Connect with us

This story originally appeared in The Algorithm, our weekly newsletter on AI. To get stories like this in your inbox first, sign up here

Human babies are fascinating creatures. Despite being completely dependent on their parents for a long time, they can do some amazing stuff. Babies have an innate understanding of the physics of our world and can learn new concepts and languages quickly, even with limited information. Even the most powerful AI systems we have today lack those abilities. Language models that power systems like ChatGPT, for example, are great at predicting the next word in a sentence but don’t have anything even close to the common sense of a toddler.  

But what if an AI could learn like a baby? AI models are trained on vast data sets consisting of billions of data points. Researchers at New York University wanted to see what such models could do when they were trained on a much smaller data set: the sights and sounds experienced by a single child learning to talk. To their surprise, their AI learned a lot—thanks to a curious baby called Sam.

The researchers strapped a camera on Sam’s head, and he wore it off and on for one and a half years, from the time he was six months old until a little after his second birthday. The material he collected allowed the researchers to teach a neural network to match words to the objects they represent, reports Cassandra Willyard in this story. (Worth clicking just for the incredibly cute pictures!) 

closeup of a smiling baby wearing a helmet camera with the bars of a crib in the background
WAI KEEN VONG

This research is just one example of how babies could take us a step closer to teaching computers to learn like humans—and ultimately build AI systems that are as intelligent as we are. Babies have inspired researchers for years. They are keen observers and excellent learners. Babies also learn through trial and error, and humans keep getting smarter as we learn more about the world. Developmental psychologists say that babies have an intuitive sense of what will happen next. For example, they know that a ball exists even though it is hidden from view, that the ball is solid and won’t suddenly change form, and that it rolls away in a continuous path and can’t suddenly teleport elsewhere.

Researchers at Google DeepMind tried to teach an AI system to have that same sense of “intuitive physics” by training a model that learns how things move by focusing on objects in videos instead of individual pixels. They trained the model on hundreds of thousands of videos to learn how an object behaves. If babies are surprised by something like a ball suddenly flying out of the window, the theory goes, it is because the object is moving in a way that violates the baby’s understanding of physics. The researchers at Google DeepMind managed to get their AI system, too, to show “surprise” when an object moved differently from the way it had learned that objects move.

Yann LeCun, a Turing Prize winner and Meta’s chief AI scientist, has argued that teaching AI systems to observe like children might be the way forward to more intelligent systems. He says humans have a simulation of the world, or a “world model,” in our brains, allowing us to know intuitively that the world is three-dimensional and that objects don’t actually disappear when they go out of view. It lets us predict where a bouncing ball or a speeding bike will be in a few seconds’ time. He’s busy building entirely new architectures for AI that take inspiration from how humans learn. We covered his big bet for the future of AI here.

The AI systems of today excel at narrow tasks, such as playing chess or generating text that sounds like something written by a human. But compared with the human brain—the most powerful machine we know of—these systems are brittle. They lack the sort of common sense that would allow them to operate seamlessly in a messy world, do more sophisticated reasoning, and be more helpful to humans. Studying how babies learn could help us unlock those abilities. 

Deeper Learning

This robot can tidy a room without any help

Robots are

Read More

————

By: Melissa Heikkilä
Title: What babies can teach AI
Sourced From: www.technologyreview.com/2024/02/06/1087793/what-babies-can-teach-ai/
Published Date: Tue, 06 Feb 2024 10:24:15 +0000

Continue Reading

Tech

The Download: hunting for new matter, and Gary Marcus’ AI critiques

1f6f0

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.

Inside the hunt for new physics at the world’s largest particle collider

In 2012, using data from CERN’s Large Hadron Collider, researchers discovered a particle called the Higgs boson. In the process, they answered a nagging question: Where do fundamental particles, such as the ones that make up all the protons and neutrons in our bodies, get their mass?

When the particle was finally found, scientists celebrated with champagne. A Nobel for two of the physicists who predicted the Higgs boson soon followed.

But now, more than a decade later, there is a sense of unease. That’s because there are still so many unanswered questions about the fundamental constituents of the universe.

So researchers are trying something new. They are repurposing detectors to search for unusual-looking particles, squeezing what they can out of the data with machine learning, and planning for entirely new kinds of colliders. Read the full story.

—Dan Garisto

This story is from the upcoming print issue of MIT Technology Review, dedicated to exploring hidden worlds. Want to get your hands on a copy when it publishes next Wednesday? Subscribe now.

I went for a walk with Gary Marcus, AI’s loudest critic

Gary Marcus, a professor emeritus at NYU, is a prominent AI researcher and cognitive scientist who has positioned himself as a vocal critic of deep learning and AI. He is a divisive figure, and can often be found engaged in spats on social media with AI heavyweights such as Yann LeCun and Geoffrey Hinton (“All attempts to socialize me have failed,” he jokes.)

Marcus does much of his tweeting on scenic walks around his hometown of Vancouver. Our senior AI reporter Melissa Heikkilä decided to join him on one such stroll while she was visiting the city, to hear his thoughts on the latest product releases and goings-on in AI. Here’s what he had to say to her.

This story is from The Algorithm, our weekly newsletter all about AI. Sign up to receive it in your inbox every Monday.

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 A new class of satellites could change everything
🛰
They’re armed with cameras powerful enough to capture peoples’ individual features. (NYT $)
A big European satellite is set to return to Earth tomorrow. (Ars Technica)
A new satellite will use Google’s AI to map methane leaks from space. (MIT Technology Review)

2 How much electricity does AI consume?
It’s a lot—but working out exact sums can be tricky. (The Verge)
Making an image with generative AI uses as much energy as charging your phone. (MIT Technology Review)

3 How Silicon Valley learned to love the military
The world is feeling like a more dangerous place these days, and that’s drowning out any ethical concerns. (WP $)
Why business is booming for military AI startups. (MIT Technology Review)
+ SpaceX is getting closer to US intelligence and military agencies. (WSJ $)
Ukraine is in desperate need of better methods to clear land mines. (Wired $)

4 The EU is investigating TikTok over child safety
It alleges the company isn’t doing enough to verify users’ ages. (Mashable)

5 It’s hard to get all that excited about Bluesky
It’s just more of the same social media. (Wired $)
How to fix the internet. (MIT Technology Review)
Why millions of people are flocking to decentralized social media services. (MIT Technology Review)

6 Ozempic is taking off in China
A lack of official approval there yet isn’t stopping anyone. (WSJ $)
We’ve never understood how hunger works. That might be about to change. (MIT Technology Review)

7 Meet the people trying to make ethical AI porn
Sex work is a sector that’s already being heavily disrupted by AI. (The Guardian)

8 Why we need DNA data drives
We’re rapidly running out of storage space, but DNA is a surprisingly viable option. (IEEE Spectrum)

9 You don’t need to keep closing your phone’s background apps
It does nothing for your battery life. In fact, it can even drain it further. (Gizmodo)
Here’s another myth worth busting: you shouldn’t put your wet phone in rice. (The Verge)

10 The mysterious math of billiard tables
If you struggle to play pool, take comfort in the fact mathematicians get stumped by it too. (Quanta $)

Quote of the day

“We realized how easy it is for people to be against something, to reject something new.” 

—Silas Heineken, a 17-year-old from Grünheide, a suburb near Berlin in Germany, tells the New York

Read More

————

By: Charlotte Jee
Title: The Download: hunting for new matter, and Gary Marcus’ AI critiques
Sourced From: www.technologyreview.com/2024/02/20/1088705/hunting-new-matter-gary-marcus/
Published Date: Tue, 20 Feb 2024 13:10:00 +0000

Continue Reading

Tech

The search for extraterrestrial life is targeting Jupiter’s icy moon Europa

PIA25496 Full scaled

>

We’ve known of Europa’s existence for more than four centuries, but for most of that time, Jupiter’s fourth-largest moon was just a pinprick of light in our telescopes—a bright and curious companion to the solar system’s resident giant. Over the last few decades, however, as astronomers have scrutinized it through telescopes and six spacecraft have flown nearby, a new picture has come into focus. Europa is nothing like our moon. 

Observations suggest that its heart is a ball of metal and rock, surrounded by a vast saltwater ocean that contains more than twice as much water as is found on Earth. That massive sea is encased in a smooth but fractured blanket of cracked ice, one that seems to occasionally break open and spew watery plumes into the moon’s thin atmosphere.

For these reasons, Europa has captivated planetary scientists interested in the geophysics of alien worlds. All that water and energy—and hints of elements essential for building organic molecules —point to another extraordinary possibility. In the depths of its ocean, or perhaps crowded in subsurface lakes or below icy surface vents, Jupiter’s big, bright moon could host life.

“We think there’s an ocean there, everywhere,” says Bob Pappalardo, a planetary scientist at NASA’s Jet Propulsion Laboratory in Pasadena, California. “Essentially everywhere on Earth that there’s water, there’s life. Could there be life on Europa?”

Pappalardo has been at the forefront of efforts to send a craft to Europa for more than two decades. Now his hope is finally coming to fruition: later this year, NASA plans to launch Europa Clipper, the largest-­ever craft designed to visit another planet. The $5 billion mission, scheduled to reach Jupiter in 2030, will spend four years analyzing this moon to determine whether it could support life. It will be joined after two years by the European Space Agency’s Juice, which launched last year and is similarly designed to look for habitable conditions, not only on Europa but also on other mysterious Jovian moons.

Neither mission will beam back a definitive answer to the question of extraterrestrial life. “Unless we get really lucky, we’re not going to be able to tell if there is life there, but we can find out if all the conditions are right for life,” says planetary geologist Louise Prockter at the Johns Hopkins Applied Physics Laboratory, a co-­investigator on the Clipper camera team.

“Essentially everywhere on Earth that there’s water, there’s life. Could there be life on Europa?”

Bob Pappalardo, planetary scientist, NASA’s Jet Propulsion Laboratory

What these spacecraft will do is get us closer than ever before to answers, by identifying the telltale chemical, physical, and geological signatures of habitability—whether a place is a suitable environment for life to emerge and thrive.

The payoff for confirming these signs on Europa would be huge. Not because humans could settle on its surface—it’s far too harsh and rugged and cold and irradiated for our delicate bodies—but because it could justify future exploration to land there and look for alien life-forms. Finding something, anything, living on Europa would offer strong evidence for an alternate path through which life could emerge. It would mean that life on Earth is not exceptional. We’d know that we have neighbors close by—even if they’re microbial, which would be the most likely life-form—and that would make it very likely that we have neighbors elsewhere in the cosmos.

PIA25496 Full 1 scaled
Engineers and technicians install reaction wheels on Europa Clipper at NASA’s Jet Propulsion Laboratory in CaliforniaNASA/JPL-CALTECH

“With the prospects of life—the prospects of vast oceans—within reach, you just have to go,” says Nicholas Makris, director of MIT’s Center for Ocean Engineering, who uses acoustics and other innovative methods to observe and explore big bodies of water. He once led a team of scientists who proposed a mission to land a spacecraft on Europa and use sound waves to explore what lies beneath the ice; he still hopes to see a lander go there one day. “You have

Read More

————

By: Stephen Ornes
Title: The search for extraterrestrial life is targeting Jupiter’s icy moon Europa
Sourced From: www.technologyreview.com/2024/02/19/1087988/nasa-europa-clipper-mission-jupiter-extraterrestrial-life/
Published Date: Mon, 19 Feb 2024 10:00:00 +0000

Continue Reading

Tech

The Download: missions to Jupiter’s moon Europa, and Uruguay’s screwworm gene drive

1f933

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 search for extraterrestrial life is targeting Jupiter’s icy moon Europa

Europa, Jupiter’s fourth-largest moon, is nothing like ours. Its surface is a vast saltwater ocean, encased in a blanket of cracked ice, one that seems to occasionally break open and spew watery plumes into the moon’s thin atmosphere.

For these reasons, Europa captivates planetary scientists. All that water and energy—and hints of elements essential for building organic molecules —point to another extraordinary possibility. Jupiter’s big, bright moon could host life.

And they may eventually get some answers. Later this year, NASA plans to launch Europa Clipper, the largest-­ever craft designed to visit another planet. The $5 billion mission, scheduled to reach Jupiter in 2030, will spend four years analyzing this moon to determine whether it could support life. Read the full story.

—Stephen Ornes

This story is from the upcoming print issue of MIT Technology Review, dedicated to exploring hidden worlds. Buy a subscription to get your hands on a copy when it publishes on February 28th! Deals start at just $8 a month

Uruguay wants to use gene drives to eradicate devastating screwworms

The New World screwworm, a parasite common in parts of South America and the Caribbean, is a disaster for cattle. It burrows into their flesh, eventually killing them. In Uruguay alone, it costs farmers between $40 million and $154 million a year. However, work is underway to fight back.

A group of researchers in Montevideo Uruguay have used the gene-editing system CRISPR to develop what’s known as a gene drive: tweaks to the screwworms genes that, if they spread, will cause a population crash.

They are about to move into the next stage of caged trials in the lab, with a view to eventually using the genetic tool to decimate the screwworm fly population. Read the full story.

—Abdullahi Tsanni

The must-reads

I’ve combed the internet to find you today’s most fun/important/scary/fascinating stories about technology.

1 The White House will spend $1.5 billion on a new chip factory
The money will go to semiconductor giant GlobalFoundries to produce advanced chips not currently made in the US. (WP $)
But work to expand US chip manufacturing keeps being plagued with problems. (NYT $)
Can the massive infusions of money rebuild the US’s industrial base? (MIT Technology Review)

2 Apple is facing its first EU fine
The EU says its music streaming services violate antitrust law. (FT $)

3 A judge ruled that Air Canada had to honor its chatbots’ discount error
This sets an important precedent as companies start to adopt AI tools. (WP $)
Judges, not politicians, are starting to dictate AI rules. (MIT Technology Review)
Should you trust an AI chatbot to plan a trip for you? (The Atlantic $)
It’s surprisingly tricky to work out when and how we’ll use generative AI. (FT $)

4 Is AI going to change how we define videos?
Systems like OpenAI’s Sora don’t make recordings. They render ideas. (New Yorker $)
Sora looks amazing—but the rest of us will have to wait to try it out. (MIT Technology Review)

5 Don’t blindly trust Google search results
AI-generated content, adverts, and ranking algorithms are really starting to spoil searches. (WSJ $)

6 The days of fast, free shipping may be coming to an end
Blame interest-rate hikes, and growing impatience from startup investors. (Insider $)

7 How New York’s legal weed revolution got derailed
The state’s plans ended up in an unholy mess. (New Yorker $)
The feud between a weed influencer and scientist over puking stoners. (MIT Technology Review)

8 A gun influencer’s conviction has done nothing to dent his popularity
In fact, YouTube is still running adverts on his channel. (NBC)
Hated that video? YouTube’s algorithm might push you another just like it. (MIT Technology Review)

9 Phone cases are getting jazzed up
🤳
They can do so much more than just protect your phone—for example, holding your lip balm. (Wired $)
Sharp-cornered smartphone cases are all the rage too. (WSJ $)

10 3D-printed chocolate sounds delicious
🍫
It’s something to do with the ridges and textures. (The Verge)

Quote of the day

“Everyone is looking around, talking about when layoffs are coming next, at what company.”

—A tech worker tells Insider that no job in the industry feels

Read More

————

By: Charlotte Jee
Title: The Download: missions to Jupiter’s moon Europa, and Uruguay’s screwworm gene drive
Sourced From: www.technologyreview.com/2024/02/19/1088699/jupiter-europa-uruguay-screwworm-gene-drive/
Published Date: Mon, 19 Feb 2024 13:10:00 +0000

Continue Reading

Trending