This is the current issue of The Download, our weekday newsletter that delivers a daily glimpse into the happenings in technology.
China has mastered EV sales. Now it must deal with their batteries.
Over the last ten years, China has experienced an electric vehicle (EV) surge, aided by governmental backing. Acquiring an electric vehicle has transformed from an uncommon choice to a commonplace one; by the end of 2025, almost 60% of newly sold cars were electric or plug-in hybrids.
However, as the batteries in China’s initial series of EVs approach the end of their life cycle, early buyers are beginning to retire their vehicles, putting the nation under pressure to determine the fate of these deteriorating components.
This situation is stressing China’s developing battery recycling sector and has led to the emergence of a gray market that frequently compromises safety and environmental regulations. National authorities and private entities are also intervening, but these initiatives have yet to keep up with the overwhelming influx of batteries being discarded. Read the entire story.
—Caiwei Chen
The AI pessimists remain unmoved
It’s an unusual moment to identify as an AI pessimist. This small yet influential group is convinced, simply put, that AI could advance to a point of becoming harmful—extremely harmful—for humanity.
The doomsayer faction has seen some significant accomplishments in recent years, such as influencing AI policies enacted by the Biden administration. However, various occurrences over the last six months have forced them to re-evaluate their positioning. Conversations about a potential AI bubble have dominated discussions as technology firms persist in pouring resources into numerous significant data centers without any assurance that future needs will correlate with their investments.
So, what is the current status for these pessimists? We reached out to some of the prominent figures within the movement to inquire whether recent obstacles and a general shift in sentiment have changed their perspectives. Discover their insights in our article.
—Garrison Lovely
This piece is part of our new Hype Correction collection, a series of articles aimed at resetting your expectations regarding the capabilities of AI—and its limitations. Explore the rest of the collection.
Take our quiz on this year in health and biotechnology
In just a few weeks, we’ll be bidding goodbye to 2025. What an eventful year it has been! Artificial intelligence is being integrated into more facets of our existence, weight-loss medications have broadened in availability, and there have been striking “omg” biotech developments in gene therapy, IVF, neurotechnology, and beyond.
Jessica Hamzelou, our senior biotech correspondent, invites you to challenge your own recollection. How closely have you been following this year’s events?
This article originally appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday and get early access to articles like this, sign up here.
The essential reads
I’ve scoured the internet to bring you the day’s most entertaining/important/scary/enlightening stories related to technology.
1 TikTok has finalized an agreement to divest its US operation
Its new owner will be a joint venture overseen by American investors including Oracle. (Axios)
+ However, the platform insists that its Chinese parent company will still own its primary US business. (FT $)
+ The acquisition is expected to conclude by January 22 of next year. (Bloomberg $)
+ This allows TikTok to avoid a US ban—at least for the time being. (The Guardian)
2 A lead on Reddit assisted in concluding the search for the Brown University shooter
The suspect, who has been discovered deceased, is also being investigated for the murder of an MIT professor. (NYT $)
+ The shooter’s motives remain unclear, according to police. (WP $)
3 Tech figures are among those included in newly-released Epstein images
Bill Gates and Google’s Sergey Brin are featured in the photographs. (FT $)
+ They’ve been extracted from a batch of over 95,000. (Wired $)
4 A Starlink satellite seems to have exploded
And is currently descending back to earth. (The Verge)
+ On-site in Ukraine’s largest Starlink repair facility. (MIT Technology Review)
5 YouTube has terminated two prominent channels that distributed fake movie trailers
Screen Culture and KH Studio produced AI-generated parody trailers totaling over a billion views. (Deadline)
+ Google is walking a fine line between embracing and rejecting generative AI. (Ars Technica)
6 Trump is tightening regulations on investments in Chinese tech companies
Legislators are becoming increasingly concerned that US capital is strengthening the nation’s surveillance apparatus. (WSJ $)
+ Meanwhile, China is enhancing its chip production capabilities. (FT $)
7 ICE has contracted an AI agent firm to pursue targets
It claims to quickly track a target’s online network. (404 Media)
8 The US aims to return to the Moon by 2028
And to establish nuclear reactors while there. (Ars Technica)
+ Southeast Asia is aspiring to find its footing in space. (MIT Technology Review)
9 UK actors are declining to be scanned for AI
They are reportedly often pressured to consent to the creation of digital representations of themselves. (The Guardian)
+ How Meta and AI entities enlisted striking actors for AI training. (MIT Technology Review)
10 Indian instructors are teaching AI usage via WhatsApp
The lessons are affordable and tailored—but the educators may not always be credible. (Rest of World)
+ How Indian health workers utilize WhatsApp to support pregnant women. (MIT Technology Review)
Quote of the day
“Trump aims to give even more control over your viewing choices to his wealthy associates. Americans deserve transparency regarding whether the president has forged another backdoor arrangement for this billionaire acquisition of TikTok.”
—Democratic senator Elizabeth Warren queries the stipulations of the agreement permitting TikTok to maintain its operations in the US in a post on Bluesky.
One last note
Synthesia’s AI avatars are becoming increasingly expressive. They will soon be able to engage in conversation.
—Rhiannon Williams
Earlier this summer, I visited Synthesia, an AI firm, to create a hyper-realistic avatar of myself. The avatars produced by the company serve as an excellent indicator of the rapid advancements made in AI over the recent years, so I was eager to see how convincingly its latest AI model, launched last month, could emulate me.
I found my avatar both unsettling and technically captivating. It is polished enough to be mistaken for a high-definition recording of a lively corporate presentation, and if you were unfamiliar with me, you might just presume that’s what it was.
My avatar illustrates how increasingly difficult it is to differentiate between the artificial and the authentic. Soon enough, these avatars will be capable of conversing with us. But how much further can they develop? What could our interactions with AI avatars lead to? Read the entire article.
We can still enjoy nice things
A haven for comfort, enjoyment, and diversion to enhance your day. (Have any suggestions? Drop me a message or send me a tweet.)
+ You can keep your beef fat—here are the food trends that should firmly remain in 2025.
+ The Library of Congress features some beautiful winter images that are entirely free for use.
+ If you have a last-minute Christmas office party tonight, avoid making these Secret Santa blunders.
+ Did you know that Billie Eilish’s hit Birds of a Feather shares the same chord progression as Wham’s Last Christmas? They actually blend quite well together.
