Plus: Donald Trump has vowed to enact struct tariffs for China, Canada and Mexico
That is immediately’s version of The Obtain, our weekday publication that gives a each day dose of what is going on on on this planet of know-how.
The way in which we measure progress in AI is horrible
Each time a brand new AI mannequin is launched, it’s sometimes touted as acing its efficiency in opposition to a collection of benchmarks. OpenAI’s GPT-4o, for instance, was launched in Could with a compilation of outcomes that confirmed its efficiency topping each different AI firm’s newest mannequin in a number of exams.
The issue is that these benchmarks are poorly designed, the outcomes laborious to copy, and the metrics they use are incessantly arbitrary, in response to new analysis. That issues as a result of AI fashions’ scores in opposition to these benchmarks decide the extent of scrutiny they obtain.
AI firms incessantly cite benchmarks as testomony to a brand new mannequin’s success, and people benchmarks already type a part of some governments’ plans for regulating AI. However proper now, they may not be adequate to make use of that method—and researchers have some concepts for the way they need to be improved.
—Scott J Mulligan
We have to begin wrestling with the ethics of AI brokers
Generative AI fashions have change into remarkably good at conversing with us, and creating pictures, movies, and music for us, however they’re not all that good at doing issues for us.
AI brokers promise to vary that. Final week researchers revealed a brand new paper explaining how they skilled simulation brokers to copy 1,000 folks’s personalities with gorgeous accuracy.
AI fashions that mimic you may exit and act in your behalf within the close to future. If such instruments change into low-cost and simple to construct, it can elevate plenty of new moral considerations, however two specifically stand out. Learn the total story.
—James O’Donnell
This story is from The Algorithm, our weekly AI publication. Join to obtain it in your inbox each Monday.
The must-reads
I’ve combed the web to search out you immediately’s most enjoyable/vital/scary/fascinating tales about know-how.
1 Donald Trump has pledged particular tariffs for China, Canada and Mexico
He says it’s to stop drug trafficking and unlawful migration into the US. (WP $)
+ The tariffs are dangerous information for Chinese language EV agency BYD’s deliberate manufacturing unit in Mexico. (WSJ $)
+ How Trump’s tariffs may drive up the price of batteries, EVs, and extra. (MIT Know-how Evaluation)
2 Maternal medical doctors are leaving Texas
Abortion restrictions make it a lot more durable to manage miscarriage care. (New Yorker $)
+ Porsha Ngumezi is the third lady identified to have died below the state’s ban. (ProPublica)
3 Bluesky has been accused of breaching EU knowledge guidelines
It’s did not declare what number of EU customers it has and the place it’s legally primarily based. (FT $)
+ Bluesky says it’s working to adjust to the disclosure guidelines. (The Info $)
4 How Amazon plans to tackle Nvidia
Its engineers are racing to get its AI chips operating reliably in knowledge facilities by the tip of the 12 months. (Bloomberg $)
+ What’s subsequent in chips. (MIT Know-how Evaluation)
5 Neuralink will check whether or not its mind implant can management a robotic arm
If it could, it’ll be the primary wi-fi brain-computer interface to take action. (Wired $)
+ Meet the opposite firms creating brain-computer interfaces. (MIT Know-how Evaluation)
6 Your Pokémon Go knowledge might be purchased by militaries and governments
Dad or mum firm Niantic hasn’t dominated it out. (404 Media)
7 Inside Google’s little-known nuclear vitality analysis group
It’s quietly been searching for to additional our understanding of nuclear vitality for years. (IEEE Spectrum)
+ Why the lifetime of nuclear vegetation is getting longer. (MIT Know-how Evaluation)
8 US farms desperately want recent water
New desalination initiatives may assist make plentiful saltwater extra plant-friendly. (Knowable Journal)
+ How we drained California dry. (MIT Know-how Evaluation)
9 Nvidia’s new AI mannequin creates fully new sounds 🎷
Together with a screaming saxophone and an indignant cello. (Ars Technica)
+ These unimaginable devices may change the way forward for music. (MIT Know-how Evaluation)
10 We could lastly know what causes mysterious radio flashes from house
Asteroids and comets bashing into neutron stars might be behind them. (New Scientist $)
Quote of the day
“Did we modify Huge Tech? My reply is not any.”
—Tommaso Valletti, an economist who labored below the European Union’s antitrust regulator Margrethe Vestager, displays on her legacy as she prepares to step all the way down to the New York Occasions.
The massive story
The best way to repair the web
October 2023
We’re in a really unusual second for the web. Everyone knows it’s damaged. However there’s a way that issues are about to vary. The stranglehold that the large social platforms have had on us for the final decade is weakening.
There’s a type of frequent knowledge that the web is irredeemably dangerous. That social platforms, hungry to revenue off your knowledge, opened a Pandora’s field that can not be closed.
However the web has additionally offered a haven for marginalized teams and a spot for help. It affords data at occasions of disaster. It will probably join you with long-lost pals. It will probably make you giggle.
The web is price combating for as a result of regardless of all of the distress, there’s nonetheless a lot good to be discovered there. And but, fixing on-line discourse is the definition of a tough drawback. However don’t fear. I’ve an thought. Learn the total story.
—Katie Notopoulos
We are able to nonetheless have good issues
A spot for consolation, enjoyable and distraction to brighten up your day. (Bought any concepts? Drop me a line or tweet ’em at me.)
+ America is tremendous into republishing traditional literature today.
+ I’m satisfied there’s nothing extra modern and daring than a hungry cat (thanks Dorothy!)
+ Gen Z famously likes to mock the way in which millennials costume, however evidently: we’ve had the final giggle.
+ How music influences math, imagine it or not.