Danijar Hafner's office in San Francisco's SoMa district is mostly empty. His new startup is still in stealth mode, and it doesn't even have its name on the door. When a reporter visited earlier this year, only one other person was there, and there was hardly any furniture. But what the office lacks in decorations, it makes up for in robots. Human-like machines of various shapes and sizes hang like puppets from racks that run down the center of the wide-open space.

Hafner, 31, won't say much about his new company yet, but he describes it as a continuation of his long-time work to help AI navigate environments it hasn't seen in training. The humanoids, which he imports from China, are the next step in this work and its physical form. Their ability to react in new situations will be key to getting robots into human spaces. For example, if you want to send a robot into someone's home, it needs to handle a floor plan and furniture it's never seen before.

To achieve this, Hafner uses something called model-based reinforcement learning. He develops world models, which are AI models designed to imitate physical reality, and trains agents within them. The agent treats the model as a real-world simulation and learns how to act there. It then uses those experiences to make predictions (or to dream or imagine, as Hafner might say) about future outcomes. That allows agents, or the robots they're embedded in, to navigate unfamiliar situations in real life.

Unlike other efforts, Hafner's technique enables agents and the robots they control to perform very complicated tasks without the real-world trial-and-error training that's traditionally been used in robotics.

Hafner grew up in a rural town in northeastern Germany, where his parents were both classical musicians. He learned programming from a neighbor, and in high school he began taking online courses about AI, which quickly became a passion. "I was always fascinated with how thinking works," he says. AI offered him a way to emulate it on a computer.

In 2015, as a second-year undergraduate studying engineering at Hasso Plattner Institute in Potsdam, he won a role as a student researcher at Google Brain. From there, he went on to a dozen internships and other positions at the company, including stints with Google Brain and Google DeepMind (the two have since merged under DeepMind) in the UK, Canada, and the US. He worked with industry legends including Geoffrey Hinton, who is often called one of the godfathers of AI, and Ashish Vaswani, coauthor of the groundbreaking research paper "Attention Is All You Need," which described the transformer technology used by today's large language models.

One of Hafner's former managers and coauthors at Google, Timothy Lillicrap, describes him as a standout among standouts. "I get to interact with a lot of really smart people in research at Google, and he easily sits in the top half of 1%," Lillicrap says. "In many cases he would build, single-handedly, things it would take entire teams of engineers to build."

Over the years, Hafner has honed and proved his approach by pitting agents trained within his world models against popular video games. His first breakthrough was PlaNet, a model that allowed agents to execute actions by planning ahead. His Dreamer 2 was the first agent to hit human-level performance playing Atari 2600 games using a world model. Dreamer 3 was the first one to solve the Minecraft Diamond challenge, successfully mining in-game gems on its own. And Dreamer 4 went a step beyond that by learning to mine diamonds from an offline data set of recorded gameplay videos, without ever interacting with the game directly.

More recently, he's begun to migrate his agents out of the virtual world and into physical reality. His DayDreamer project used the Dreamer algorithm to let robots operate themselves in novel environments and react to new experiences (such as being pushed over) without any specific training.

Today, Hafner is working on his new startup, which he left Google DeepMind to form in the fall of 2025. Though he's coy about his next steps, it's clear he's dreaming big: "I was interested in solving a problem," he hints, "that would change the world."

Yuancheng (Ryan) Lu is fascinated by aging and eyes. When he walks out of the Whitehead Institute in Cambridge, Massachusetts, his aviator glasses automatically get darker in the sun. Age-related blindness runs in his family. A great-aunt in China was killed crossing a road because she couldn't see oncoming traffic. And Lu's own 23andMe test showed he has a mutation for macular degeneration, a leading cause of vision loss in old age. Bright sunlight is another risk factor—hence the shades. "They protect me," he says. "Plus, they look cool."

Lu, 34, works on gene therapies to prevent age-related vision loss. "I think the eye is a really unique system to study aging and rejuvenation," he says. "I could give a whole presentation." I push up my reading glasses and lean in to listen.

Lu is behind one of the most exciting results in rejuvenation science—and in eye research. In 2018, while getting his PhD at Harvard Medical School, he used an age-reversal technique called reprogramming to repair the optic nerves of mice. He crushed the nerves, blinding the animals, and then injected the cells with a gene therapy meant to restore them to a youthful state. Sixteen days later, the nerves were growing back, their axons showing up through a microscope as spidery orange filaments.

As hype around age reversal swirls, Lu has been busy in the lab searching for what he calls "the next generation of rejuvenation therapies."

The head of that lab, longevity scientist David Sinclair, remembers when Lu texted him the pictures: "He asked me, 'What do you see here?' And I said, 'I see the future.'" Later tests in a box with rotating bars of light showed the mice were tracking the changes. They could see again.

This year, nearly the exact genetic therapy Lu created for mice entered human clinical trials. On June 9, the startup Life Biosciences, which Sinclair cofounded and in which Lu owns a small stake, announced it had injected the treatment into the eye of a person with glaucoma. The trial has been big news. A headline in the New York Times suggested the technology could "change humanity." Posters on X gushed, with one declaring that "the fountain of youth is here."

"It's remarkable that what he developed as a student is now going into humans," says Sinclair of the treatment, now called ER-100. "It's barely even changed since he built it."

Reprogramming refers to an age-restoring process that takes place inside an embryo. It's why babies are born young, not old: The DNA they've inherited from their parents has been scrubbed and reset. In 2006, Japanese researchers showed they could cause the process to occur in the lab by introducing just four key genes, known by the acronym OSKM. Add these to a cell from a 100-year-old and it will turn into a stem cell that acts as if it was plucked from an embryo.

That's powerful stuff. But we don't want to turn people into blobs of stem-cell protoplasm. Lu figured out a way to control the effect. He trimmed the list of genes to just OSK—leaving out M, for Myc, the one most likely to cause dangerous changes like cancer. His extra flash of insight was that reprogramming could be tested on the optic nerve; the eye is particularly accessible.

Lu's result, published in Nature in 2020, helped set off an investment rush. Since then, US tech billionaires have placed huge bets on private companies like Altos Labs and NewLimit to explore reprogramming and anti-aging medicine. The day I spoke with Lu, he'd spent the morning meeting with the business magnate Zhong Shanshan, one of China's richest people.

Still, as hype around age reversal swirls, Lu has been notably absent from the public conversation. He's been busy in the lab searching for what he calls "the next generation of rejuvenation therapies." With a sigh, Lu describes the grueling effort over the last six years to understand what OSK really does. The treatment remains toxic to many cell types, and he says it's becoming obvious that different factors drive aging in each kind. This year, for example, he identified a gene responsible for protecting the retina from damage by free radicals—the main cause of age-related macular degeneration.

While Sinclair, his former boss, believes humans could live to be 200, Lu disagrees. There's just too much that goes wrong as we age. His work with OSK, he says, was more a proof of concept than a silver bullet. But it did change the conversation. "Six years ago, you couldn't talk about rejuvenation. We didn't use that word—there was pushback," Lu tells me. "But I think people have accepted the concept that you can really reverse molecular age."

Every year, MIT Technology Review picks some of the brightest young minds in science and technology. Their list, called 35 Innovators Under 35, highlights people whose work could shape the future. This year, nine of them are transforming biotech. Here are five of their stories, from lifesaving devices to groundbreaking age-reversal tech.

1. Preventing maternal deaths

Paschal Kija, 28, has created a device to treat postpartum hemorrhage—a dangerous bleeding complication after childbirth. In his home country of Tanzania, it causes about 29% of maternal deaths. His device, called Mkanda Salama (Swahili for "Safe Wrap"), is easy to use and costs only $70. A study found it stopped bleeding in 73% of women within 20 minutes.

2. Making brain electrodes inspired by Japanese art

For decades, scientists have implanted electrodes into the brain to study activity and treat neurological disorders. But these devices can damage brain tissue. Xiao Yang, 34, is developing ultra-small, flexible electrodes that look like actual neurons. She hopes they will cause less harm. She also creates sheets of electrodes to study brain cells in the lab. Inspired by kirigami—the Japanese art of cutting paper to form 3D shapes—she made a honeycombed sheet shaped like a spiral basket. She's already using it to study brain cells.

3. Developing an all-new treatment for baby KJ

In 2024, Kyle "KJ" Muldoon Jr. was born with a rare and potentially fatal genetic disorder. Sarah Grandinette, now 26, was part of a team that developed a personalized gene-editing therapy for him. The therapy was designed to correct a genetic "misspelling." Grandinette created cells with KJ's genetic variant, used them to screen gene-editing approaches, and tested potential medicines in mice and monkeys. KJ received his first dose at about seven months old. He responded well and was eventually discharged from the hospital. He's "doing pretty great," she says.

4. Reversing the aging process to treat eye disease

The hottest area in longevity research right now is reprogramming—trying to rewind the age of cells by resetting them to an embryonic-like state. In a 2020 study, Yuancheng (Ryan) Lu, now 34, and his colleagues showed that a reprogramming therapy reversed vision loss in aged, blind mice. Now, an almost identical therapy is being tested in people with eye disease. Life Biosciences, the company developing the drug, gave its first volunteer a dose in June.

5. Using AI to design new viruses

Last year, Samuel King used a generative AI model to design new genetic blueprints for bacteriophages—tiny viruses that infect bacteria. He printed the blueprints as strands of DNA. In experiments, the AI-designed viruses could copy themselves, burst out of bacterial cells, and infect other bacteria. Viruses aren't alive, but King, 27, hopes that AI-designed life forms might one day be used to make drugs or clean up pollution.

You can read more about these innovators and others on the biotech list here.

This article first appeared in The Checkup, MIT Technology Review's weekly biotech newsletter. To receive it in your inbox every Thursday, sign up here.

Every year, our reporters and editors argue about which new technologies will shape the future. Once a year, we make our best guesses and share them with you. Here are the advances we believe will drive progress or cause the biggest changes—for better or worse—in the coming years.

CLIMATE CHANGE AND ENERGY Sodium-ion batteries Sodium-ion batteries are made from common materials like salt. They are becoming a cheaper and safer alternative to lithium batteries. With support from big companies and public money, they are ready to power electricity grids and affordable electric vehicles around the world.

ARTIFICIAL INTELLIGENCE Generative coding AI coding tools are changing how we write, test, and launch code. They make it easier and faster to build complex websites, games, and other apps than ever before. But you should always double-check what the AI produces.

CLIMATE CHANGE AND ENERGY Next-gen nuclear Nuclear power already provides steady electricity to grids worldwide without producing greenhouse gases. New designs use different fuels and cooling systems, or take up less space. This could help more reactors start up faster.

DEEP DIVE How next-generation nuclear reactors break out of the 20th-century blueprint From molten salt to TRISO fuel, here’s how technological advancements could upend an old power technology. READ NOW

ARTIFICIAL INTELLIGENCE AI companions Every day, millions of people talk with AI chatbots. Some form what feel like close, personal bonds with the bots. There is growing evidence that this can be dangerous, and politicians are finally paying attention.

BIOTECHNOLOGY Base-edited baby When he was just seven months old, baby KJ became the first person to receive a personalized gene-editing treatment. A clinical trial is now planned, and custom gene-editing drugs could be approved within the next few years.

BIOTECHNOLOGY Gene resurrection Growing collections of gene information from extinct creatures are providing clues for new treatments and suggesting solutions to climate change—and may help save endangered species.

ARTIFICIAL INTELLIGENCE Mechanistic interpretability Nobody knows exactly how large language models work, which means we don’t have a clear idea of their limitations. But that’s changing: clever research techniques are giving us the best glimpse yet of what’s really going on inside the black box.

DEEP DIVE Meet the new biologists treating large language models like aliens By studying large language models as if they were living things instead of computer programs, scientists are discovering some of their secrets for the first time. READ NOW

SPACE Commercial space stations Space tourism may seem fanciful, but this might be the year paying customers can finally check into a room with a galactic view. The shiny new modules will also support research missions by scientists and space agencies.

BIOTECHNOLOGY Embryo scoring Screening embryos for genetic diseases is relatively common practice in fertility clinics today, and it can give potential parents some peace of mind. Now, though, new startups are making bold claims about using similar techniques to predict certain traits, including intelligence.

ARTIFICIAL INTELLIGENCE Hyperscale AI data centers The race for AI supremacy has supercharged data centers. Hyperscale AI data centers pack powerful computer chips into synchronized clusters that work like giant, high-speed supercomputers—sizzling hot, power-hungry behemoths pushing infrastructure to its limits.

Vote for the 11th breakthrough Humanoid robots 0% Reasoning models 0% Artificial wombs 0% Uncrewed fighter jets 0%

LEARN ABOUT THE WINNER Humanoid robots Machines designed in our image have sparked public fascination and substantial investments.

The gig workers who are training humanoids at home By Michelle Kim

10 Breakthrough Technologies ABOUT THE LIST MIT Technology Review’s annual list of 10 Breakthrough Technologies aims to help our audience know which emerging technologies are really worth paying attention to right now. Our editors and reporters pitch advances that they think should make the list, and we discuss and debate the merits of each over the course of several months.

We look for breakthroughs that will have a broad impact and make meaningful differences in our lives and work. We give equal consideration to those that we think will have a positive effect or bring negative consequences. Every year, we have to pass on many worthy candidates. Here’s what didn’t make the list this year.

This is the 25th year we’ve produced this list. Predicting the future is tough and we don’t always get it right. Look back through our archives to see what’s worked out—and what hasn’t. Or read this special analysis by an MIT researcher about some of the flops.

Explore the archives CREDITS Editorial Special projects editor: Amy Nordrum Editing: Rachel Courtland, Niall Firth, Mary Beth Griggs, Mat Honan, Charlotte Jee, Amy Nordrum, Adam Rogers, Amanda Silverman Copy editing: Linda Lowenthal, Teresa Elsey Engagement: Juliet Beauchamp, Abby Ivory-Ganja Fact checking: Ena Alvarado, Lauren Colella, Kelsey Kudak, Rachael Lallensack

Art Illustration: Vichhika Tep, Daniel Liévano, Stuart Bradford Art direction: Mariya Sitnova, Stephanie Arnett, Eric Mongeon

Design & Technology Lead developer: Andre Vitorio Design: Vichhika Tep Product management & analytics: Allison Chase Creative direction & product strategy: Mariya Sitnova CTO: Drake Martinet