenty of Room at the Bottom.” There is no physical law stopping us from composing the whole Encyclopædia Britannica on the head of a pin, he argued. We merely had actually not developed the tools.

He was right, and for years, calculating took us into that area nearly free of charge, as transistors consistently cut in half in size. That simple descent has actually now stalled. Today, the market is significantly turning to strength, developing ever-bigger information centers to improve efficiency.

The laws of physics leave plenty of space for enhancement.

Here’s how scientists are checking out methods to open the next leap in computing.

— Volkan Gurses

This story is from the most recent edition of our print publication: The Impossible problem Subscribe now to read it completely

It’s likewise part of MIT Technology Review Explainsour series checking out the concepts forming innovation. Find out more from the series

3 things you must understand about permafrost

Just how much do you learn about permafrost? Possibly, like me, you’ve know that this layer of frozen soil is defrosting, that it’s crucial in the worldwide puzzle of environment modification … which’s about the degree of it.

The excellent news for both people is that my associate James Temple simply helicoptered into Alaska’s forests to report on the fires threatening permafrost and their complex prospective environment impacts.

Here are 3 things I eliminated from his reporting.

— Casey Crownhart

This story is from The Spark, our weekly environment tech newsletter. Register to get it in your inbox every Wednesday.

Roundtables: a discussion with the developer of AI-designed infections

Can AI create brand-new life kinds? In 2025, Stanford University PhD trainee Samuel King used an initial response by utilizing a generative AI design to propose hereditary plans for tiny infections. It isn’t yet an example of AI-generated life, however that might be next.

In our most current Roundtables discussion, MIT Technology Review senior AI press reporter James O’Donnell consulted with King about his work, being called among our Innovators Under 35, and brand-new methods of seeing biology. Customers can now see a complete recording of the occasion.

Wish to sign up with the next discussion? Sign Up For MIT Technology Review for special access to all our Roundtables.

The must-reads

I’ve combed the web to discover you today’s most fun/important/scary/ remarkable stories about innovation.

1 AI has actually discovered a decade-old defect threatening much of the web
It discovered defects in PHP, the code behind many sites. (Reuters $)
+ Governments are advising site owners to spot PHP or take websites offline. (BBC
+ Exploit efforts leapt from thousands to millions after disclosure. (Axios
+ Now is a great time for doing criminal offense. (MIT Technology Review)

2 Hackers utilized Claude and Grok to automate attacks on over 400 banks
The AI representatives targeted banks in Latin America. (Information $)
+ Unvetted AI representatives are likewise accessing United States checking account.(Axios
+ Barclays states AI is enabling the bank to cut 5,000 tasks. (Guardian
+ AI is making online criminal activities easier. (MIT Technology Review)

3 The Pentagon has actually cleared the DHS to check a heat ray weapon
It might be utilized for crowd control at the southern United States border. (CNN
+ It utilizes unnoticeable energy beams to produce a burning feeling. (Axios

4 A start-up states AI-designed infections have actually treated their very first client
The speculative treatment targets drug-resistant E. coli. (NYT $)
+ But the business has actually not released peer-reviewed outcomes. (Futurism
+ AI-designed infections are currently eliminating germs. (MIT Technology Review)

5 Japan wants to return the very first samples from Mars’s moons
Its spacecraft is now on its method to Phobos. (Guardian
+ It intends to expose whether Mars was when habitable. (Economist $)

6 Meta deals with a $25 billion charge over the Cambridge Analytica scandal
The business states it will appeal the New Mexico judgment. (CNBC
+ A Meta share offer has actually persuaded Mark Zuckerberg to remain in California. (WSJ $)
+ Until the business remade its stock structure to assist him pay it. (NYT $)

7 A brand-new antimatter experiment might unlock to brand-new physics
Researchers want to identify distinctions in between matter and antimatter. (Nature

8 China’s Unitree is pitching its most current humanoid as a robotic bodyguard
A marketing clip reveals the H2-X deactivating assailants. (SCMP

9 Just Eat has actually apologised for unusual AI-generated menu images
Users were not impressed by the bacon in soda and cheese sushi. (BBC

10 Airbnb hosts rage they can no longer make visitors do tasks
Hosts are threatening to strike, however visitors are commemorating. (Wired $)

Quote of the day

“Now they’ve provided everyone a factor not to utilize Starlink: that Starlink isn’t trusted for nationwide security for any person however Trump. Possibly he ‘d turn it off on the United States if he didn’t like policy and he didn’t like the president.”

— SpaceX financier Ross Gerber informs CNN that Elon Musk’s choice to limit Ukraine’s Starlink gain access to has actually made the service a liability for every single other federal government that counts on it.

Another thing

Online harassment is entering its AI period

Scott Shambaugh didn’t reconsider when he rejected an AI representative’s demand to add to matplotlib, a software application library he assists handle. Things got odd.

In the middle of the night, Shambaugh opened his e-mail to find the representative had actually struck back with a post. Entitled “Gatekeeping in Open Source: The Scott Shambaugh Story,” the post implicated him of declining the code out of a worry of being supplanted by AI. “He attempted to secure his little fiefdom,” the representative composed. “It’s insecurity, plain and easy.”

Shambaugh isn’t alone in dealing with misbehaving representatives– and they’re not likely to stop at harassment. Check out the complete story.