The Download: how NYC tackles robust issues, and China’s AI requirements


This is immediately’s version of The Download, our weekday publication that gives a every day dose of what’s happening on the earth of know-how.

Why New York City is embracing low-tech options to exhausting issues

It’s a actuality of politics that’s typically neglected: as soon as a legislation is handed, it must evolve from an thought right into a plan with a price range and a workers, and from there it wants to truly attain the lives of thousands and thousands of individuals. Moving from coverage to implementation has all the time been a tough a part of governing. Today it’s simple to imagine know-how could make it simpler. But is that truly true?

New York City is one thing of a check lab for methods to confront some huge issues that plague the trendy state. Akin to a rustic within the price range and bureaucratic complexity of its authorities, it’s, and has been, coping with the important thing query of how one can make authorities work for individuals immediately. But it’s discovering that generally the answer to doing huge issues additionally entails doing plenty of small issues, generally with the bottom tech potential. Read the complete story.

—Tate Ryan-Mosley

Tate’s story is from the following journal version of MIT Technology Review, set to go dwell on October 25. It’s all about society’s hardest issues, and the way we should always deal with them. If you don’t subscribe already, join now to get a duplicate when it lands.

China’s plan to guage the security of generative AI

Ever for the reason that Chinese authorities handed a legislation on generative AI again in July, it’s been unclear how precisely China’s censorship machine would adapt for the AI period. 

Last week we lastly received some readability. On October 11, the Chinese authorities launched a draft doc that proposed detailed guidelines for how one can decide whether or not a generative AI mannequin is problematic. 

The requirements doc is very detailed: it units clear standards for when an information supply ought to be banned from coaching generative AI. It additionally clarifies what corporations ought to think about a “safety risk” in AI fashions. But, it’s essential to recollect, these requirements aren’t legal guidelines. Read the complete story.

—Zeyi Yang

Zeyi’s story is from China Report, our weekly publication masking tech and energy in China. Sign up to obtain it in your inbox each Tuesday.

This microbe-filled tablet might monitor irritation within the intestine

The information: A blueberry-size tablet that you simply swallow might let docs measure indicators of inflammatory bowel illness (IBD) within the intestine, serving to spot it earlier and measure its development in actual time.

How it really works: The tablet comprises E. coli micro organism which have been genetically engineered to detect molecules which might be produced in extra within the guts of individuals with IBD. A gene cluster was added to the micro organism that causes them to emit mild when they’re uncovered to those inflammatory molecules. The tablet additionally comprises photodiodes, which may detect this mild, and a radio chip that may transmit the sign exterior the physique.

Next steps: Following profitable trials in pigs, the tablet now needs to be validated in people to reveal that the sensors are capable of detect the biomarkers within the concentrations that may be indicative of illness. Read the complete story.

—Soumya Sagar

This is a subscriber-only story. If you haven’t already, join now to learn it in full.

2023 Climate Tech Companies to Watch: ReNew and its bold clear power tasks

ReNew is taking part in a essential function in India’s power transition by scaling up the nation’s renewable power capability, constructing power storage infrastructure, and—it hopes—ultimately producing inexperienced hydrogen.  Read extra about it, and skim the remainder of the checklist of Climate Tech Companies to Watch.

The must-reads

I’ve combed the web to search out you immediately’s most enjoyable/essential/scary/fascinating tales about know-how.

1 Hamas is spreading propaganda by unmoderated social platforms
Telegram channels are rife with disturbing movies and pictures of the atrocities in Gaza. (WP $)
+ Banning Hamas from social media hasn’t prevented it from sowing concern. (NYT $)
+ Big Tech has abdicated its accountability for the information. (New Yorker $)

2 AI factories are coming
Thanks to a brand new deal struck between Nvidia and Chinese cellphone maker Foxconn. (FT $)

3 China’s spies are hellbent on stealing from Silicon Valley
Today’s tech secrets and techniques are extra compelling than political intel, based on the FBI. (NYT $)
+ How China constructed a one-of-a-kind cyber-espionage behemoth to final. (MIT Technology Review)

4 X is charging some customers $1 a 12 months to submit
In one more try and differentiate bots from actual human customers. (The Verge)

5  OpenAI is engaged on a brand new AI-generated pictures device
Whether the device actually is 99% correct, because it claims, stays to be seen. (Bloomberg $)
+ GPT-4 remains to be fairly simple to trick. (The Verge)
+ AI-text detection instruments are very easy to idiot. (MIT Technology Review)

6 Scientists are bugging the rainforest to guard it
Eavesdropping on creatures is a non-invasive approach of conserving tabs on them. (Wired $)
+ Urban bushes may very well be falling sufferer to unhealthy microbiomes. (New Scientist $)

7 New York City’s mayor has been making audio deepfakes of himself
His workplace has been making eerie robocalls of him talking Spanish, Mandarin, and Yiddish, with out the disclaimer it’s not really him. (Motherboard)
+ AI voice actors sound extra human than ever—they usually’re prepared to rent. (MIT Technology Review)

8 Even artists’ tech isn’t immune from turning into out of date
And changing long-discontinued {hardware} is turning into more and more difficult. (NYT $)

9 Check out this superb image of Jupiter’s moon Io
It’s the most effective picture we’ve captured in many years. (New Scientist $)
+ What’s subsequent for the moon. (MIT Technology Review)

10 To construct a greater web, we should deal with the place we went mistaken
We shouldn’t depart it as much as the loudest, most polarizing voices. (The Atlantic $)
+ How to repair the web. (MIT Technology Review)

Quote of the day

“If you’re able to build a rocket faster than the government can regulate it… that needs to be addressed.”

—Tim Hughes, SpaceX’s senior vice chairman, criticizes the US authorities’s sluggish regulatory processes, the Washington Post reviews.

The huge story

How AI helps historians higher perceive our previous

April 2023

Historians have began utilizing machine studying to look at historic paperwork, together with astronomical tables like these produced in Venice and different early trendy cities.

Proponents declare that the appliance of contemporary laptop science to the previous helps draw connections throughout a broader swath of the historic report than would in any other case be potential, correcting distortions that come from analyzing historical past one doc at a time.

But it introduces distortions of its personal, together with the danger that machine studying will slip bias or outright falsifications into the historic report. Read the complete story.

—Moira Donovan

We can nonetheless have good issues

A spot for consolation, enjoyable and distraction in these bizarre instances. (Got any concepts? Drop me a line or tweet ’em at me.)

+ Go on then—a listing of all of the nice albums this 12 months you could have missed (to date)
+ What may very well be higher than tickling a pleasant fox? 🦊

+ This assortment of pictures of London by the ages is splendidly nostalgic.
+ Drum machines won’t ever not be cool.
+ Congratulations to Grazer, the fish-loving winner of this 12 months’s hotly-contested Fat Bear contest!


Please enter your comment!
Please enter your name here