Connect with us

Technology & Innovation

What babies can teach AI

Diane Davis

Published

on

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

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!) 

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 good at certain tasks. They’re great at picking up and moving objects, for example, and they’re even getting better at cooking. But while robots may easily complete tasks like these in a laboratory, getting them to work in an unfamiliar environment where there’s little data available is a real challenge.

Technology & Innovation

The citizen scientists chronicling life at a neglected but vital Mexican river

Diane Davis

Published

on

""

Since the fall of 2021, Viaje al Microcosmos has organized walks on the river to encourage the general public to explore the space and make new connections with this misunderstood body of water. In the last few years, participants have created their own microscopes to study water samples they’ve collected; created an immersive art installation to showcase the species that live on the river; produced a podcast; developed a template to help people document water quality; contributed to the Río Santa Catarina’s Wikipedia page; and more. Each activity helps build a community of city dwellers who believe the river is a natural resource worth protecting. The effort seems only more urgent against the backdrop of a regional drought, rising temperatures, and a government that continues to place environmental issues on the back burner. 

A woman peeks into a microscope at Viaje al Microcosmos’s stand at the citizen science event.

LORENA RíOS

Viaje al Microcosmos’s goal is multifaceted and ambitious. The activists want to build a community of citizen scientists that pushes back against the belief that the river is for sale. They want to produce information about the river that can shape public policy and aid in its preservation. But perhaps most important, they want to instill the revolutionary thought that the Santa Catarina River can, in fact, be just that: a river. 

Lorena Ríos is a freelance journalist based in Monterrey, Mexico.

Continue Reading

Technology & Innovation

What Luddites can teach us about resisting an automated future

Diane Davis

Published

on

A story in comic format. In this first panel, two figures in silhouette look out at a modern city skyline.  The text reads, "The future is here already. AI art has arrived. Simply write a prompt..."

A person's smiling headshot being uploaded.  The text reads, "Or allow a company unrestricted access to your likeness..."
The headshot from the previous panel with distroted features and a wavy new background. The text reads, "...And ta-da! Some absolutely serviceable hotel lobby art. But calling this 'AI art' uncritically buys into the AI hype machine."
Two people look at the blank space where the framed picture of a flower has been stolen by a giant robot hand. The text reads, "Algorithmically generated art theft" might be more accurate.

Two panels. In the first, a scrabbly line resembling a signature. The text reads, "Machine-learning software is "taught" by feeding it existing art without consent, credit, or compensation. To the extent that mangled signature remnants remain visible on AI art." In the second, the text reads, "Artists have begun to push back." over the image of protestors led by Matthew Butterick, co-counsel in a class action lawsuit brought against AI art by artists, who says, "Everybody would creates for a living should be in code red."
The text reads, "A common refrain from defenders of AI art has been to label these naysayers:"  above multiple speech bubbles saying "Luddite!" "Luddites!" and "Luddism!"
Text across the top continues, "...a term synonymous with technophobia, anti-progress, and reactionism. It's even used to describe being hapless with new tech." A group of people eating dinner together, and one person is saying, "Oh, I can't use TikTok! I'm such a Luddite!" The below the image reads, "In truth, the Luddites were skilled with machines. They were simply fighting for better worker rights."

A man in a cravat holding a plume at his writing desk smiles, while behind in shadow are two women at textile machines. The text reads, "In 1799, the British government passed legislation that prohibited trader unions and collective bargaining. Mill owners introduced more machines in their factories, reducing wages."
A man with his shirtsleeves rolled up, raising a sledgehammer.  The text reads, "And so, in 1811, after years of frustrating negotiations, a spate of coordinated attacks on mill frameworks erupted across the United Kingdom, led by "King Ned Ludd."

Sheets of paper flutter down to a gallows in the foreground as soldiers with bayonets raised corral a crowd of civilians behind.  The text reads, "But them, 14000 soldiers were sent to protect the mills and quash riots. New legislation made frame breaking a capital offense. Eventually, key Luddite organizers were identified, arrested and executed.
Lord Byron reciting from "Song for the Luddites" says with his eyes closed, "This is the dew, which the tree shall renew Of Liberty, planted by Ludd!"  The text above reads, "Yet the movement lived on. Outbreaks of machine breaking continued for years in the UK and beyond.
A woman smashing something with a sledgehammer. The text reads, "In 1837, a spinning jenny in Chalabre, France was destroyed by workers. The women workers 'made themselves conspicuous by their fury and violence.' From local newspaper, L'Aude."

The text reads, "In the 1870s, English textile designer, author, and socialist William Morris began taking an interest in the production process of his designs. He was disgusted by the poor living conditions of workers and the pollution caused by the textile industry." William Morris in profile saying, "Why does a reasonable man use a machine? Surely to save his labour. Under capitalism, machines were primarily used to increase production, thereby increasing the worker's drudgery."
The text reads, "People have been making similar arguments against AI art." A tweet from CatBastardQuinn, @QuinnCat13 reads, "We could automate menial jobs so people have time to make art and music but apparently we'd rather automate art and music so people have time for menial jobs."
An office worker at a laptop with face in hands and an Amazon worker in a pile of shipping boxes. The text reads" Likewise, Morris was interested in what he called 'worthy work'.' He wanted people to take pleasure in their work rather than 'mere toiling to live that we may live to toil.'"
The text reads, "He understood that machines were only as progressive as the people who used them." Morris facing us from a field of app notifications, says "[An unfree man is a] slave to machinery; the new machine MUST be invented, and when invented he MUST - I will not say use it, but be used by it whether he likes it or not."
Frederick Taylor overseeing factory workers standing around machines with a stopwatch.  The text reads, "In America, mechanical engineer Frederick W. Taylor began using what he called the 'Scientific Management' in his factories. Taylor timed each worker's every movement, breaking down their work into a set of discrete tasks. Then demanded workers speed them up."
The text reads, "Taylorism, as it became known, was less a science than a political ideology concerned with remolding workers into pliant subjects. Taylor faces us to say, "A complete mental revolution on the part of the workingman...toward their work, toward their fellow men, and toward their employers."
The text reads, "But the spirit of Luddism lived on. Seemingly out of nowhere, there was a rash of mechanical breakdowns at Taylor's factories." An angry Taylor tells us, "These men were deliberately breaking their machines." while two men stand behind him shrugging and smiling slightly while a plume a smoke rises from an unseen area.

A cover of a  book entitled, "Sabotage: Its History, Philosophy & Function by author Walker C Smith. shows a black cat climbing atop a bag of money.  The text reads, "Despite setbacks and intervention from Congress, Taylorism spread. The Industrial Workers of the World responded by publishing two tracts on the topic of sabotage in 1913. 'The aim is to hit the employer in his vital spot, his heart and soul, in other words, his pocketbook.;"
Hands reach to an assembly line where the words "Automation - as it was later dubbed by Delmar Harder, vice president of Ford Motor Company, in 1947 - continued unabated." repeat on each panel of the conveyor.

The text reads, "The legacy of the Luddites lived on in the latter half of the 20th century. Confined to the lowest-paying jobs, black workers were the first to be targeted by the midcentury push for automation.  Robert L. Allen quotes from his book, Black Awakening in Capitalist America (1969) "Not only is the economic situation of the masses of blacks grim, but the prospects are that it will not improve, rather it will deteriorate. This is due partly to the unregulated impact of automation."
The text reads, "The Black Panthers were also quick to recognize <a href=technology and automation were not politically neutral." Huey Newton, cofounder of the Black Panther Party says from a podium, "If the ruling circle remains in power it seems to me that capitalists will continue to develop their technological machinery because they are not interested in the people…Every worker is in jeopardy because of the ruling circle."” class=”wp-image-1088535″ />
A group of Black Panthers standing together.  The text reads, "The Black Panther Party's Ten Point Program was updated in 1972. 'People's community control of modern <a href=technology‘ was added to the demands for ‘land, bread, housing, education, clothing, justice, peace."” class=”wp-image-1088536″ />

A conveyor belt with a chart, titled "Overrepresentation of African Americans in 3 occupation categories with the highest expected displacement, %" The statistics are Office Support e.g. secretaries with a displacement rate of 36%, Food Services, e.g. fast-food cooks with a displacement rate of 35% and Production work e.g. machinists with a displacement rate of 34%. The text reads, "A 2019 McKinsey Global Intitute report found that: 'African Americans are overrepresented in occupations likely to be the most affected by automation.' And the 2030 outlook doesn't look much better."
A crowd of students holding a sign that says "STRIKE" and a button that says, "I am a human being: do not fold, spindle, or mutilate" The text reads, "The introduction of the computer in the 1970s only exacerbated growing tensions. Punch cards used in universities, similar to cards used for drafting recruits for Vietnam, were seen as a symbol of bureaucracy and alienation. Students burned, vandalized, and otherwise destroyed punch cards for course registration."
Lisa Gitelman, author of Always Already New: Media, History and the Data of Culture (2008) says, "Many people simply stopped drawing distinctions between one card-enabled system and another. Whether the cards registered draftees or pupils, they helped 'the system.""
Men in a warroom facing a map. The text reads, "Vietnam had become the first computational war, bringing a shift toward strategies rooted in quantitative data collection and automated analysis."

A fighter plane with the name, "Tom Cat" on the side. The text reads, "Sensor arrays and unmanned drones became a big part of the war. An ongoing experiment with the intention of the eventual replacement of human pilots. American soldiers were sabotaging equipment, staging protests and refusing to fight. Automation of war as with the automation of industry, was a political strategy designed to, once again, reassert control over rebellious workers."
A person at the self-checkout frustrated by the machines error message which says, "Unexpected item in the bagging area!"  The text reads "Now, as more of our daily lives are automated, people are finding that it doesn't always make our lives easier."
The text reads, "In 2016, a study revealed that physicians spend two hours of computer work for every hour spent speaking with a patient face to face." Atul Gawande, surgeon and public health researcher tells us, "I've come to feel that a system that promised to increase my mastery over my work has, instead, increased my work's mastery over me."

The text above people using computers reads, "In the rush to claim success with new AI software, this invisible human work becomes even more insidious. Researcher Jathan Sadowski calls it 'Potemkin AI.'  Sadowski says to us, "There is a long list of services that purport to be powered by sophisticated software, but actually rely on humans acting like robots."
Writer and filmmaker Astra Taylor calls these modern-day Mechanical Turks an example of "fauxtomation" which she says "reinforces the idea that work has no value if it is unpaid and acclimates us to the idea that one day we won't be needed."
A sad person hold s a robot mask to their face. The text reads, "As a 2019 report by the think tank Data & Society concludes, 'Automated and AI technologies tend to mask the human labour that allows them to be fully integrated into a social context while profoundly changing the conditions and quality of labour that is at stake."

The text reads, "Opposition to 21st century tech can be found in unlikely places. Silicon Valley executives are restricting their own children's screen time and sending them to tech-free schools." Taewoo Kim, chief AI engineer at machine learning startup One Smart Lab, tells us from in front of a Waldorf school, "You can't put your face in a device and expect to develop a long-term attention span."
A bullet shaped robot outside near grass has sauce on it. The text reads, "In San Francisco, security robots sent to harass the homeless have been repeatedly assaulted, with one being covered in BBQ sauce and wrapped in tarp."
An autonomous vehicle with a broken window.  The text reads, "In Arizona, people are slashing the tires of driverless cars after Uber struck and killed a woman in Tempe." A quote from Douglas Rushkoff, author of Team Human (2019) reads, "People are lashing out justifiably. There's a growing sense that the giant corporations honing driverless technologies do not have our best interests at heart."
A pie chart with the words, "A pew research poll found that 85% of Americans favored the restriction of automation to only the most dangerous forms of work."

People standing around and looking at their cell phones. The text reads, "We're all living through an era in which we have become the product."  Shoshana Zuboff, author of The Age of Surveillance Capitalism (2018) tells us, "The age of surveillance capitalism is a titanic struggle between capital and each one of us. It is a direct intervention into free will, an assault on human autonomy. To tune and herd and shape and push us in the direction that creates the highest probability of their business success. [There's no way] to dress this up as anything but behavioral modification."
A facial recognition target over the face of a Black person next to a hand pointing at the word "Error." The text reads, "Many are also drawing attention to the biases of software made by an almost entirely male, predominantly white workforce."  A quote from Timnit Gebru, cofounder of Black in AI says, "I'm worried about groupthink, insularity, and arrogance in the AI community, [...] If many are actively excluded from its creation, this <a href=technology will benefit a few while harming a great many."” class=”wp-image-1088552″ />
The text reads, "VPNs, the dark web, and plugins like RequestPolicy are arguably Luddite responses to new <a href=technology. Computer science students have already developed Glaze- a tool to prevent AI models from mimicking artist styles." Maxigas, in Resistance to the Current: The Dialectics of Hacking (2022) says, "A retrograde attempt to rewind web history: a Luddite machine that, as they say ‘breaks’ the essential mechanisms of websites."” class=”wp-image-1088553″ />

Gavin Mueller in Breaking Things at Work (2021) says, "Luddism contains a critical perspective on <a href=technology that pays particular attention to technology‘s relationship to the labor process and working conditions. In other words, it views technology not as neutral but as a site of struggle. Luddism rejects production for production’s sake: it is critical of ‘efficiency’ as an end goal."” class=”wp-image-1088554″ />
Above a person tearing a line chart, the text reads, "'Degrowth,' 'slow living,' 'quiet quitting,' and the 'I Do Not Dream of Labour' movements could all be described as forms of modern neo-Luddism."  Jason Hickel, author of Less is More: How Degrowth will Save the World (2020)  says, "Lashed to the growth imperative, <a href=technology is used not to do the same amount of stuff in less time, but rather to do more stuff in the same amount of time."” class=”wp-image-1088555″ />

A man with a sledgehammer resting on his shoulder as hi looks out at the horizon. The text reads, "Questioning and resisting the worst excesses of <a href=technology isn’t antithetical to progress. If your concept of ‘progress’ doesn’t put people at the center of it, is it even progress? Maybe those of us who are apprehensive about AI art are Luddites. Maybe we should wear that badge with pride. Welcome to the future. Sabotage it."” class=”wp-image-1088556″ />

Tom Humberstone is a comic artist and illustrator based in Edinburgh, Scotland.

Continue Reading

Technology & Innovation

China’s next cultural export could be TikTok-style short soap operas

Diane Davis

Published

on

Poster of the short drama "Mr. Williams! Madame Is Dying," showing the two protaganists.

Web novels are a unique form of literature that has been popular on the Chinese internet for much of the last two decades: long stories that are written and posted chapter by chapter every day. Each chapter can be read in less than 10 minutes, but installments will keep being added for months if not years. Readers become avid fans, waiting for the new chapter to come out every day and paying a few cents to access it.

While some talented Chinese book authors got their big break by writing web novels, the majority of these works are the popcorn of literature, offering daily bite-size dopamine hits. For a while in the 2010s, some found an audience overseas too, with Chinese companies setting up websites to translate web novels into English.

But in the age of TikTok, long text posts have become less popular online, and the web-novel industry is looking to pivot. Business executives have realized they can adapt these novels into super-short dramas. Both forms aim for the same market: people who want something quick to kill time in their commute, or during breaks and lunch.

Many of the leading Chinese short-drama apps today work closely with Chinese web-novel companies. ReelShort is partially owned by COL Group, one of the largest digital publishers in China, with a treasure trove of novels that are ready for adaptation.

Poster of the short drama Mr. Williams! Madame Is Dying.

COURTESY OF FLEXTV

To get a quick sense of what these stories are like, you just need to take a look at their titles: President’s Sexy Wife, The Bride of the Wolf King, Boss Behind the Scenes Is My Husband, or The New Rich Family Grudge.

One of the highest-grossing shows on FlexTV is called Mr. Williams! Madame Is Dying. It’s a corny romance story about a love triangle, ultra-rich families, cancer, rebirth, and redemption, and it was adapted from a Chinese web novel that has nearly 1,300 chapters. The original story has been turned into a Chinese short drama, but FlexTV decided to shoot another version in Los Angeles for an international audience.

These short dramas prioritize quick, oversimplified stories of love, wealth, betrayal, and revenge, sometimes featuring mythical creatures like vampires and werewolves. Stories of marrying into a rich family attract men, while stories with a powerful female protagonist in control of her life appeal to women, says Gao, the COO of FlexTV. 

“Quibi mostly served the [artistic] pursuits of directors and producers. They thought their tastes were better than the general public and their work was to be appreciated by the elites,” he says, “What we are making is more like fast-moving consumer goods. It’s rooted in the needs of ordinary users.”

Continue Reading

Trending