This week, OpenAI announced what it calls the 12 days of OpenAI, or 12 days of shipmas. On December 4, CEO Sam Altman took to X to announce that the company would be “doing 12 days of openai. each weekday, we will have a livestream with a launch or demo, some big ones and some stocking stuffers.”
The company will livestream about new products every morning for 12 business days in a row during December. It’s an impressive-sounding (and media-savvy) schedule, to be sure. But it also speaks to how tight the race between the AI bigs has become, and also how much OpenAI is scrambling to build more revenue.
While it remains to be seen whether or not they’ve got AGI in a pear tree up their sleeve, and maybe putting aside whether or not Sam Altman is your true love, the man can ship. OpenAI has been a monster when it comes to actually getting new products out the door and into the hands of users. It’s hard for me to believe that it was just two years ago, almost exactly, that it released ChatGPT. That was a world-changing release, but was also just one of many. The company has been on an absolute tear: Since 2022, it’s shipped DALL-E 2, DALL-E 3, GPT-4, ChatGPT Plus, a realtime API, GPT-4o, an advanced voice mode, a preview version of a new model called o1, and a web search engine. And that’s just a partial list.
When it kicked off its 12-days shenanigans on Thursday, it was with an official roll out of OpenAI o1 and a new, $200-per-month service called ChatGPT Pro. Friday morning, it followed that up with an announcement about a new model customization technique.
If the point you have taken away from all this is that OpenAI is very, very bad at naming things, you would be right. But! There’s another point to be made, which is that the stuff it is shipping is not coming out in a vacuum anymore, as it was two years ago. When DALL-E 2 shipped, OpenAI seemed a little like the only game in town. That was still mostly true when ChatGPT came out a few months later. But those releases sent Google into full-on freakout mode, issuing a “code red” to catch up. And then it was off to the races.
Now, there is a full-scale sprint happening between OpenAI, Google (which released its Gemini models to the public almost exactly a year ago), Anthropic (which was founded by a bunch of OpenAI formers), Meta, and, to some extent, Microsoft (OpenAI’s partner).
To wit: A little over a month ago, Anthropic unveiled a bananas demo of its chatbot Claude’s ability to use a computer. On Thursday (aka: the first day of shipmas), Microsoft announced a version of CoPilot that can follow along with you while you browse the web using AI vision. And ahead of what is widely predicted to be OpenAI’s biggest release of shipmas, its new video generation tool Sora, Google jumped ahead with its own generative video product, Veo (although it has not released it widely to the public yet).
Oh. There was also one other announcement from OpenAI, just ahead of shipmas, that seems relevant. On Wednesday, it announced a new partnership with defense contractor Anduril. Some of you may remember that OpenAI is the company that had once pledged not to let its technology be used for weapons development or the military. As James O’Donnell points out, “OpenAI’s policies banning military use of its technology unraveled in less than a year.”
This is notable in its own right, but also in crystallizing just how much OpenAI needs cold hard cash. See also: the new $200-per-month ChatGPT Pro tier. (And while recurring revenue from users will bring in some much-needed cash flow, there is a fortune in defense spending.) In addition, the company is looking into bringing paid advertisements to its services, according to its CFO Sarah Friar in an interview with the FT way back in … (checks watch) … Monday.
As has been oft-discussed, OpenAI is just incinerating piles of money. It’s on track to lose billions and billions of dollars for several more years. It has to start bringing in more revenue, lots more. And to do that it has to stay ahead of its rivals. And to do that, it has to get new, compelling products to market that are better in some way than what its competitors offer. Which means it has to ship. And monetize. And ship. And monetize. Because Google and Anthropic and Meta and a host of others are all going to keep coming out with new products, and new services too.
The arms race is on. And while the 12 days of shipmas may seem jolly, internally I bet it feels a lot more like Santa’s workshop on December 23. Pressure’s on. Time to deliver.
If someone forwarded you this edition of The Debrief, you can subscribe here. I appreciate your feedback on this newsletter. Drop me a line at mat.honan@technologyreview.com with any and all thoughts. And of course, I love tips.
Every week, I talk to one of MIT Technology Review’s journalists to go behind the scenes of a story they are working on. This week, I hit up Amanda Silverman, our features and investigations editor, about our big story on the way the war in Ukraine is reshaping the tech sector in eastern Europe.
Amanda: I think that’s pretty spot on. Though maybe it’s more accurate to say, less expensive, more-easily-built tech products. It’s all relative, right? Like, the retrofitted consumer drones that have been so prevalent in Ukraine over the past few years are vastly cheaper than traditional weapons systems, and what we’re seeing now is that lots of other tech that was initially developed for civilian purposes—like, Pete reported on a type of scooter—are being sent to the front. And again, these are much, much cheaper than traditional weaponry. And they can be developed and shipped out really quickly.
The other thing Pete found was that this tech is being quickly reworked to respond to battlefield feedback—like that scooter has been customized to carry NATO standard-sized bullet boxes. I can’t imagine that happening in the old way of doing things.
Mat: It’s move fast and (hope not to) break things, but for war…. There is also this other, much scarier idea in there, which is that the war is changing, maybe has changed, Eastern Europe’s tech sector. What did Pete find is happening there?
Amanda: So a lot of the countries neighboring Ukraine are understandably pretty freaked out by what happened there and how the country had to turn on a dime to respond to the full-scale invasion by Russia. At the same time, Pete found that a lot of people in these countries, particularly in Latvia and particularly leading tech startups, have been inspired by how Ukrainians mobilized for the war and they’re trying to sort of get ahead of the potential enemy and get ready for a conflict within their borders. It’s not all scary, to be clear. It’s arguably somewhat thrilling to see all this innovation happening so quickly and to have some of the more burdensome red tape removed.
Mat: Okay so Russia’s neighbors are freaked out, as you say, understandably. Did anything about this story freak you out?
Amanda: Yeah, it’s impossible to ignore that there is a huge, scary risk here, too: as these companies develop new tech for war, they have an unprecedented opportunity to test it out in Ukraine without going through the traditional development and procurement process—which can be slow and laborious, sure, but also includes a lot of important testing, checks and balances, and more to prevent fraud and lots of other abuses and dangers. Like, Pete nods to how Clearview AI was deploying its tech to identify Russian war dead, which is scary in and of itself and also may violate the Geneva Conventions.
Mat: And then I’m curious, what do you look for when you are assigning a story like this? What caught your attention?
Amanda: I felt like I’d read quite a bit about the total mobilization of Ukrainian society (including a story from Pete inWired). But I had sort of thought about all this activity as happening in a bit of a vacuum. Or at least in a limited sense, within Ukrainian borders. Of course, the US and our European allies are sending loads of money and loads of weapons but (at least as I understand it) they’re largely weapons we already have in our arsenals. So when Pete pitched us this story about how the war was reshaping the tech sector of Ukraine’s neighbors, particularly civilian tech, I was really intrigued.
The Recommendation
Several weeks ago, we had our e-bike stolen. Some guy with an angle iron cut the lock. And as it turned out, our insurance didn’t cover the loss because the bike (like almost all e-bikes) had a top speed above 15 mph. As I came to learn, this is not uncommon. But you know what is common? E-bike theft. The police told us there is little chance of recovering our bike—in large part because we did not have a tracker attached to it. It was an all-around frustrating experience. We replaced the bike, and this time I’ve invested in one of these Elevation Labs waterproof mounts to affix an AirTag to the frame, hidden away below the seat. They have a whole line of mounts, a few of which are bike-specific. Much cheaper than a new bike. They make a good stocking stuffer.
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.
The world’s next big environmental problem could come from space
In September, a unique chase took place in the skies above Easter Island. From a rented jet, a team of researchers captured a satellite’s last moments as it fell out of space and blazed into ash across the sky, using cameras and scientific equipment. Their hope was to gather priceless insights into the physical and chemical processes that occur when satellites burn up as they fall to Earth at the end of their missions.
This kind of study is growing more urgent. The number of satellites in the sky is rapidly rising—with a tenfold increase forecast by the end of the decade. Letting these satellites burn up in the atmosphere at the end of their lives helps keep the quantity of space junk to a minimum. But doing so deposits satellite ash in the Earth’s atmosphere. This metallic ash could potentially alter the climate, and we don’t yet know how serious the problem is likely to be. Read the full story.
—Tereza Pultarova
OpenAI’s “12 days of shipmas” tell us a lot about the AI arms race
Last week, OpenAI announced what it calls the 12 days of OpenAI, or 12 days of shipmas. On December 4, CEO Sam Altman took to X to announce that the company would be “doing 12 days of openai. each weekday, we will have a livestream with a launch or demo, some big ones and some stocking stuffers.”
The company will livestream about new products every morning for 12 business days in a row during December. It’s an impressive-sounding (and media-savvy) schedule, to be sure. But it also speaks to how tight the race between the AI bigs has become, and also how much OpenAI is scrambling to build more revenue. Read the full story.
—Mat Honan
This story originally appeared in The Debrief with Mat Honan, our weekly take on what’s really going on behind the biggest tech headlines. The story is subscriber-only so nab a subscription too, if you haven’t already! Or you can sign up to the newsletter for free to get the next edition in your inbox on Friday.
The must-reads
I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.
1 The USDA is launching a national program to test milk for bird flu A full nine months after the current outbreak was first detected in dairy cows. (STAT) + The risk of a bird flu pandemic is rising. (MIT Technology Review)
2 Here’s what sets OpenAI’s new models apart They’re shifting from predicting to reasoning, which could be a huge deal. (The Atlantic $) + Regardless of whether capabilities are slowing, AI’s impact is only poised to grow. (Vox) + It may be comforting to dismiss AI as hype—but it misses the point. (Platformer)
3 A federal appeals court has upheld the US TikTok ban But what happens next is anyone’s guess. (WSJ $) + Whether TikTok is banned or not, the actions against it have had a big impact. (MIT Technology Review)
4 Top internet sleuths are sitting out the hunt for the UnitedHealthcare CEO killer In fact, some are even criticizing people who are trying to help. (NBC) + Why so many Americans are at best indifferent to this particular murder. (New Yorker $)
5 Schools are attempting to stop teens self-harming before they even try The AI tools they’re adopting could be doing far more damage than help, though. (NYT $)
6 China is building its own Starlink system The Qianfan constellation could eventually grow to nearly 14,000 satellites. (The Economist $) + The end of the ISS will usher in a more commercialized future in space. (The Verge)
7 This was an exciting year for superconductors Superconductivity—the flow of electric current with no resistance—was discovered in three new materials. (Quanta $)
8 Meet the world’s least productive programmers It seems a small minority of disillusioned ‘ghost engineers’ do pretty much no work at all. (WP $)
9 Why people are turning their backs on dating apps There’s a large degree of fatigue, and a feeling that they’re somehow detached from reality. (The Guardian)
10 Fake snacks are racking up millions of views on Instagram There’s even a word for this trend: snackfishing. (Wired $)
Quote of the day
“I think Twitter and now X is like a crack addiction for him, though. He is clearly chasing a particular hit all the time and he has ended up self-radicalising himself with the platform he has purchased.”
—A former Twitter employee in London tells The Guardian how Elon Musk has changed since he purchased the platform.
The big story
How electricity could help tackle a surprising climate villain
Sublime Systems
BOB O’CONNOR
January 2024
Cement is used to build everything from roads and buildings to dams and basement floors. But it’s also a climate threat. Cement production accounts for more than 7% of global carbon dioxide emissions—more than sectors like aviation, shipping, or landfills.
One solution to this climate catastrophe might be coursing through the pipes at Sublime Systems. The startup is developing an entirely new way to make cement. Instead of heating crushed-up rocks in lava-hot kilns, Sublime’s technology zaps them in water with electricity, kicking off chemical reactions that form the main ingredients in its cement.
But it faces huge challenges: competing with established industry players, and persuading builders to use its materials in the first place. Read the full story.
—Casey Crownhart
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 or tweet ’em at me.)
+ Who will be the Lord of Misrule in your household this Christmas? + People’s Wikipedia browsing data always makes for interesting reading. + Wait, so we’ve been mispronouncing these words all along? (Apart from espresso, c’mon) + The Muppet Christmas Carol might just be the greatest festive film.
MIT Technology Review’s How Toseries helps you get things done.
Today, OpenAI released its video generation model Sora to the public. The announcement comes on the fifth day of the company’s “shipmas” event, a 12-day marathon of tech releases and demos. Here’s what you should know—and how you can use the video model right now.
What is Sora?
Sora is a powerful AI video generation model that can create videos from text prompts, animate images, or remix videos in new styles. OpenAI first previewed the model back in February, but today is the first time the company is releasing it for broader use.
What’s new about this release?
The core function of Sora—creating impressive videos with simple prompts—remains similar to what was previewed in February, but OpenAI worked to make the model faster and cheaper ahead of this wider release. There are a few new features, and two stand out.
One is called Storyboard. With it, you can create multiple AI-generated videos and then assemble them together on a timeline, much the way you would with conventional video editors like Adobe Premiere Pro.
The second is a feed that functions as a sort of creative gallery. Users can post their Sora-generated videos to the feed, see the prompts behind certain videos, tweak them, and generally get inspiration, OpenAI says.
How much can you do with it?
You can generate videos from text prompts, change the style of videos and change elements with a tool called Remix, and assemble multiple clips together with Storyboard. Sora also provides preset styles you can apply to your videos, like moody film noir or cardboard and papercraft, which gives a stop-motion feel. You can also trim and loop the videos that you make.
Who can use it?
To generate videos with Sora, you’ll need to subscribe to one of OpenAI’s premium plans—either ChatGPT Plus ($20 per month) or ChatGPT Pro ($200 per month). Both subscriptions include access to other OpenAI products as well. Users with ChatGPT Plus can generate videos as long as five seconds with a resolution up to 720p. This plan lets you create 50 videos per month.
Users with a ChatGPT Pro subscription can generate longer, higher-resolution videos, capped at a resolution of 1080p and a duration of 20 seconds. They can also have Sora generate up to five variations of a video at once from a single prompt, making it possible to review options faster. Pro users are limited to 500 videos per month but can also create unlimited “relaxed” videos, which are not generated in the moment but rather queued for when site traffic is low.
Both subscription levels make it possible to create videos in three aspect ratios: vertical, horizontal, and square.
If you don’t have a subscription, you’ll be limited to viewing the feed of Sora-generated videos.
OpenAI is starting its global launch of Sora today, but it will take longer to launch in “most of Europe,” the company said.
OPENAI
Where can I access it?
OpenAI has broken Sora out from ChatGPT. To access it, go to Sora.com and log in with your ChatGPT Plus or Pro account. (MIT Technology Review was unable to access the site at press time—a note on the site indicated that signups were paused because they were “currently experiencing heavy traffic.”)
How’d we get here?
A number of things have happened since OpenAI first unveiled Sora back in February. Other tech companies have also launched video generation tools, like Meta Movie Gen and Google Veo. There’s also been plenty of backlash. For example, artists who had early access to experiment with Sora leaked the tool to protest the way OpenAI has trained it on artists’ work without compensation.
What’s next?
As with any new release of a model, it remains to be seen what steps OpenAI has taken to keep Sora from being used for nefarious, illegal, or unethical purposes, like the creation of deepfakes. On the question of moderation and safety, an OpenAI employee said they “might not get it perfect on day one.”
Another looming question is how much computing capacity and energy Sora will use up every time it creates a video. Generating a video uses much more computing time, and therefore energy, than generating a typical text response in a tool like ChatGPT. The AI boom has already been an energy hog, presenting a challenge to tech companies aiming to rein in their emissions, and the wide availability of Sora and other video models like it has the potential to make that problem worse.
This article is from The Spark, MIT Technology Review’s weekly climate newsletter. To receive it in your inbox every Wednesday, sign up here.
Last week, we celebrated Thanksgiving here in the US, and I had hearty helpings of ham and turkey alongside my mashed potatoes and green bean casserole.
Meat is often the star on our plates, but our love of animal-based foods is a problem for the climate. Depending on how you count it up, livestock accounts for somewhere between 10% and 20% of all greenhouse gas emissions.
A growing number of alternative foods seek to mimic or replace options that require raising and slaughtering animals. These include plant-based products and newly approved cultivated (or lab-grown) meats. An increasing number of companies are even raising microbes in the lab in the hopes that we’ll add them to the menu, as I covered in a story this week.
But as one of my colleagues always puts it when I tell him about some alternative food product, the key question is, will anyone eat it?
Food might just be one of the trickiest climate problems to solve. Technically, none of us has to be eating any of the highest-emissions foods—like beef—that are worst for the climate. But what we eat is deeply personal, and it’s often tied up with our culture and our social lives. Many people want hamburgers at a barbecue and nice steak dinners.
The challenge of our food system’s climate impact is only getting more tricky: richer countries tend to eat more meat, and so as populations grow and the standard of living rises around the world, we’re going to see emissions from livestock production rise, too.
In an effort to combat that trend, alternative food products aim to deliver foods similar to the ones we know and love with less harmful effects on the climate. Plant-based options like those from Beyond Meat and Impossible Foods have exploded in recent years, finding their way into supermarkets and even onto the menus of major fast-food brands like Burger King.
The problem is, a lot of alternative products have been struggling lately. Unit sales of meat alternatives in the US were down by 26% between 2021 and 2023, and fewer households are buying plant-based alternative meat options, according to a report from the Good Food Institute. Consumers say that alternatives still aren’t up to par on taste and price, two key factors that determine what people decide to eat.
So companies are racing to invent better products. I’ve spent a lot of time covering cultivated (or lab-grown) meats. To make these products, animal cells are grown in the lab and processed into things like chicken nuggets. Two companies got approval to sell cultivated chicken in the US in 2023, and we’ve seen both offer their products in limited runs at high-end restaurants.
But these products are still not quite the same thing as the meat we’re used to. When I tried a burger that contained cells grown in a lab, it was similar to plant-based ones that have a softer texture than I’m used to. Chicken from Upside Foods, served at a Michelin-starred restaurant, had similar textural differences. And these products are still only available at very small scales, if at all, and they’re expensive.
LANZATECH
One key issue that comes up again and again as I report on these new products is what to call them. The industry strongly prefers cultivated, not “lab-grown.” Probably better to not remind people that they’re eating something grown in vats in a laboratory. As the companies that make these products often point out, we don’t typically use this sort of language for the animal-based products we’re used to. You’d never find the phrase “slaughtered baby cow” on a menu, just “veal.”
I was thinking about this issue of language and marketing again recently as I reported a story about a company looking to grow bacteria, dry it, and sell it to feed animals or people. I found myself a little weirded out by the prospect of dried microbe powder finding its way into my diet. But I don’t have a problem drinking wine or eating cheese, two products that rely on microbes and a fermentation process to exist.
Maybe LanzaTech will come up with a marketing plan that makes their microbe powder an easy addition to my Thanksgiving table. Ultimately though, no matter how well they’re marketed, I’m not sure how much we can rely on alternative products to solve the climate challenge that is our food system.
As is often the case when it comes to addressing climate change, we’re going to need not only some behavioral changes, but also technical solutions like cattle burp pills and new fertilizer options, as well as policy to help nudge our food system in the right direction.
Now read the rest of The Spark
Related reading
A new crop of biotech startups is looking to grow food out of thin air. Read more about a few of the leading businesses in this story from earlier this fall.
Rumin8 and Pivot Bio, two of our Climate Tech Companies to Watch this year, are both working to address emissions from agriculture.
Keeping up with climate
China announced it would ban the export to the US of several rare minerals that are crucial in technology like semiconductors. The move follows efforts by the US to shift supply chains away from China. (New York Times)
Donald Trump has pledged to ramp up tariffs on Chinese goods, while other nations around the world have already put such policies in place. (Rest of World)
Australia is on track to meet its 2030 emissions target. The country’s climate pollution is projected to fall more than 42% below 2005 levels by the end of the decade. (Bloomberg)
Talks to form an international plastic treaty fell apart this week. Some countries favored cutting down plastic production, while others, including oil-rich nations, pushed back. (Washington Post)
The US Department of Energy announced a nearly $7 billion loan to Stellantis and Samsung for two battery factories that will supply batteries for EVs. (New York Times) → That follows a $6.6 billion loan to Rivian to help the company build a stalled factory in Georgia. (Associated Press) → The Biden administration is racing to lock in loans and safeguard them against rollbacks before Donald Trump takes office in January. (E&E News)
California could increase use of ethanol, a move the state says could lower gas prices. But experts warn that expanded use of ethanol made from corn can have negative consequences for climate progress and the environment. (Inside Climate News)
Norway’s government is blocking plans to mine the sea bed. There were plans to begin offering permits in the first half of 2025, and preparations will continue during the suspension. (Reuters) → These deep-sea “potatoes” could be the future of mining for battery materials. (MIT Technology Review)
A decade ago, sea surface temperatures in the Pacific shot up in a dramatic marine heat wave. Now, scientists are looking for clues in that event to understand what rising temperatures will mean for the ocean. (New York Times)
MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here.
This week, China banned exports of several critical minerals to the US, marking the latest move in an escalating series of tit-for-tat trade restrictions between the world’s two largest economies.
In explicitly cutting off, rather than merely restricting, materials of strategic importance to the semiconductor, defense, and electric vehicle sectors, China has clearly crossed a new line in the long-simmering trade war.
At the same time, it selected minerals that won’t cripple any industries—which leaves China plenty of ammunition to inflict greater economic pain in response to any further trade restrictions that the incoming Trump administration may impose.
The president-elect recently pledged to impose an additional 10% tariff on all Chinese goods, and he floated tariff rates as high as 60% to 100% during his campaign. But China, which dominates the supply chains for numerous critical minerals essential to high-tech sectors, seems to be telegraphing that it’s prepared to hit back hard.
“It’s a sign of what China is capable of,” says Gracelin Baskaran, director of the Critical Minerals Security Program at the Center for Strategic and International Studies, a bipartisan research nonprofit in Washington, DC. “Shots have been fired.”
What drove the decision?
China’s announcement directly followed the Biden administration’s decision to further restrict exports of chips and other technologies that could help China develop advanced semiconductors used in cutting-edge weapon systems, artificial intelligence, and other applications.
Throughout his presidency, Biden has enacted a series of increasingly aggressive export controls aimed at curbing China’s military strength, technological development, and growing economic power. But the latest clampdown crossed a “clear line in the sand for China,” by threatening its ability to protect national security or shift toward production of more advanced technologies, says Cory Combs, associate director at Trivium China, a research firm.
“It is very much indicative of where Beijing feels its interests lie,” he says.
What exactly did China ban?
In response to the US’s new chip export restrictions, China immediately banned exports of gallium, germanium, antimony, and so called “superhard materials” used heavily in manufacturing, arguing that they have both military and civilian applications, according to the New York Times. China had already placed limits on the sale of most of these goods to the US.
The nation said it may also further restrict sales of graphite, which makes up most of the material in the lithium-ion battery anodes used in electric vehicles, grid storage plants, and consumer electronics.
What will the bans do?
Experts say, for the most part, the bans won’t have major economic impacts. This is in part because China already restricted exports of these minerals months ago, and also because they are mostly used for niche categories within the semiconductor industry. US imports of these materials from China have already fallen as US companies figured out new sources or substitutes for the materials.
But a recent US Geological Survey study found that outright bans on gallium and germanium by China could cut US gross domestic product by $3.4 billion. In addition, these are materials that US politicians will certainly take note of, because they “touch on many forms of security: economic, energy, and defense,” Baskaran says.
Antimony, for example, is used in “armor-piercing ammunition, night-vision goggles, infrared sensors, bullets, and precision optics,” Baskaran and a colleague noted in a recent essay.
Companies rely on gallium to produce a variety of military and electronics components, including satellite systems, power converters, LEDs, and the high-powered chips used in electric vehicles. Germanium is used in fiber optics, infrared optics, and solar cells.
Before it restricted the flow of these materials, China accounted for more than half of US imports of gallium and germanium, according to the US Geological Survey. Together, China and Russia control 50% of the worldwide reserves of antimony.
How does it affect climate tech?
Any tightened restrictions on graphite could have a pronounced economic impact on US battery and EV makers, in part because there are so few other sources for it. China controls about 80% of graphite output from mines and processes around 70% of the material, according to the International Energy Agency.
“It would be very significant for batteries,” says Seaver Wang, co-director of the climate and energy team at the Breakthrough Institute, where his research is focused on minerals and manufacturing supply chains. “By weight, you need way more graphite per terawatt hour than nickel, cobalt, or lithium. And the US has essentially no operating production.”
Anything that pushes up the costs of EVs threatens to slow the shift away from gas-guzzlers in the US, as their lofty price tags remain one of the biggest hurdles for many consumers.
How does this impact China’s economy?
There are real economic risks in China’s decision to cut off the sale of materials it dominates, as it creates incentives for US companies to seek out new sources around the world, switch to substitute materials, and work to develop more domestic supplies where geology allows.
“The challenge China faces is that most of its techniques to increase pain by disrupting supply chains would also impact China, which itself is connected to these supply chains,” says Chris Miller, a professor at Tufts University and author of Chip War: The Fight for the World’s Most Critical Technology.
Notably, the latest announcement could compel US companies to develop their own sources of gallium and germanium, which can be extracted as by-products of zinc and aluminum mining. There are a number of zinc mines in Alaska and Tennessee, and limited extraction of bauxite, which produces aluminum, in Arkansas, Alabama, and Georgia.
Gallium can also be recycled from numerous electronics, providing another potential domestic path for US companies, Combs notes.
The US has already taken steps to counter China’s dominance over the raw ingredients of essential industries, including by issuing a $150 million loan to an Australian company, Syrah Resources, to accelerate the development of graphite mining in Mozambique.
In addition, the mining company Perpetua Resources has proposed reopening a gold mine near Yellow Pine, Idaho, in part to extract antimony trisulfide for use in military applications. The US Department of Defense has provided tens of millions of dollars to help the company conduct environmental studies, though it will still take years for the mine to come online, noted Baskaran and her colleague.
Wang says that China’s ban might prove “shortsighted,” as any success in diversifying these global supply chains will weaken the nation’s grip in the areas it now dominates.
What happens next?
The US is also likely to pay very high economic costs in an escalating trade war with China.
Should the nation decide to enact even stricter trade restrictions, Combs says China could opt to inflict greater economic pain on the US through a variety of means. These could include further restricting or fully banning graphite, as well other crucial battery materials like lithium; cutting off supplies of tungsten, which is used heavily in the aerospace, military, and nuclear power sectors; and halting the sale of copper, which is used in power transmission lines, solar panels, wind turbines, EVs, and many other products.
China may also decide to take further steps to prevent US firms from selling their goods into the massive market of Chinese consumers and industries, Miller adds. Or it might respond to stricter export restrictions by turning to the US’s economic rivals for advanced technologies.
In the end, it’s not clear either nation wins in a protracted and increasingly combative trade war. But it’s also not apparent that mutually assured economic damage will prove to be an effective deterrent. Indeed, China may well feel the need to impose stricter measures in the coming months or years, as there are few signs that President-elect Trump intends to tone down his hawkish stance toward China.
“It’s hard to see a Trump 2.0 de-escalating with China,” Baskaran says. “We’re on a one-way trajectory toward continued escalation; the question is the pace and the form. It’s not really an ‘if” question.”
This article first appeared in The Checkup, MIT Technology Review’s weekly biotech newsletter. To receive it in your inbox every Thursday, and read articles like this first, sign up here.
There’s a new film about IVF out on Netflix. And “everyone in the field [of reproductive medicine] has watched it,” according to one embryologist I spoke to recently. Joy is a lovely watch about the birth of the field, thanks to the persistent efforts of Robert Edwards, Jean Purdy, and Patrick Steptoe in the face of significant opposition.
IVF is a success story for embryo research. But today, valuable embryos that could be used for research are being wasted, say researchers who gathered at a conference in central London earlier this week.
The conference was organized by the Progress Educational Trust, a UK-based charity that aims to provide information to the public on genomics and infertility. The event marked 40 years since the publication of the Warnock Report, which followed a governmental inquiry into infertility treatment and embryological research. The report is considered to be the first to guide recognition of the embryo’s “special” status in law and helped establish regulation of the nascent technology in the UK.
The report also endorsed the “14-day rule,” which limits the growth of embryos in a lab to this two-week point. The rule, since adopted around the world, is designed to prevent scientists from growing embryos to the point where they develop a structure called the primitive streak. At this point, the development of tissues and organs begins, and the embryo is no longer able to split to form twins.
The embryos studied in labs have usually been created for IVF but are no longer needed by the people whose cells created them. Those individuals might have completed their families, or they might not be able to use the embryos because their circumstances have changed. Sometimes the embryos have genetic abnormalities that make them unlikely to survive a pregnancy.
These embryos can be used to learn more about how humans develop before birth, and to discover potential treatments for developmental disorders like spina bifida or heart defects, for example. Research on embryos can help reveal clues about our fundamental biology, and provide insight into pregnancy and miscarriage.
A survey conducted by the Human Fertility and Embryology Authority, which regulates reproductive technology in the UK, found that the majority of patients would rather donate their embryos to research than allow them to “perish,” Geraldine Hartshorne, director of the Coventry Centre for Reproductive Medicine, told the audience.
There are a few reasons why embryos aren’t making it into research labs, says Hartshorne. Part of the problem is that most IVF cycles happen at clinics that don’t have links with academic research centers.
As things stand, embryos tend to be stored at the clinics where they were created. It can be difficult to get them to research centers—clinic staff don’t have the time, energy, or head space to manage the paperwork legally required to get embryos donated to specific research projects, said Hartshorne. It would make more sense to have some large, central embryo bank where people could send embryos to donate for research, she added.
A particular problem is the paperwork. While the UK is rightly praised for its rigorous approach to regulation of reproductive technologies, which embryologists around the globe tend to describe as “world-leading,” there are onerous levels of bureaucracy to contend with, said Hartshorne. “When patients contact me and say ‘I’d like to give my embryos or my eggs to your research project,’ I usually have to turn them away, because it would take me a year to get through the paperwork necessary,” she said.
Perhaps there’s a balance to be struck. Research on embryos has the potential to be hugely valuable. As the film Joy reminds us, it can transform medical practice and change lives.
“Without research, there would be no progress, and there would be no change,” Hartshorne said. “That is definitely not something that I think we should aspire to for IVF and reproductive science.”
Mexico’s Sinaloa cartel is recruiting young chemistry students from colleges to make fentanyl. Specifically, the students are being tasked with the often dangerous job of trying to synthesize precursor chemicals that must currently be imported. They also try to design stronger versions of the drug that are more likely to get users hooked. (New York Times)
Billionaire Greg Lindberg is running his own “baby project.” Having duped, misled, and paid off a series of egg donors and surrogates, the disgraced insurance tycoon currently has 12 children, nine of whom were born in the last five years or so. He is the sole parent caring for eight of them, despite facing significant jail time since being convicted of bribery and pleading guilty to money laundering and fraud conspiracy charges for crimes unrelated to the baby project. The scale of his project is an indictment of the US fertility industry. (Bloomberg Businessweek)
The UK government has agreed to a contract for more than 5 million doses of a vaccine designed to protect people from the H5 bird flu virus. The vaccine is being procured as part of pandemic preparedness plans and will be used only if the virus starts spreading among humans. (UK Health Security Agency)
Last week, MPs voted in favor of a bill to legalize assisted dying in England and Wales. In the past few months, the debate over the bill has included horror stories of painful deaths. Most deaths are “ordinary,” but we all stand to benefit from talking about, and understanding, what death involves. (New Statesman)
An unknown disease has killed 143 people in southwest Congo, according to local authorities. The number of infections continues to rise, and the situation is extremely worrying. (Reuters)
Brian Thompson, the 50-year-old CEO of US health insurance company UnitedHealthcare, was fatally shot in New York city on Wednesday. The New York Times is reporting that bullet casings found at the scene appear to have been marked with the words “delay” and “deny.” The words may refer to strategies used by insurance companies to avoid covering healthcare costs. (New York Times)
Next month, MIT Technology Review will unveil the 2025 list of 10 Breakthrough Technologies. Every year, our newsroom looks across the fields we cover for technologies that are having a true breakthrough moment. This annual package highlights the technologies that we think matter most right now.
We define ‘breakthrough’ in a few ways—perhaps there’s been a scientific advance that now makes a new technology possible, or a company has earned regulatory approval for a vital medical treatment. Maybe a consumer device has reached a tipping point in its adoption, or an industrial technology has passed the critical pilot phase with flying colors. In the 2025 edition, which comes out in January, you’ll see some of the latest advances in automation, medicine, and the physical sciences (just to name a few) that we expect will have a major impact on our lives.
In the meantime, here are three technologies that we considered including on the 2025 list but ultimately decided to leave off. Though these nominees didn’t make the cut this year, they’re still worth keeping an eye on. We certainly will be.
Virtual power plants
Virtual power plants are energy systems that link together many different technologies to both generate and store power. They allow utility companies to connect solar panels and wind turbines with grid batteries and electric vehicles, and to better manage the flow of power across the grid.
During times of peak electricity usage, software linked to smart meters may one day automatically decide to power someone’s home by drawing electricity from a fully charged EV sitting in a neighbor’s garage, thereby reducing demand on the grid. The software could also work out how to compensate the EV owner accordingly.
In the US, an estimated 500 virtual power plants now provide up to 60 gigawatts of capacity (that’s about as much total capacity as the US grid will add this year). Some such systems are also up and running in China, Japan, Croatia, and Taiwan. But lots more virtual power plants would need to be configured before they start to affect the grid as a whole.
Useful AI agents
AI agents are all the rage right now. These AI-powered helpers will, supposedly, schedule our meetings and book our trips and carry out all kinds of tasks online on our behalf. Agents employ generative models to learn how to navigate websites and desktop software (and manage our passwords and credit card details). They will perhaps interact and coordinate with other people’s agents along the way.
And there is real development power behind them—Salesforce just launched a platform where companies can make their own customer service agents, and Anthropic’s Claude model is gaining the ability to navigate a computer by using a mouse and keyboard, just like people.
However, many challenges remain in getting these agents to know what you mean when you make specific requests, and enabling them to carry out the necessary actions reliably. Given the formidable hurdles, we think it may be a little while before they are good enough to be truly useful. AI agents may be coming, but not just yet.
eVTOLs
The acronym is a mouthful, but you can think of electric vertical takeoff and landing (eVTOL) aircraft as being kind of like electric helicopters. Most versions in development are not designed to be personal vehicles; they’d be flown by pilots to transport commuters in from the suburbs, or whisk visitors downtown from the airport. Someday, these air taxis may fly themselves.
There’s been real progress toward getting eVTOLs off the ground. Earlier this year, manufacturer EHang received the first Chinese certificate to mass-produce this type of vehicle, and it has begun taking orders. South Korea and the UAE have put policies in place to allow eVTOLs to operate there. And in the US, Archer recently earned its FAA certification to begin commercial operations. Then, in October, the FAA finalized rules for training pilots and operating eVTOLs—marking the first time in decades that the agency has approved such rules for a new category of aircraft.
Interest and momentum have built in recent years. Major players in the aviation industry, including Boeing and Airbus, have invested in startups or funded internal R&D projects to develop these futuristic aircraft. However, no eVTOL company has actually begun commercial operations yet, so we’ll keep watching for that.
The US Department of Defense has invested $2.4 million over two years in deepfake detection technology from a startup called Hive AI. It’s the first contract of its kind for the DOD’s Defense Innovation Unit, which accelerates the adoption of new technologies for the US defense sector. Hive AI’s models are capable of detecting AI-generated video, image, and audio content.
Although deepfakes have been around for the better part of a decade, generative AI has made them easier to create and more realistic-looking than ever before, which makes them ripe for abuse in disinformation campaigns or fraud. Defending against these sorts of threats is now crucial for national security, says Captain Anthony Bustamante, a project manager and cyberwarfare operator for the Defense Innovation Unit.
“This work represents a significant step forward in strengthening our information advantage as we combat sophisticated disinformation campaigns and synthetic-media threats,” says Bustamante. Hive was chosen out of a pool of 36 companies to test its deepfake detection and attribution technology with the DOD. The contract could enable the department to detect and counter AI deception at scale.
Defending against deepfakes is “existential,” says Kevin Guo, Hive AI’s CEO. “This is the evolution of cyberwarfare.”
Hive’s technology has been trained on a large amount of content, some AI-generated and some not. It picks up on signals and patterns in AI-generated content that are invisible to the human eye but can be detected by an AI model.
“Turns out that every image generated by one of these generators has that sort of pattern in there if you know where to look for it,” says Guo. The Hive team constantly keeps track of new models and updates its technology accordingly.
The tools and methodologies developed through this initiative have the potential to be adapted for broader use, not only addressing defense-specific challenges but also safeguarding civilian institutions against disinformation, fraud, and deception, the DOD said in a statement.
Hive’s technology provides state-of-the-art performance in detecting AI-generated content, says Siwei Lyu, a professor of computer science and engineering at the University at Buffalo. He was not involved in Hive’s work but has tested its detection tools.
Ben Zhao, a professor at the University of Chicago, who has also independently evaluated Hive AI’s deepfake technology, agrees but points out that it is far from foolproof.
“Hive is certainly better than most of the commercial entities and some of the research techniques that we tried, but we also showed that it is not at all hard to circumvent,” Zhao says. The team found that adversaries could tamper with images in a way that bypassed Hive’s detection.
And given the rapid development of generative AI technologies, it is not yet certain how it will fare in real-world scenarios that the defense sector might face, Lyu adds.
Guo says Hive is making its models available to the DOD so that the department can use the tools offline and on their own devices. This keeps sensitive information from leaking.
But when it comes to protecting national security against sophisticated state actors, off-the-shelf products are not enough, says Zhao: “There’s very little that they can do to make themselves completely robust to unforeseen nation-state-level attacks.”
At first glance, the Mosphera scooter may look normal—just comically oversized. It’s like the monster truck of scooters, with a footplate seven inches off the ground that’s wide enough to stand on with your feet slightly apart—which you have to do to keep your balance, because when you flip the accelerator with a thumb, it takes off like a rocket. While the version I tried in a parking lot in Riga’s warehouse district had a limiter on the motor, the production version of the supersized electric scooter can hit 100 kilometers (62 miles) per hour on the flat. The all-terrain vehicle can also go 300 kilometers on a single charge and climb 45-degree inclines.
Latvian startup Global Wolf Motors launched in 2020 with a hope that the Mosphera would fill a niche in micromobility. Like commuters who use scooters in urban environments, farmers and vintners could use the Mosphera to zip around their properties; miners and utility workers could use it for maintenance and security patrols; police and border guards could drive them on forest paths. And, they thought, maybe the military might want a few to traverse its bases or even the battlefield—though they knew that was something of a long shot.
When co-founders Henrijs Bukavs and Klavs Asmanis first went to talk to Latvia’s armed forces, they were indeed met with skepticism—a military scooter, officials implied, didn’t make much sense—and a wall of bureaucracy. They found that no matter how good your pitch or how glossy your promo video (and Global Wolf’s promo is glossy: a slick montage of scooters jumping, climbing, and speeding in formation through woodlands and deserts), getting into military supply chains meant navigating layer upon layer of officialdom.
Then Russia launched its full-scale invasion of Ukraine in February 2022, and everything changed. In the desperate early days of the war, Ukrainian combat units wanted any equipment they could get their hands on, and they were willing to try out ideas—like a military scooter—that might not have made the cut in peacetime. Asmanis knew a Latvian journalist heading to Ukraine; through the reporter’s contacts, the startup arranged to ship two Mospheras to the Ukrainian army.
Within weeks, the scooters were at the front line—and even behind it, being used by Ukrainian special forces scouts on daring reconnaissance missions. It was an unexpected but momentous step for Global Wolf, and an early indicator of a new demand that’s sweeping across tech companies along Ukraine’s borders: for civilian products that can be adapted quickly for military use.
COURTESY OF GLOBAL WOLF
Global Wolf’s high-definition marketing materials turned out to be nowhere near as effective as a few minutes of grainy phone footage from the war. The company has since shipped out nine more scooters to the Ukrainian army, which has asked for another 68. Where Latvian officials once scoffed, the country’s prime minister went to see Mosphera’s factory in April 2024, and now dignitaries and defense officials from the country are regular visitors.
It might have been hard a few years ago to imagine soldiers heading to battle on oversized toys made by a tech startup with no military heritage. But Ukraine’s resistance to Russia’s attacks has been a miracle of social resilience and innovation—and the way the country has mobilized is serving both a warning and an inspiration to its neighbors. They’ve watched as startups, major industrial players, and political leaders in Ukraine have worked en masse to turn civilian technology into weapons and civil defense systems. They’ve seen Ukrainian entrepreneurs help bootstrap a military-industrial complex that is retrofitting civilian drones into artillery spotters and bombers, while software engineers become cyberwarriors and AI companies shift to battlefield intelligence. Engineers work directly with friends and family on the front line, iterating their products with incredible speed.
Their successes—often at a fraction of the cost of conventional weapons systems—have in turn awakened European governments and militaries to the potential of startup-style innovation and startups to the potential dual uses of their products, meaning ones that have legitimate civilian applications but can be modified at scale to turn them into weapons.
This heady mix of market demand and existential threat is pulling tech companies in Latvia and the other Baltic states into a significant pivot. Companies that can find military uses for their products are hardening them and discovering ways to get them in front of militaries that are increasingly willing to entertain the idea of working with startups. It’s a turn that may only become more urgent if the US under incoming President Donald Trump becomes less willing to underwrite the continent’s defense.
But while national governments, the European Union, and NATO are all throwing billions of dollars of public money into incubators and investment funds—followed closely by private-sector investors—some entrepreneurs and policy experts who have worked closely with Ukraine warn that Europe might have only partially learned the lessons from Ukraine’s resistance.
If Europe wants to be ready to meet the threat of attack, it needs to find new ways of working with the tech sector. That includes learning how Ukraine’s government and civil society adapted to turn civilian products into dual-use tools quickly and cut through bureaucracy to get innovative solutions to the front. Ukraine’s resilience shows that military technology isn’t just about what militaries buy but about how they buy it, and about how politics, civil society, and the tech sector can work together in a crisis.
“[Ukraine], unfortunately, is the best defense technology experimentation ground in the world right now. If you are not in Ukraine, then you are not in the defense business.”
“I think that a lot of tech companies in Europe would do what is needed to do. They would put their knowledge and skills where they’re needed,” says Ieva Ilves, a veteran Latvian diplomat and technology policy expert. But many governments across the continent are still too slow, too bureaucratic, and too worried that they might appear to be wasting money, meaning, she says, that they are not necessarily “preparing the soil for if [a] crisis comes.”
“The question is,” she says, “on a political level, are we capable of learning from Ukraine?”
Waking up the neighbors
Many Latvians and others across the Baltic nations feel the threat of Russian aggression more viscerally than their neighbors in Western Europe. Like Ukraine, Latvia has a long border with Russia and Belarus, a large Russian-speaking minority, and a history of occupation. Also like Ukraine, it has been the target of more than a decade of so-called “hybrid war” tactics—cyberattacks, disinformation campaigns, and other attempts at destabilization—directed by Moscow.
Since Russian tanks crossed into Ukraine two-plus years ago, Latvia has stepped up its preparations for a physical confrontation, investing more than €300 million ($316 million) in fortifications along the Russian border and reinstating a limited form of conscription to boost its reserve forces. Since the start of this year, the Latvian fire service has been inspecting underground structures around the country, looking for cellars, parking garages, and metro stations that could be turned into bomb shelters.
And much like Ukraine,Latvia doesn’t have a huge military-industrial complex that can churn out artillery shells or tanks en masse.
What it and other smaller European countries can produce for themselves—and potentially sell to their allies—are small-scale weapons systems, software platforms, telecoms equipment, and specialized vehicles. The country is now making a significant investment in tools like Exonicus, a medical technology platform founded 11 years ago by Latvian sculptor Sandis Kondrats. Users of its augmented-reality battlefield-medicine training simulator put on a virtual reality headset that presents them with casualties, which they have to diagnose and figure out how to treat. The all-digital training saves money on mannequins, Kondrats says, and on critical field resources.
“If you use all the medical supplies on training, then you don’t have any medical supplies,” he says. Exonicus has recently broken into the military supply chain, striking deals with the Latvian, Estonian, US, and German militaries, and it has been training Ukrainian combat medics.
Medical technology company Exonicus has created an augmented-reality battlefield-medicine training simulator that presents users with casualties, which they have to diagnose and figure out how to treat.
GATIS ORLICKIS/BALTIC PICTURES
There’s also VR Cars, a company founded by two Latvian former rally drivers, that signed a contract in 2022 to develop off-road vehicles for the army’s special forces. And there is Entangle, a quantum encryption company that sells widgets that turn mobile phones into secure communications devices, and has recently received an innovation grant from the Latvian Ministry of Defense.
Unsurprisingly, a lot of the focus in Latvia has been on unmanned aerial vehicles (UAVs), or drones, which have become ubiquitous on both sides fighting in Ukraine, often outperforming weapons systems that cost an order of magnitude more. In the early days of the war, Ukraine found itself largely relying on machines bought from abroad, such as the Turkish-made Bayraktar strike aircraft and jury-rigged DJI quadcopters from China. It took a while, but within a year the country was able to produce home-grown systems.
As a result, a lot of the emphasis in defense programs across Europe is on UAVs that can be built in-country. “The biggest thing when you talk to [European ministries of defense] now is that they say, ‘We want a big amount of drones, but we also want our own domestic production,’” says Ivan Tolchinsky, CEO of Atlas Dynamics, a drone company headquartered in Riga. Atlas Dynamics builds drones for industrial uses and has now made hardened versions of its surveillance UAVs that can resist electronic warfare and operate in battlefield conditions.
Agris Kipurs founded AirDog in 2014 to make drones that could track a subject autonomously; they were designed for people doing outdoor sports who wanted to film themselves without needing to fiddle with a controller. He and his co-founders sold the company to a US home security company, Alarm.com, in 2020. “For a while, we did not know exactly what we would build next,” Kipurs says. “But then, with the full-scale invasion of Ukraine, it became rather obvious.”
His new company, Origin Robotics, has recently “come out of stealth mode,” he says, after two years of research and development. Origin has built on the team’s experience in consumer drones and its expertise in autonomous flight to begin to build what Kipurs calls “an airborne precision-guided weapon system”—a guided bomb that a soldier can carry in a backpack.
The Latvian government has invested in encouraging startups like these, as well as small manufacturers, to develop military-capable UAVs by establishing a €600,000 prize fund for domestic drone startups and a €10 million budget to create a new drone program, working with local and international manufacturers.
VR Cars was founded by two Latvian former rally drivers and has developed off-road vehicles for the army’s special forces.
Latvia is also the architect and co-leader, with the UK, of the Drone Coalition, a multicountry initiative that’s directing more than €500 million toward building a drone supply chain in the West. Under the initiative, militaries run competitions for drone makers, rewarding high performers with contracts and sending their products to Ukraine. Its grantees are often not allowed to publicize their contracts, for security reasons. “But the companies which are delivering products through that initiative are new to the market,” Kipurs says. “They are not the companies that were there five years ago.”
Even national telecommunications company LMT, which is partly government owned, is working on drones and other military-grade hardware, including sensor equipment and surveillance balloons. It’s developing a battlefield “internet of things” system—essentially, a system that can track in real time all the assets and personnel in a theater of war. “In Latvia, more or less, we are getting ready for war,” says former naval officer Kaspars Pollaks, who heads an LMT division that focuses on defense innovation. “We are just taking the threat really seriously. Because we will be operationally alone [if Russia invades].”
The Latvian government’s investments are being mirrored across Europe: NATO has expanded its Defence Innovation Accelerator for the North Atlantic (DIANA) program, which runs startup incubators for dual-use technologies across the continent and the US, and launched a separate €1 billion startup fund in 2022. Adding to this, the European Investment Fund, a publicly owned investment company, launched a €175 million fund-of-funds this year to support defense technologies with dual-use potential. And the European Commission has earmarked more than €7 billion for defense research and development between now and 2027.
Private investors are also circling, looking for opportunities to profit from the boom. Figures from the European consultancy Dealroom show that fundraising by dual-use and military-tech companies on the continent was just shy of $1 billion in 2023—up nearly a third over 2022, despite an overall slowdown in venture capital activity.
Atlas Dynamics builds drones for industrial uses and now makes hardened versions that can resist electronic warfare and operate in battlefield conditions.
ATLAS AERO
When Atlas Dynamics started in 2015, funding was hard to come by, Tolchinsky says: “It’s always hard to make it as a hardware company, because VCs are more interested in software. And if you start talking about the defense market, people say, ‘Okay, it’s a long play for 10 or 20 years, it’s not interesting.’” That’s changed since 2022. “Now, what we see because of this war is more and more venture capital that wants to invest in defense companies,” Tolchinsky says.
But while money is helping startups get off the ground, to really prove the value of their products they need to get their tools in the hands of people who are going to use them. When I asked Kipurs if his products are currently being used in Ukraine, he only said: “I’m not allowed to answer that question directly. But our systems are with end users.”
Battle tested
Ukraine has moved on from the early days of the conflict, when it was willing to take almost anything that could be thrown at the invaders. But that experience has been critical in pushing the government to streamline its procurement processes dramatically to allow its soldiers to try out new defense-tech innovations.
Origin Robotics has built on a history of producing consumer drones to create a guided bomb that a soldier can carry in a backpack.
Technology that doesn’t work at the front puts soldiers at risk, so in many cases they have taken matters into their own hands. Two Ukrainian drone makers tell me that military procurement in the country has been effectively flipped on its head: If you want to sell your gear to the armed forces, you don’t go to the general staff—you go directly to the soldiers and put it in their hands. Once soldiers start asking their senior officers for your tool, you can go back to the bureaucrats and make a deal.
Many foreign companies have simply donated their products to Ukraine—partly out of a desire to help, and partly because they’ve identified a (potentially profitable) opportunity to expose them to the shortened innovation cycles of conflict and to get live feedback from those fighting. This can be surprisingly easy as some volunteer units handle their own parallel supply chains through crowdfunding and donations, and they are eager to try out new tools if someone is willing to give them freely. One logistics specialist supplying a front line unit, speaking anonymously as he’s not authorized to talk to the media, tells me that this spring, they turned to donated gear from startups in Europe and the US to fill gaps left by delayed US military aid, including untested prototypes of UAVs and communications equipment.
All of this has allowed many companies to bypass the traditionally slow process of testing and demonstrating their products, for better and worse.
Tech companies’ rush into the conflict zone has unnerved some observers, who are worried that by going to war, companies have sidestepped ethical and safety concerns over their tools. Clearview AI gave Ukraine access to its controversial facial recognition tools to help identify Russia’s war dead, for example, sparking moral and practical questions over accuracy, privacy, and human rights—publishing images of those killed in war is arguably a violation of the Geneva Convention. Some high-profile tech executives, including Palantir CEO Alex Karp and former Google CEO-turned-military-tech-investor Eric Schmidt, have used the conflict to try to shift the global norms for using artificial intelligence in war, building systems that let machines select targets for attacks—which some experts worry is a gateway into autonomous “killer robots.”
LMT’s Pollaks says he has visited Ukraine often since the war began. Though he declines to give more details, he euphemistically describes Ukraine’s wartime bureaucracy as “nonstandardized.” If you want to blow something up in front of an audience in the EU, he says, you have to go through a whole lot of approvals, and the paperwork can take months, even years. In Ukraine, plenty of people are willing to try out your tools.
“[Ukraine], unfortunately, is the best defense technology experimentation ground in the world right now,” Pollaks says. “If you are not in Ukraine, then you are not in the defense business.”
Jack Wang, principal at UK-based venture capital fund Project A, which invests in military-tech startups, agrees that the Ukraine “track” can be incredibly fruitful. “If you sell to Ukraine, you get faster product and tech iteration, and live field testing,” he says. “The dollars might vary. Sometimes zero, sometimes quite a bit. But you get your product in the field faster.”
The feedback that comes from the front is invaluable. Atlas Dynamics has opened an office in Ukraine, and its representatives there work with soldiers and special forces to refine and modify their products. When Russian forces started jamming a wide band of radio frequencies to disrupt communication with the drones, Atlas designed a smart frequency-hopping system, which scans for unjammed frequencies and switches control of the drone over to them, putting soldiers a step ahead of the enemy.
At Global Wolf, battlefield testing for the Mosphera has led to small but significant iterations of the product, which have come naturally as soldiers use it. One scooter-related problem on the front turned out to be resupplying soldiers in entrenched positions with ammunition. Just as urban scooters have become last-mile delivery solutions in cities, troops found that the Mosphera was well suited to shuttling small quantities of ammo at high speeds across rough ground or through forests. To make this job easier, Global Wolf tweaked the design of the vehicle’s optional extra trailer so that it perfectly fits eight NATO standard-sized bullet boxes.
Within weeks of Russia’s full-scale invasion, Mosphera scooters were at Ukraine’s front line—and even behind it, being used by Ukrainian special forces scouts.
GLOBAL WOLF
Some snipers prefer the electric Mosphera to noisy motorbikes or quads, using the vehicles to weave between trees to get into position. But they also like to shoot from the saddle—something they couldn’t do from the scooter’s footplate. So Global Wolf designed a stable seat that lets shooters fire without having to dismount. Some units wanted infrared lights, and the company has made those, too. These types of requests give the team ideas for new upgrades: “It’s like buying a car,” Asmanis says. “You can have it with air conditioning, without air conditioning, with heated seats.”
Being battle-tested is already proving to be a powerful marketing tool. Bukavs told me he thinks defense ministers are getting closer to moving from promises toward “action.” The Latvian police have bought a handful of Mospheras, and the country’s military has acquired some, too, for special forces units. (“We don’t have any information on how they’re using them,” Asmanis says. “It’s better we don’t ask,” Bukavs interjects.) Military distributors from several other countries have also approached them to market their units locally.
Although they say their donations were motivated first and foremost by a desire to help Ukraine resist the Russian invasion, Bukavs and Asmanis admit that they have been paid back for their philanthropy many times over.
Of course, all this could change soon, and the Ukraine “track” could very well be disrupted when Trump returns to office in January. The US has provided more than $64 billion worth of military aid to Ukraine since the start of the full-scale invasion. A significant amount of that has been spent in Europe, in what Wang calls a kind of “drop-shipping”—Ukraine asks for drones, for instance, and the US buys them from a company in Europe, which ships them directly to the war effort.
Wang showed me a recent pitch deck from one European military-tech startup. In assessing the potential budgets available for its products, it compares the Ukrainian budget, which was in the tens of millions of dollars, and the “donated from everybody else” budget, which was a billion dollars. A large amount of that “everybody else” money comes from the US.
If, as many analysts expect, the Trump administration dramatically reduces or entirely stops US military aid to Ukraine, these young companies focused on military tech and dual-use tech will likely take a hit. “Ideally, the European side will step up their spending on European companies, but there will be a short-term gap,” Wang says.
A lasting change?
Russia’s full-scale invasion exposed how significantly the military-industrial complex in Europe has withered since the Cold War. Across the continent, governments have cut back investments in hardware like ships, tanks, and shells, partly because of a belief that wars would be fought on smaller scales, and partly to trim their national budgets.
“After decades of Europe reducing its combat capability,” Pollaks says, “now we are in the situation we are in. [It] will be a real challenge to ramp it up. And the way to do that, at least from our point of view, is real close integration between industry and the armed forces.”
This would hardly be controversial in the US, where the military and the defense industry often work closely together to develop new systems. But in Europe, this kind of collaboration would be “a bit wild,” Pollaks says. Militaries tend to be more closed off, working mainly with large defense contractors, and European investors have tended to be more squeamish about backing companies whose products could end up going to war.
As a result, despite the many positive signs for the developers of military tech, progress in overhauling the broader supply chain has been slower than many people in the sector would like.
Several founders of dual-use and military-tech companies in Latvia and the other Baltic states tell me they are often invited to events where they pitch to enthusiastic audiences of policymakers, but they never see any major orders afterward. “I don’t think any amount of VC blogging or podcasting will change how the military actually procures technology,” says Project A’s Wang. Despite what’s happening next door, Ukraine’s neighbors are still ultimately operating in peacetime. Government budgets remain tight, and even if the bureaucracy has become more flexible, layers upon layers of red tape remain.
Soldiers of the Latvian National Defense Service learn field combat skills in a training exercise.
GATIS INDRēVICS/ LATVIAN MINISTRY OF DEFENSE
Even Global Wolf’s Bukavs laments that a caravan of political figures has visited their factory but has not rewarded the company with big contracts. Despite Ukraine’s requests for the Mosphera scooters, for instance, they ultimately weren’t included in Latvia’s 2024 package of military aid due to budgetary constraints.
What this suggests is that European governments have learned a partial lesson from Ukraine—that startups can give you an edge in conflict. But experts worry that the continent’s politics means it may still struggle to innovate at speed. Many Western European countries have built up substantial bureaucracies to protect their democracies from corruption or external influences. Authoritarian states aren’t so hamstrung, and they, too, have been watching the war in Ukraine closely. Russian forces are reportedly testing Chinese and Iranian drones at the front line. Even North Korea has its own drone program.
The solution isn’t necessarily to throw out the mechanisms for accountability that are part of democratic society. But the systems that have been built up for good governance have led to fragility, sometimes leading governments to worry more about the politics of procurement than preparing for crises, according to Ilves and other policy experts I spoke to.
“Procurement problems grow bigger and bigger when democratic societies lose trust in leadership,” says Ilves, who now advises Ukraine’s Ministry of Digital Transformation on cybersecurity policy and international cooperation. “If a Twitter [troll] starts to go after a defense procurement budget, he can start to shape policy.”
That makes it hard to give financial support to a tech company whose products you don’t need now, for example, but whose capabilities might be useful to have in an emergency—a kind of merchant marine for technology, on constant reserve in case it’s needed. “We can’t push European tech to keep innovating imaginative crisis solutions,” Ilves says. “Business is business. It works for money, not for ideas.”
Even in Riga the war can feel remote, despite the Ukrainian flags flying from windows and above government buildings. Conversations about ordnance delivery and electronic warfare held in airy warehouse conversions can feel academic, even faintly absurd. In one incubator hub I visited in April, a company building a heavy-duty tracked ATV worked next door to an accounting software startup. On the top floor, bean bag chairs were laid out and a karaoke machine had been set up for a party that evening.
A sense of crisis is needed to jolt politicians, companies, and societies into understanding that the front line can come to them, Ilves says: “That’s my take on why I think the Baltics are ahead. Unfortunately not because we are so smart, but because we have this sense of necessity.”
Nevertheless, she says her experience over the past few years suggests there’s cause for hope if, or when, danger breaks through a country’s borders. Before the full-scale invasion, Ukraine’s government wasn’t exactly popular among the domestic business and tech communities. “And yet, they came together and put their brains and resources behind [the war effort],” she says. “I have a feeling that our societies are sometimes better than we think.”
Google DeepMind has unveiled an AI model that’s better at predicting the weather than the current best systems. The new model, dubbed GenCast, is published in Nature today.
This is the second AI weather model that Google has launched in just the past few months. In July, it published details of NeuralGCM, a model that combined AI with physics-based methods like those used in existing forecasting tools. That model performed similarly to conventional methods but used less computing power.
GenCast is different, as it relies on AI methods alone. It works sort of like ChatGPT, but instead of predicting the next most likely word in a sentence, it produces the next most likely weather condition. In training, it starts with random parameters, or weights, and compares that prediction with real weather data. Over the course of training, GenCast’s parameters begin to align with the actual weather.
The model was trained on 40 years of weather data (1979 to 2018) and then generated a forecast for 2019. In its predictions, it was more accurate than the current best forecast, the Ensemble Forecast, ENS, 97% of the time, and it was better at predicting wind conditions and extreme weather like the path of tropical cyclones. Better wind prediction capability increases the viability of wind power, because it helps operators calculate when they should turn their turbines on and off. And better estimates for extreme weather can help in planning for natural disasters.
Google DeepMind isn’t the only big tech firm that is applying AI to weather forecasting. Nvidia released FourCastNet in 2022. And in 2023 Huawei developed its Pangu-Weather model, which trained on 39 years of data. It produces deterministic forecasts—those providing a single number rather than a range, like a prediction that tomorrow will have a temperature of 30 °F or 0.7 inches of rainfall.
GenCast differs from Pangu-Weather in that it produces probabilistic forecasts—likelihoods for various weather outcomes rather than precise predictions. For example, the forecast might be “There is a 40% chance of the temperature hitting a low of 30 °F” or “There is a 60% chance of 0.7 inches of rainfall tomorrow.” This type of analysis helps officials understand the likelihood of different weather events and plan accordingly.
These results don’t mean the end of conventional meteorology as a field. The model is trained on past weather conditions, and applying them to the far future may lead to inaccurate predictions for a changing and increasingly erratic climate.
GenCast is still reliant on a data set like ERA5, which is an hourly estimate of various atmospheric variables going back to 1940, says Aaron Hill, an assistant professor at the School of Meteorology at the University of Oklahoma, who was not involved in this research. “The backbone of ERA5 is a physics-based model,” he says.
In addition, there are many variables in our atmosphere that we don’t directly observe, so meteorologists use physics equations to figure out estimates. These estimates are combined with accessible observational data to feed into a model like GenCast, and new data will always be required. “A model that was trained up to 2018 will do worse in 2024 than a model trained up to 2023 will do in 2024,” says Ilan Price, researcher at DeepMind and one of the creators of GenCast.
In the future, DeepMind plans to test models directly using data such as wind or humidity readings to see how feasible it is to make predictions on observation data alone.
There are still many parts of forecasting that AI models still struggle with, like estimating conditions in the upper troposphere. And while the model may be good at predicting where a tropical cyclone may go, it underpredicts the intensity of cyclones, because there’s not enough intensity data in the model’s training.
The current hope is to have meteorologists working in tandem with GenCast. “There’s actual meteorological experts that are looking at the forecast, making judgment calls, and looking at additional data if they don’t trust a particular forecast,” says Price.
Hill agrees. “It’s the value of a human being able to put these pieces together that is significantly undervalued when we talk about AI prediction systems,” he says. “Human forecasters look at way more information, and they can distill that information to make really good forecasts.”