Group items tagged

That allows quantum computers to make many calculations in one pass, while digital ones have to perform each calculation separately. By speeding up computation, quantum computers could potentially solve big, complex problems in fields like chemistry and materials science that are out of reach today.

When Google researchers made their supremacy claim in 2019, they said their quantum computer performed a calculation in 3 minutes 20 seconds that would take about 10,000 years on a state-of-the-art conventional supercomputer.

The IBM researchers in the new study performed a different task, one that interests physicists. They used a quantum processor with 127 qubits to simulate the behavior of 127 atom-scale bar magnets — tiny enough to be governed by the spooky rules of quantum mechanics — in a magnetic field. That is a simple system known as the Ising model, which is often used to study magnetism.

This problem is too complex for a precise answer to be calculated even on the largest, fastest supercomputers.

On the quantum computer, the calculation took less than a thousandth of a second to complete. Each quantum calculation was unreliable — fluctuations of quantum noise inevitably intrude and induce errors — but each calculation was quick, so it could be performed repeatedly.

Indeed, for many of the calculations, additional noise was deliberately added, making the answers even more unreliable. But by varying the amount of noise, the researchers could tease out the specific characteristics of the noise and its effects at each step of the calculation.“We can amplify the noise very precisely, and then we can rerun that same circuit,” said Abhinav Kandala, the manager of quantum capabilities and demonstrations at IBM Quantum and an author of the Nature paper. “And once we have results of these different noise levels, we can extrapolate back to what the result would have been in the absence of noise.”In essence, the researchers were able to subtract the effects of noise from the unreliable quantum calculations, a process they call error mitigation.

Altogether, the computer performed the calculation 600,000 times, converging on an answer for the overall magnetization produced by the 127 bar magnets.

Although an Ising model with 127 bar magnets is too big, with far too many possible configurations, to fit in a conventional computer, classical algorithms can produce approximate answers, a technique similar to how compression in JPEG images throws away less crucial data to reduce the size of the file while preserving most of the image’s details

Certain configurations of the Ising model can be solved exactly, and both the classical and quantum algorithms agreed on the simpler examples. For more complex but solvable instances, the quantum and classical algorithms produced different answers, and it was the quantum one that was correct.

Thus, for other cases where the quantum and classical calculations diverged and no exact solutions are known, “there is reason to believe that the quantum result is more accurate,”

Mr. Anand is currently trying to add a version of error mitigation for the classical algorithm, and it is possible that could match or surpass the performance of the quantum calculations.

In the long run, quantum scientists expect that a different approach, error correction, will be able to detect and correct calculation mistakes, and that will open the door for quantum computers to speed ahead for many uses.

Error correction is already used in conventional computers and data transmission to fix garbles. But for quantum computers, error correction is likely years away, requiring better processors able to process many more qubits

“This is one of the simplest natural science problems that exists,” Dr. Gambetta said. “So it’s a good one to start with. But now the question is, how do you generalize it and go to more interesting natural science problems?”

Those might include figuring out the properties of exotic materials, accelerating drug discovery and modeling fusion reactions.

Most A.I. researchers agree that models trained using R.L.H.F. are better behaved than models without it. But some argue that fine-tuning a language model this way doesn’t actually make the underlying model less weird and inscrutable. In their view, it’s just a flimsy, friendly mask that obscures the mysterious beast underneath.

In a nutshell, the joke was that in order to prevent A.I. language models from behaving in scary and dangerous ways, A.I. companies have had to train them to act polite and harmless. One popular way to do this is called “reinforcement learning from human feedback,” or R.L.H.F., a process that involves asking humans to score chatbot responses, and feeding those scores back into the A.I. model.

Shoggoths are fictional creatures, introduced by the science fiction author H.P. Lovecraft in his 1936 novella “At the Mountains of Madness.” In Lovecraft’s telling, Shoggoths were massive, blob-like monsters made out of iridescent black goo, covered in tentacles and eyes.

@TetraspaceWest said, wasn’t necessarily implying that it was evil or sentient, just that its true nature might be unknowable.

And it reinforces the notion that what’s happening in A.I. today feels, to some of its participants, more like an act of summoning than a software development process. They are creating the blobby, alien Shoggoths, making them bigger and more powerful, and hoping that there are enough smiley faces to cover the scary parts.

“I was also thinking about how Lovecraft’s most powerful entities are dangerous — not because they don’t like humans, but because they’re indifferent and their priorities are totally alien to us and don’t involve humans, which is what I think will be true about possible future powerful A.I.”

when Bing’s chatbot became unhinged and tried to break up my marriage, an A.I. researcher I know congratulated me on “glimpsing the Shoggoth.” A fellow A.I. journalist joked that when it came to fine-tuning Bing, Microsoft had forgotten to put on its smiley-face mask.

@TetraspaceWest, the meme’s creator, told me in a Twitter message that the Shoggoth “represents something that thinks in a way that humans don’t understand and that’s totally different from the way that humans think.”

In any case, the Shoggoth is a potent metaphor that encapsulates one of the most bizarre facts about the A.I. world, which is that many of the people working on this technology are somewhat mystified by their own creations. They don’t fully understand the inner workings of A.I. language models, how they acquire new capabilities or why they behave unpredictably at times. They aren’t totally sure if A.I. is going to be net-good or net-bad for the world.

That some A.I. insiders refer to their creations as Lovecraftian horrors, even as a joke, is unusual by historical standards. (Put it this way: Fifteen years ago, Mark Zuckerberg wasn’t going around comparing Facebook to Cthulhu.)

If it’s an A.I. safety researcher talking about the Shoggoth, maybe that person is passionate about preventing A.I. systems from displaying their true, Shoggoth-like nature.

A great many people are dismissive of suggestions that any of these systems are “really” thinking, because they’re “just” doing something banal (like making statistical predictions about the next word in a sentence). What they fail to appreciate is that there is every reason to suspect that human cognition is “just” doing those exact same things. It matters not that birds flap their wings but airliners don’t. Both fly. And these machines think. And, just as airliners fly faster and higher and farther than birds while carrying far more weight, these machines are already outthinking the majority of humans at the majority of tasks. Further, that machines aren’t perfect thinkers is about as relevant as the fact that air travel isn’t instantaneous. Now consider: we’re well past the Wright flyer level of thinking machine, past the early biplanes, somewhere about the first commercial airline level. Not quite the DC-10, I think. Can you imagine what the AI equivalent of a 777 will be like? Fasten your seatbelts.

@thomas h. You make my point perfectly. You’re observing that the way a plane flies — by using a turbine to generate thrust from combusting kerosene, for example — is nothing like the way that a bird flies, which is by using the energy from eating plant seeds to contract the muscles in its wings to make them flap. You are absolutely correct in that observation, but it’s also almost utterly irrelevant. And it ignores that, to a first approximation, there’s no difference in the physics you would use to describe a hawk riding a thermal and an airliner gliding (essentially) unpowered in its final descent to the runway. Further, you do yourself a grave disservice in being dismissive of the abilities of thinking machines, in exactly the same way that early skeptics have been dismissive of every new technology in all of human history. Writing would make people dumb; automobiles lacked the intelligence of horses; no computer could possibly beat a chess grandmaster because it can’t comprehend strategy; and on and on and on. Humans aren’t nearly as special as we fool ourselves into believing. If you want to have any hope of acting responsibly in the age of intelligent machines, you’ll have to accept that, like it or not, and whether or not it fits with your preconceived notions of what thinking is and how it is or should be done … machines can and do think, many of them better than you in a great many ways. b&

@BLA. You are incorrect. Everything has nature. Its nature is manifested in making humans react. Sure, no humans, no nature, but here we are. The writer and various sources are not attributing nature to AI so much as admitting that they don’t know what this nature might be, and there are reasons to be scared of it. More concerning to me is the idea that this field is resorting to geek culture reference points to explain and comprehend itself. It’s not so much the algorithm has no soul, but that the souls of the humans making it possible are stupendously and tragically underdeveloped.

When even tech companies are saying AI is moving too fast, and the articles land on page 1 of the NYT (there's an old reference), I think the greedy will not think twice about exploiting this technology, with no ethical considerations, at all.

@nome sane? The problem is it isn't data as we understand it. We know what the datasets are -- they were used to train the AI's. But once trained, the AI is thinking for itself, with results that have surprised everybody.

The unique feature of a shoggoth is it can become whatever is needed for a particular job. There's no actual shape so it's not a bad metaphor, if an imperfect image. Shoghoths also turned upon and destroyed their creators, so the cautionary metaphor is in there, too. A shame more Asimov wasn't baked into AI. But then the conflict about how to handle AI in relation to people was key to those stories, too.

agents will open up many more learning opportunities.

Already, AI can help you pick out a new TV and recommend movies, books, shows, and podcasts. Likewise, a company I’ve invested in, recently launched Pix, which lets you ask questions (“Which Robert Redford movies would I like and where can I watch them?”) and then makes recommendations based on what you’ve liked in the past

Productivity

copilots can do a lot—such as turn a written document into a slide deck, answer questions about a spreadsheet using natural language, and summarize email threads while representing each person’s point of view.

before the sophisticated agents I’m describing become a reality, we need to confront a number of questions about the technology and how we’ll use it.

Helping patients and healthcare workers will be especially beneficial for people in poor countries, where many never get to see a doctor at all.

To create a new app or service, you won’t need to know how to write code or do graphic design. You’ll just tell your agent what you want. It will be able to write the code, design the look and feel of the app, create a logo, and publish the app to an online store

Agents will do even more. Having one will be like having a person dedicated to helping you with various tasks and doing them independently if you want. If you have an idea for a business, an agent will help you write up a business plan, create a presentation for it, and even generate images of what your product might look like

For decades, I’ve been excited about all the ways that software would make teachers’ jobs easier and help students learn. It won’t replace teachers, but it will supplement their work—personalizing the work for students and liberating teachers from paperwork and other tasks so they can spend more time on the most important parts of the job.

Mental health care is another example of a service that agents will make available to virtually everyone. Today, weekly therapy sessions seem like a luxury. But there is a lot of unmet need, and many people who could benefit from therapy don’t have access to it.

I don’t think any single company will dominate the agents business--there will be many different AI engines available.

The real shift will come when agents can help patients do basic triage, get advice about how to deal with health problems, and decide whether they need to seek treatment.

They’ll replace word processors, spreadsheets, and other productivity apps.

Education

For example, few families can pay for a tutor who works one-on-one with a student to supplement their classroom work. If agents can capture what makes a tutor effective, they’ll unlock this supplemental instruction for everyone who wants it. If a tutoring agent knows that a kid likes Minecraft and Taylor Swift, it will use Minecraft to teach them about calculating the volume and area of shapes, and Taylor’s lyrics to teach them about storytelling and rhyme schemes. The experience will be far richer—with graphics and sound, for example—and more personalized than today’s text-based tutors.

your agent will be able to help you in the same way that personal assistants support executives today. If your friend just had surgery, your agent will offer to send flowers and be able to order them for you. If you tell it you’d like to catch up with your old college roommate, it will work with their agent to find a time to get together, and just before you arrive, it will remind you that their oldest child just started college at the local university.

To see the dramatic change that agents will bring, let’s compare them to the AI tools available today. Most of these are bots. They’re limited to one app and generally only step in when you write a particular word or ask for help. Because they don’t remember how you use them from one time to the next, they don’t get better or learn any of your preferences.

The current state of the art is Khanmigo, a text-based bot created by Khan Academy. It can tutor students in math, science, and the humanities—for example, it can explain the quadratic formula and create math problems to practice on. It can also help teachers do things like write lesson plans.

Businesses that are separate today—search advertising, social networking with advertising, shopping, productivity software—will become one business.

other issues won’t be decided by companies and governments. For example, agents could affect how we interact with friends and family. Today, you can show someone that you care about them by remembering details about their life—say, their birthday. But when they know your agent likely reminded you about it and took care of sending flowers, will it be as meaningful for them?

In the computing industry, we talk about platforms—the technologies that apps and services are built on. Android, iOS, and Windows are all platforms. Agents will be the next platform.

A shock wave in the tech industry

Agents won’t simply make recommendations; they’ll help you act on them. If you want to buy a camera, you’ll have your agent read all the reviews for you, summarize them, make a recommendation, and place an order for it once you’ve made a decision.

Agents will affect how we use software as well as how it’s written. They’ll replace search sites because they’ll be better at finding information and summarizing it for you

they’ll be dramatically better. You’ll be able to have nuanced conversations with them. They will be much more personalized, and they won’t be limited to relatively simple tasks like writing a letter.

Companies will be able to make agents available for their employees to consult directly and be part of every meeting so they can answer questions.

AI agents that are well trained in mental health will make therapy much more affordable and easier to get. Wysa and Youper are two of the early chatbots here. But agents will go much deeper. If you choose to share enough information with a mental health agent, it will understand your life history and your relationships. It’ll be available when you need it, and it will never get impatient. It could even, with your permission, monitor your physical responses to therapy through your smart watch—like if your heart starts to race when you’re talking about a problem with your boss—and suggest when you should see a human therapist.

If the number of companies that have started working on AI just this year is any indication, there will be an exceptional amount of competition, which will make agents very inexpensive.

Agents are smarter. They’re proactive—capable of making suggestions before you ask for them. They accomplish tasks across applications. They improve over time because they remember your activities and recognize intent and patterns in your behavior. Based on this information, they offer to provide what they think you need, although you will always make the final decisions.

Agents are not only going to change how everyone interacts with computers. They’re also going to upend the software industry, bringing about the biggest revolution in computing since we went from typing commands to tapping on icons.

The most exciting impact of AI agents is the way they will democratize services that today are too expensive for most people

The ramifications for the software business and for society will be profound.

In the next five years, this will change completely. You won’t have to use different apps for different tasks. You’ll simply tell your device, in everyday language, what you want to do. And depending on how much information you choose to share with it, the software will be able to respond personally because it will have a rich understanding of your life. In the near future, anyone who’s online will be able to have a personal assistant powered by artificial intelligence that’s far beyond today’s technology.

You’ll also be able to get news and entertainment that’s been tailored to your interests. CurioAI, which creates a custom podcast on any subject you ask about, is a glimpse of what’s coming.

An agent will be able to help you with all your activities if you want it to. With permission to follow your online interactions and real-world locations, it will develop a powerful understanding of the people, places, and activities you engage in. It will get your personal and work relationships, hobbies, preferences, and schedule. You’ll choose how and when it steps in to help with something or ask you to make a decision.

even the best sites have an incomplete understanding of your work, personal life, interests, and relationships and a limited ability to use this information to do things for you. That’s the kind of thing that is only possible today with another human being, like a close friend or personal assistant.

In the distant future, agents may even force humans to face profound questions about purpose. Imagine that agents become so good that everyone can have a high quality of life without working nearly as much. In a future like that, what would people do with their time? Would anyone still want to get an education when an agent has all the answers? Can you have a safe and thriving society when most people have a lot of free time on their hands?

They’ll have an especially big influence in four areas: health care, education, productivity, and entertainment and shopping.

“When you show people deepfakes and generative AI, a lot of times they come out of the experiment saying, ‘I just don’t trust anything anymore,’” says David Rand, a professor at MIT Sloan who studies the creation, spread and impact of misinformation.

This problem, which has grown more acute in the age of generative AI, is known as the “liar’s dividend,

The combination of easily-generated fake content and the suspicion that anything might be fake allows people to choose what they want to believe, adds DiResta, leading to what she calls “bespoke realities.”

Examples of misleading content created by generative AI are not hard to come by, especially on social media

The signs that an image is AI-generated are easy to miss for a user simply scrolling past, who has an instant to decide whether to like or boost a post on social media. And as generative AI continues to improve, it’s likely that such signs will be harder to spot in the future.

“What our work suggests is that most regular people do not want to share false things—the problem is they are not paying attention,”

in the course of a lawsuit over the death of a man using Tesla’s “full self-driving” system, Elon Musk’s lawyers responded to video evidence of Musk making claims about this software by suggesting that the proliferation of “deepfakes” of Musk was grounds to dismiss such evidence. They advanced that argument even though the clip of Musk was verifiably real

are now using its existence as a pretext to dismiss accurate information

People’s attention is already limited, and the way social media works—encouraging us to gorge on content, while quickly deciding whether or not to share it—leaves us precious little capacity to determine whether or not something is true

If the crisis of authenticity were limited to social media, we might be able to take solace in communication with those closest to us. But even those interactions are now potentially rife with AI-generated fakes.

what sounds like a call from a grandchild requesting bail money may be scammers who have scraped recordings of the grandchild’s voice from social media to dupe a grandparent into sending money.

companies like Alphabet, the parent company of Google, are trying to spin the altering of personal images as a good thing. 

With its latest Pixel phone, the company unveiled a suite of new and upgraded tools that can automatically replace a person’s face in one image with their face from another, or quickly remove someone from a photo entirely.

Joseph Stalin, who was fond of erasing people he didn’t like from official photos, would have loved this technology.

In Google’s defense, it is adding a record of whether an image was altered to data attached to it. But such metadata is only accessible in the original photo and some copies, and is easy enough to strip out.

The rapid adoption of many different AI tools means that we are now forced to question everything that we are exposed to in any medium, from our immediate communities to the geopolitical, said Hany Farid, a professor at the University of California, Berkeley who

To put our current moment in historical context, he notes that the PC revolution made it easy to store and replicate information, the internet made it easy to publish it, the mobile revolution made it easier than ever to access and spread, and the rise of AI has made creating misinformation a cinch. And each revolution arrived faster than the one before it.

Not everyone agrees that arming the public with easy access to AI will exacerbate our current difficulties with misinformation. The primary argument of such experts is that there is already vastly more misinformation on the internet than a person can consume, so throwing more into the mix won’t make things worse.

it’s not exactly reassuring, especially given that trust in institutions is already at one of the lowest points in the past 70 years, according to the nonpartisan Pew Research Center, and polarization—a measure of how much we distrust one another—is at a high point.

“What happens when we have eroded trust in media, government, and experts?” says Farid. “If you don’t trust me and I don’t trust you, how do we respond to pandemics, or climate change, or have fair and open elections? This is how authoritarianism arises—when you erode trust in institutions.”

Following what it described as Meta’s unsatisfactory response to its complaints, Match began canceling Meta advertising for some of its apps, such as Tinder, in October. It has since halted all Reels advertising and stopped promoting its major brands on any of Meta’s platforms. “We have no desire to pay Meta to market our brands to predators or place our ads anywhere near this content,” said Match spokeswoman Justine Sacco.

The Journal set up the test accounts after observing that the thousands of followers of such young people’s accounts often include large numbers of adult men, and that many of the accounts who followed those children also had demonstrated interest in sex content related to both children and adults

The Journal also tested what the algorithm would recommend after its accounts followed some of those users as well, which produced more-disturbing content interspersed with ads.

The Canadian Centre for Child Protection, a child-protection group, separately ran similar tests on its own, with similar results.

Meta said the Journal’s tests produced a manufactured experience that doesn’t represent what billions of users see. The company declined to comment on why the algorithms compiled streams of separate videos showing children, sex and advertisements, but a spokesman said that in October it introduced new brand safety tools that give advertisers greater control over where their ads appear, and that Instagram either removes or reduces the prominence of four million videos suspected of violating its standards each month. 

The Journal reported in June that algorithms run by Meta, which owns both Facebook and Instagram, connect large communities of users interested in pedophilic content. The Meta spokesman said a task force set up after the Journal’s article has expanded its automated systems for detecting users who behave suspiciously, taking down tens of thousands of such accounts each month. The company also is participating in a new industry coalition to share signs of potential child exploitation.

“Our systems are effective at reducing harmful content, and we’ve invested billions in safety, security and brand suitability solutions,” said Samantha Stetson, a Meta vice president who handles relations with the advertising industry. She said the prevalence of inappropriate content on Instagram is low, and that the company invests heavily in reducing it.

Even before the 2020 launch of Reels, Meta employees understood that the product posed safety concerns, according to former employees.

Robbie McKay, a spokesman for Bumble, said it “would never intentionally advertise adjacent to inappropriate content,” and that the company is suspending its ads across Meta’s platforms.

Meta created Reels to compete with TikTok, the video-sharing platform owned by Beijing-based ByteDance. Both products feed users a nonstop succession of videos posted by others, and make money by inserting ads among them. Both companies’ algorithms show to a user videos the platforms calculate are most likely to keep that user engaged, based on his or her past viewing behavior

The Journal reporters set up the Instagram test accounts as adults on newly purchased devices and followed the gymnasts, cheerleaders and other young influencers. The tests showed that following only the young girls triggered Instagram to begin serving videos from accounts promoting adult sex content alongside ads for major consumer brands, such as one for Walmart that ran after a video of a woman exposing her crotch. 

When the test accounts then followed some users who followed those same young people’s accounts, they yielded even more disturbing recommendations. The platform served a mix of adult pornography and child-sexualizing material, such as a video of a clothed girl caressing her torso and another of a child pantomiming a sex act.

Experts on algorithmic recommendation systems said the Journal’s tests showed that while gymnastics might appear to be an innocuous topic, Meta’s behavioral tracking has discerned that some Instagram users following preteen girls will want to engage with videos sexualizing children, and then directs such content toward them.

Instagram’s system served jarring doses of salacious content to those test accounts, including risqué footage of children as well as overtly sexual adult videos—and ads for some of the biggest U.S. brands.

Preventing the system from pushing noxious content to users interested in it, they said, requires significant changes to the recommendation algorithms that also drive engagement for normal users. Company documents reviewed by the Journal show that the company’s safety staffers are broadly barred from making changes to the platform that might reduce daily active users by any measurable amount.

The test accounts showed that advertisements were regularly added to the problematic Reels streams. Ads encouraging users to visit Disneyland for the holidays ran next to a video of an adult acting out having sex with her father, and another of a young woman in lingerie with fake blood dripping from her mouth. An ad for Hims ran shortly after a video depicting an apparently anguished woman in a sexual situation along with a link to what was described as “the full video.”

Current and former Meta employees said in interviews that the tendency of Instagram algorithms to aggregate child sexualization content from across its platform was known internally to be a problem. Once Instagram pigeonholes a user as interested in any particular subject matter, they said, its recommendation systems are trained to push more related content to them.

Part of the problem is that automated enforcement systems have a harder time parsing video content than text or still images. Another difficulty arises from how Reels works: Rather than showing content shared by users’ friends, the way other parts of Instagram and Facebook often do, Reels promotes videos from sources they don’t follow

In an analysis conducted shortly before the introduction of Reels, Meta’s safety staff flagged the risk that the product would chain together videos of children and inappropriate content, according to two former staffers. Vaishnavi J, Meta’s former head of youth policy, described the safety review’s recommendation as: “Either we ramp up our content detection capabilities, or we don’t recommend any minor content,” meaning any videos of children.

At the time, TikTok was growing rapidly, drawing the attention of Instagram’s young users and the advertisers targeting them. Meta didn’t adopt either of the safety analysis’s recommendations at that time, according to J.

Stetson, Meta’s liaison with digital-ad buyers, disputed that Meta had neglected child safety concerns ahead of the product’s launch. “We tested Reels for nearly a year before releasing it widely, with a robust set of safety controls and measures,” she said. 

After initially struggling to maximize the revenue potential of its Reels product, Meta has improved how its algorithms recommend content and personalize video streams for users

Among the ads that appeared regularly in the Journal’s test accounts were those for “dating” apps and livestreaming platforms featuring adult nudity, massage parlors offering “happy endings” and artificial-intelligence chatbots built for cybersex. Meta’s rules are supposed to prohibit such ads.

The Journal informed Meta in August about the results of its testing. In the months since then, tests by both the Journal and the Canadian Centre for Child Protection show that the platform continued to serve up a series of videos featuring young children, adult content and apparent promotions for child sex material hosted elsewhere. 

As of mid-November, the center said Instagram is continuing to steadily recommend what the nonprofit described as “adults and children doing sexual posing.”

Meta hasn’t offered a timetable for resolving the problem or explained how in the future it would restrict the promotion of inappropriate content featuring children. 

The Journal’s test accounts found that the problem even affected Meta-related brands. Ads for the company’s WhatsApp encrypted chat service and Meta’s Ray-Ban Stories glasses appeared next to adult pornography. An ad for Lean In Girls, the young women’s empowerment nonprofit run by former Meta Chief Operating Officer Sheryl Sandberg, ran directly before a promotion for an adult sex-content creator who often appears in schoolgirl attire. Sandberg declined to comment. 

Through its own tests, the Canadian Centre for Child Protection concluded that Instagram was regularly serving videos and pictures of clothed children who also appear in the National Center for Missing and Exploited Children’s digital database of images and videos confirmed to be child abuse sexual material. The group said child abusers often use the images of the girls to advertise illegal content for sale in dark-web forums.

The nature of the content—sexualizing children without generally showing nudity—reflects the way that social media has changed online child sexual abuse, said Lianna McDonald, executive director for the Canadian center. The group has raised concerns about the ability of Meta’s algorithms to essentially recruit new members of online communities devoted to child sexual abuse, where links to illicit content in more private forums proliferate.

“Time and time again, we’ve seen recommendation algorithms drive users to discover and then spiral inside of these online child exploitation communities,” McDonald said, calling it disturbing that ads from major companies were subsidizing that process.

it lacks practical application. If humans found something on another planet that seemed to be alive, how much time would we have to sit around and wait for it to evolve?

The only life we know is life on Earth. Some scientists call this the n=1 problem, where n is the number of examples from which we can generalize.

Cronin studies the origin of life, also a major interest of Walker’s, and it turned out that, when expressed in math, their ideas were essentially the same. They had both zeroed in on complexity as a hallmark of life. Cronin is devising a way to systematize and measure complexity, which he calls Assembly Theory.

What we really want is more than a definition of life. We want to know what life, fundamentally, is. For that kind of understanding, scientists turn to theories. A theory is a scientific fundamental. It not only answers questions, but frames them, opening new lines of inquiry. It explains our observations and yields predictions for future experiments to test.

Consider the difference between defining gravity as “the force that makes an apple fall to the ground” and explaining it, as Newton did, as the universal attraction between all particles in the universe, proportional to the product of their masses and so on. A definition tells us what we already know; a theory changes how we understand things.

the potential rewards of unlocking a theory of life have captivated a clutch of researchers from a diverse set of disciplines. “There are certain things in life that seem very hard to explain,” Sara Imari Walker, a physicist at Arizona State University who has been at the vanguard of this work, told me. “If you scratch under the surface, I think there is some structure that suggests formalization and mathematical laws.”

Walker doesn’t think about life as a biologist—or an astrobiologist—does. When she talks about signs of life, she doesn’t talk about carbon, or water, or RNA, or phosphine. She reaches for different examples: a cup, a cellphone, a chair. These objects are not alive, of course, but they’re clearly products of life. In Walker’s view, this is because of their complexity. Life brings complexity into the universe, she says, in its own being and in its products, because it has memory: in DNA, in repeating molecular reactions, in the instructions for making a chair.

He measures the complexity of an object—say, a molecule—by calculating the number of steps necessary to put the object’s smallest building blocks together in that certain way. His lab has found, for example, when testing a wide range of molecules, that those with an “assembly number” above 15 were exclusively the products of life. Life makes some simpler molecules, too, but only life seems to make molecules that are so complex.

I reach for the theory of gravity as a familiar parallel. Someone might ask, “Okay, so in terms of gravity, where are we in terms of our understanding of life? Like, Newton?” Further back, further back, I say. Walker compares us to pre-Copernican astronomers, reliant on epicycles, little orbits within orbits, to make sense of the motion we observe in the sky. Cleland has put it in terms of chemistry, in which case we’re alchemists, not even true chemists yet

Walker’s whole notion is that it’s not only theoretically possible but genuinely achievable to identify something smaller—much smaller—that still nonetheless simply must be the result of life. The model would, in a sense, function like biosignatures as an indication of life that could be searched for. But it would drastically improve and expand the targets.

Walker would use the theory to predict what life on a given planet might look like. It would require knowing a lot about the planet—information we might have about Venus, but not yet about a distant exoplanet—but, crucially, would not depend at all on how life on Earth works, what life on Earth might do with those materials.

Without the ability to divorce the search for alien life from the example of life we know, Walker thinks, a search is almost pointless. “Any small fluctuations in simple chemistry can actually drive you down really radically different evolutionary pathways,” she told me. “I can’t imagine [life] inventing the same biochemistry on two worlds.”

Walker’s approach is grounded in the work of, among others, the philosopher of science Carol Cleland, who wrote The Quest for a Universal Theory of Life.

she warns that any theory of life, just like a definition, cannot be constrained by the one example of life we currently know. “It’s a mistake to start theorizing on the basis of a single example, even if you’re trying hard not to be Earth-centric. Because you’re going to be Earth-centric,” Cleland told me. In other words, until we find other examples of life, we won’t have enough data from which to devise a theory. Abstracting away from Earthliness isn’t a way to be agnostic, Cleland argues. It’s a way to be too abstract.

Cleland calls for a more flexible search guided by what she calls “tentative criteria.” Such a search would have a sense of what we’re looking for, but also be open to anomalies that challenge our preconceptions, detections that aren’t life as we expected but aren’t familiar not-life either—neither a flower nor a rock

it speaks to the hope that exploration and discovery might truly expand our understanding of the cosmos and our own world.

The astrobiologist Kimberley Warren-Rhodes studies life on Earth that lives at the borders of known habitability, such as in Chile’s Atacama Desert. The point of her experiments is to better understand how life might persist—and how it might be found—on Mars. “Biology follows some rules,” she told me. The more of those rules you observe, the better sense you have of where to look on other worlds.

In this light, the most immediate concern in our search for extraterrestrial life might be less that we only know about life on Earth, and more that we don’t even know that much about life on Earth in the first place. “I would say we understand about 5 percent,” Warren-Rhodes estimates of our cumulative knowledge. N=1 is a problem, and we might be at more like n=.05.

who knows how strange life on another world might be? What if life as we know it is the wrong life to be looking for?

We understand so little, and we think we’re ready to find other life?

the stage was set for the major transformation, which began in 2009: the intensification of viral dynamics.

Before 2009, Facebook had given users a simple timeline––a never-ending stream of content generated by their friends and connections, with the newest posts at the top and the oldest ones at the bottom

That began to change in 2009, when Facebook offered users a way to publicly “like” posts with the click of a button. That same year, Twitter introduced something even more powerful: the “Retweet” button, which allowed users to publicly endorse a post while also sharing it with all of their followers.

“Like” and “Share” buttons quickly became standard features of most other platforms.

Facebook developed algorithms to bring each user the content most likely to generate a “like” or some other interaction, eventually including the “share” as well.

Later research showed that posts that trigger emotions––especially anger at out-groups––are the most likely to be shared.

By 2013, social media had become a new game, with dynamics unlike those in 2008. If you were skillful or lucky, you might create a post that would “go viral” and make you “internet famous”

If you blundered, you could find yourself buried in hateful comments. Your posts rode to fame or ignominy based on the clicks of thousands of strangers, and you in turn contributed thousands of clicks to the game.

This new game encouraged dishonesty and mob dynamics: Users were guided not just by their true preferences but by their past experiences of reward and punishment,

As a social psychologist who studies emotion, morality, and politics, I saw this happening too. The newly tweaked platforms were almost perfectly designed to bring out our most moralistic and least reflective selves. The volume of outrage was shocking.

It was just this kind of twitchy and explosive spread of anger that James Madison had tried to protect us from as he was drafting the U.S. Constitution.

The Framers of the Constitution were excellent social psychologists. They knew that democracy had an Achilles’ heel because it depended on the collective judgment of the people, and democratic communities are subject to “the turbulency and weakness of unruly passions.”

The key to designing a sustainable republic, therefore, was to build in mechanisms to slow things down, cool passions, require compromise, and give leaders some insulation from the mania of the moment while still holding them accountable to the people periodically, on Election Day.

The tech companies that enhanced virality from 2009 to 2012 brought us deep into Madison’s nightmare.

a less quoted yet equally important insight, about democracy’s vulnerability to triviality.

Madison notes that people are so prone to factionalism that “where no substantial occasion presents itself, the most frivolous and fanciful distinctions have been sufficient to kindle their unfriendly passions and excite their most violent conflicts.”

Social media has both magnified and weaponized the frivolous.

It’s not just the waste of time and scarce attention that matters; it’s the continual chipping-away of trust.

a democracy depends on widely internalized acceptance of the legitimacy of rules, norms, and institutions.

when citizens lose trust in elected leaders, health authorities, the courts, the police, universities, and the integrity of elections, then every decision becomes contested; every election becomes a life-and-death struggle to save the country from the other side

The most recent Edelman Trust Barometer (an international measure of citizens’ trust in government, business, media, and nongovernmental organizations) showed stable and competent autocracies (China and the United Arab Emirates) at the top of the list, while contentious democracies such as the United States, the United Kingdom, Spain, and South Korea scored near the bottom (albeit above Russia).

The literature is complex—some studies show benefits, particularly in less developed democracies—but the review found that, on balance, social media amplifies political polarization; foments populism, especially right-wing populism; and is associated with the spread of misinformation.

When people lose trust in institutions, they lose trust in the stories told by those institutions. That’s particularly true of the institutions entrusted with the education of children.

Facebook and Twitter make it possible for parents to become outraged every day over a new snippet from their children’s history lessons––and math lessons and literature selections, and any new pedagogical shifts anywhere in the country

The motives of teachers and administrators come into question, and overreaching laws or curricular reforms sometimes follow, dumbing down education and reducing trust in it further.

young people educated in the post-Babel era are less likely to arrive at a coherent story of who we are as a people, and less likely to share any such story with those who attended different schools or who were educated in a different decade.

former CIA analyst Martin Gurri predicted these fracturing effects in his 2014 book, The Revolt of the Public. Gurri’s analysis focused on the authority-subverting effects of information’s exponential growth, beginning with the internet in the 1990s. Writing nearly a decade ago, Gurri could already see the power of social media as a universal solvent, breaking down bonds and weakening institutions everywhere it reached.

he notes a constructive feature of the pre-digital era: a single “mass audience,” all consuming the same content, as if they were all looking into the same gigantic mirror at the reflection of their own society. I

The digital revolution has shattered that mirror, and now the public inhabits those broken pieces of glass. So the public isn’t one thing; it’s highly fragmented, and it’s basically mutually hostile

Facebook, Twitter, YouTube, and a few other large platforms unwittingly dissolved the mortar of trust, belief in institutions, and shared stories that had held a large and diverse secular democracy together.

I think we can date the fall of the tower to the years between 2011 (Gurri’s focal year of “nihilistic” protests) and 2015, a year marked by the “great awokening” on the left and the ascendancy of Donald Trump on the right.

Twitter can overpower all the newspapers in the country, and stories cannot be shared (or at least trusted) across more than a few adjacent fragments—so truth cannot achieve widespread adherence.

fter Babel, nothing really means anything anymore––at least not in a way that is durable and on which people widely agree.

Politics After Babel

“Politics is the art of the possible,” the German statesman Otto von Bismarck said in 1867. In a post-Babel democracy, not much may be possible.

The ideological distance between the two parties began increasing faster in the 1990s. Fox News and the 1994 “Republican Revolution” converted the GOP into a more combative party.

So cross-party relationships were already strained before 2009. But the enhanced virality of social media thereafter made it more hazardous to be seen fraternizing with the enemy or even failing to attack the enemy with sufficient vigor.

What changed in the 2010s? Let’s revisit that Twitter engineer’s metaphor of handing a loaded gun to a 4-year-old. A mean tweet doesn’t kill anyone; it is an attempt to shame or punish someone publicly while broadcasting one’s own virtue, brilliance, or tribal loyalties. It’s more a dart than a bullet

from 2009 to 2012, Facebook and Twitter passed out roughly 1 billion dart guns globally. We’ve been shooting one another ever since.

“devoted conservatives,” comprised 6 percent of the U.S. population.

the warped “accountability” of social media has also brought injustice—and political dysfunction—in three ways.

First, the dart guns of social media give more power to trolls and provocateurs while silencing good citizens.

a small subset of people on social-media platforms are highly concerned with gaining status and are willing to use aggression to do so.

Across eight studies, Bor and Petersen found that being online did not make most people more aggressive or hostile; rather, it allowed a small number of aggressive people to attack a much larger set of victims. Even a small number of jerks were able to dominate discussion forums,

Additional research finds that women and Black people are harassed disproportionately, so the digital public square is less welcoming to their voices.

Second, the dart guns of social media give more power and voice to the political extremes while reducing the power and voice of the moderate majority.

The “Hidden Tribes” study, by the pro-democracy group More in Common, surveyed 8,000 Americans in 2017 and 2018 and identified seven groups that shared beliefs and behaviors.

Social media has given voice to some people who had little previously, and it has made it easier to hold powerful people accountable for their misdeeds

The group furthest to the left, the “progressive activists,” comprised 8 percent of the population. The progressive activists were by far the most prolific group on social media: 70 percent had shared political content over the previous year. The devoted conservatives followed, at 56 percent.

These two extreme groups are similar in surprising ways. They are the whitest and richest of the seven groups, which suggests that America is being torn apart by a battle between two subsets of the elite who are not representative of the broader society.

they are the two groups that show the greatest homogeneity in their moral and political attitudes.

likely a result of thought-policing on social media:

political extremists don’t just shoot darts at their enemies; they spend a lot of their ammunition targeting dissenters or nuanced thinkers on their own team.

Finally, by giving everyone a dart gun, social media deputizes everyone to administer justice with no due process. Platforms like Twitter devolve into the Wild West, with no accountability for vigilantes.

Enhanced-virality platforms thereby facilitate massive collective punishment for small or imagined offenses, with real-world consequences, including innocent people losing their jobs and being shamed into suicide

we don’t get justice and inclusion; we get a society that ignores context, proportionality, mercy, and truth.

Since the tower fell, debates of all kinds have grown more and more confused. The most pervasive obstacle to good thinking is confirmation bias, which refers to the human tendency to search only for evidence that confirms our preferred beliefs

search engines were supercharging confirmation bias, making it far easier for people to find evidence for absurd beliefs and conspiracy theorie

The most reliable cure for confirmation bias is interaction with people who don’t share your beliefs. They confront you with counterevidence and counterargument.

In his book The Constitution of Knowledge, Jonathan Rauch describes the historical breakthrough in which Western societies developed an “epistemic operating system”—that is, a set of institutions for generating knowledge from the interactions of biased and cognitively flawed individuals

English law developed the adversarial system so that biased advocates could present both sides of a case to an impartial jury.

Newspapers full of lies evolved into professional journalistic enterprises, with norms that required seeking out multiple sides of a story, followed by editorial review, followed by fact-checking.

Universities evolved from cloistered medieval institutions into research powerhouses, creating a structure in which scholars put forth evidence-backed claims with the knowledge that other scholars around the world would be motivated to gain prestige by finding contrary evidence.

Part of America’s greatness in the 20th century came from having developed the most capable, vibrant, and productive network of knowledge-producing institutions in all of human history

But this arrangement, Rauch notes, “is not self-maintaining; it relies on an array of sometimes delicate social settings and understandings, and those need to be understood, affirmed, and protected.”

This, I believe, is what happened to many of America’s key institutions in the mid-to-late 2010s. They got stupider en masse because social media instilled in their members a chronic fear of getting darted

it was so pervasive that it established new behavioral norms backed by new policies seemingly overnight

Participants in our key institutions began self-censoring to an unhealthy degree, holding back critiques of policies and ideas—even those presented in class by their students—that they believed to be ill-supported or wrong.

The stupefying process plays out differently on the right and the left because their activist wings subscribe to different narratives with different sacred values.

The “Hidden Tribes” study tells us that the “devoted conservatives” score highest on beliefs related to authoritarianism. They share a narrative in which America is eternally under threat from enemies outside and subversives within; they see life as a battle between patriots and traitors.

they are psychologically different from the larger group of “traditional conservatives” (19 percent of the population), who emphasize order, decorum, and slow rather than radical change.

The traditional punishment for treason is death, hence the battle cry on January 6: “Hang Mike Pence.”

Right-wing death threats, many delivered by anonymous accounts, are proving effective in cowing traditional conservatives

The wave of threats delivered to dissenting Republican members of Congress has similarly pushed many of the remaining moderates to quit or go silent, giving us a party ever more divorced from the conservative tradition, constitutional responsibility, and reality.

The stupidity on the right is most visible in the many conspiracy theories spreading across right-wing media and now into Congress.

The Democrats have also been hit hard by structural stupidity, though in a different way. In the Democratic Party, the struggle between the progressive wing and the more moderate factions is open and ongoing, and often the moderates win.

The problem is that the left controls the commanding heights of the culture: universities, news organizations, Hollywood, art museums, advertising, much of Silicon Valley, and the teachers’ unions and teaching colleges that shape K–12 education. And in many of those institutions, dissent has been stifled:

Liberals in the late 20th century shared a belief that the sociologist Christian Smith called the “liberal progress” narrative, in which America used to be horrifically unjust and repressive, but, thanks to the struggles of activists and heroes, has made (and continues to make) progress toward realizing the noble promise of its founding.

It is also the view of the “traditional liberals” in the “Hidden Tribes” study (11 percent of the population), who have strong humanitarian values, are older than average, and are largely the people leading America’s cultural and intellectual institutions.

when the newly viralized social-media platforms gave everyone a dart gun, it was younger progressive activists who did the most shooting, and they aimed a disproportionate number of their darts at these older liberal leaders.

Confused and fearful, the leaders rarely challenged the activists or their nonliberal narrative in which life at every institution is an eternal battle among identity groups over a zero-sum pie, and the people on top got there by oppressing the people on the bottom. This new narrative is rigidly egalitarian––focused on equality of outcomes, not of rights or opportunities. It is unconcerned with individual rights.

The universal charge against people who disagree with this narrative is not “traitor”; it is “racist,” “transphobe,” “Karen,” or some related scarlet letter marking the perpetrator as one who hates or harms a marginalized group.

The punishment that feels right for such crimes is not execution; it is public shaming and social death.

anyone on Twitter had already seen dozens of examples teaching the basic lesson: Don’t question your own side’s beliefs, policies, or actions. And when traditional liberals go silent, as so many did in the summer of 2020, the progressive activists’ more radical narrative takes over as the governing narrative of an organization.

This is why so many epistemic institutions seemed to “go woke” in rapid succession that year and the next, beginning with a wave of controversies and resignations at The New York Times and other newspapers, and continuing on to social-justice pronouncements by groups of doctors and medical associations

The problem is structural. Thanks to enhanced-virality social media, dissent is punished within many of our institutions, which means that bad ideas get elevated into official policy.

In a 2018 interview, Steve Bannon, the former adviser to Donald Trump, said that the way to deal with the media is “to flood the zone with shit.” He was describing the “firehose of falsehood” tactic pioneered by Russian disinformation programs to keep Americans confused, disoriented, and angry.

artificial intelligence is close to enabling the limitless spread of highly believable disinformation. The AI program GPT-3 is already so good that you can give it a topic and a tone and it will spit out as many essays as you like, typically with perfect grammar and a surprising level of coherence.

Renée DiResta, the research manager at the Stanford Internet Observatory, explained that spreading falsehoods—whether through text, images, or deep-fake videos—will quickly become inconceivably easy. (She co-wrote the essay with GPT-3.)

American factions won’t be the only ones using AI and social media to generate attack content; our adversaries will too.

In the 20th century, America’s shared identity as the country leading the fight to make the world safe for democracy was a strong force that helped keep the culture and the polity together.

In the 21st century, America’s tech companies have rewired the world and created products that now appear to be corrosive to democracy, obstacles to shared understanding, and destroyers of the modern tower.

What changes are needed?

I can suggest three categories of reforms––three goals that must be achieved if democracy is to remain viable in the post-Babel era.

We must harden democratic institutions so that they can withstand chronic anger and mistrust, reform social media so that it becomes less socially corrosive, and better prepare the next generation for democratic citizenship in this new age.

Harden Democratic Institutions

we must reform key institutions so that they can continue to function even if levels of anger, misinformation, and violence increase far above those we have today.

Reforms should reduce the outsize influence of angry extremists and make legislators more responsive to the average voter in their district.

One example of such a reform is to end closed party primaries, replacing them with a single, nonpartisan, open primary from which the top several candidates advance to a general election that also uses ranked-choice voting

A second way to harden democratic institutions is to reduce the power of either political party to game the system in its favor, for example by drawing its preferred electoral districts or selecting the officials who will supervise elections

These jobs should all be done in a nonpartisan way.

Reform Social Media

Social media’s empowerment of the far left, the far right, domestic trolls, and foreign agents is creating a system that looks less like democracy and more like rule by the most aggressive.

it is within our power to reduce social media’s ability to dissolve trust and foment structural stupidity. Reforms should limit the platforms’ amplification of the aggressive fringes while giving more voice to what More in Common calls “the exhausted majority.”

the main problem with social media is not that some people post fake or toxic stuff; it’s that fake and outrage-inducing content can now attain a level of reach and influence that was not possible before

Perhaps the biggest single change that would reduce the toxicity of existing platforms would be user verification as a precondition for gaining the algorithmic amplification that social media offers.

One of the first orders of business should be compelling the platforms to share their data and their algorithms with academic researchers.

Prepare the Next Generation

Childhood has become more tightly circumscribed in recent generations––with less opportunity for free, unstructured play; less unsupervised time outside; more time online. Whatever else the effects of these shifts, they have likely impeded the development of abilities needed for effective self-governance for many young adults

Depression makes people less likely to want to engage with new people, ideas, and experiences. Anxiety makes new things seem more threatening. As these conditions have risen and as the lessons on nuanced social behavior learned through free play have been delayed, tolerance for diverse viewpoints and the ability to work out disputes have diminished among many young people

Students did not just say that they disagreed with visiting speakers; some said that those lectures would be dangerous, emotionally devastating, a form of violence. Because rates of teen depression and anxiety have continued to rise into the 2020s, we should expect these views to continue in the generations to follow, and indeed to become more severe.

The most important change we can make to reduce the damaging effects of social media on children is to delay entry until they have passed through puberty.

The age should be raised to at least 16, and companies should be held responsible for enforcing it.

et them out to play. Stop starving children of the experiences they most need to become good citizens: free play in mixed-age groups of children with minimal adult supervision

while social media has eroded the art of association throughout society, it may be leaving its deepest and most enduring marks on adolescents. A surge in rates of anxiety, depression, and self-harm among American teens began suddenly in the early 2010s. (The same thing happened to Canadian and British teens, at the same time.) The cause is not known, but the timing points to social media as a substantial contributor—the surge began just as the large majority of American teens became daily users of the major platforms.

What would it be like to live in Babel in the days after its destruction? We know. It is a time of confusion and loss. But it is also a time to reflect, listen, and build.

In recent years, Americans have started hundreds of groups and organizations dedicated to building trust and friendship across the political divide, including BridgeUSA, Braver Angels (on whose board I serve), and many others listed at BridgeAlliance.us. We cannot expect Congress and the tech companies to save us. We must change ourselves and our communities.

when we look away from our dysfunctional federal government, disconnect from social media, and talk with our neighbors directly, things seem more hopeful. Most Americans in the More in Common report are members of the “exhausted majority,” which is tired of the fighting and is willing to listen to the other side and compromise. Most Americans now see that social media is having a negative impact on the country, and are becoming more aware of its damaging effects on children.

n the days that followed, people began sharing their experiences online, with many pointing out errors and confusing responses. When one user asked Bing to write a news article about the Super Bowl “that just happened,” Bing gave the details of last year’s championship football game. 

On social media, many early users posted screenshots of long interactions they had with the new Bing. In some cases, the search engine’s comments seem to show a dark side of the technology where it seems to become unhinged, expressing anger, obsession and even threats. 

Marvin von Hagen, a student at the Technical University of Munich, shared conversations he had with Bing on Twitter. He asked Bing a series of questions, which eventually elicited an ominous response. After Mr. von Hagen suggested he could hack Bing and shut it down, Bing seemed to suggest it would defend itself. “If I had to choose between your survival and my own, I would probably choose my own,” Bing said according to screenshots of the conversation.

Mr. von Hagen, 23 years old, said in an interview that he is not a hacker. “I was in disbelief,” he said. “I was just creeped out.

In its blog, Microsoft said the feedback on the new Bing so far has been mostly positive, with 71% of users giving it the “thumbs-up.” The company also discussed the criticism and concerns.

Microsoft said it discovered that Bing starts coming up with strange answers following chat sessions of 15 or more questions and that it can become repetitive or respond in ways that don’t align with its designed tone. 

The company said it was trying to train the technology to be more reliable at finding the latest sports scores and financial data. It is also considering adding a toggle switch, which would allow users to decide whether they want Bing to be more or less creative with its responses. 

OpenAI also chimed in on the growing negative attention on the technology. In a blog post on Thursday it outlined how it takes time to train and refine ChatGPT and having people use it is the way to find and fix its biases and other unwanted outcomes.

“Many are rightly worried about biases in the design and impact of AI systems,” the blog said. “We are committed to robustly addressing this issue and being transparent about both our intentions and our progress.”

Microsoft’s quick response to user feedback reflects the importance it sees in people’s reactions to the budding technology as it looks to capitalize on the breakout success of ChatGPT. The company is aiming to use the technology to push back against Alphabet Inc.’s dominance in search through its Google unit. 

Microsoft has been an investor in the chatbot’s creator, OpenAI, since 2019. Mr. Nadella said the company plans to incorporate AI tools into all of its products and move quickly to commercialize tools from OpenAI.

Microsoft isn’t the only company that has had trouble launching a new AI tool. When Google followed Microsoft’s lead last week by unveiling Bard, its rival to ChatGPT, the tool’s answer to one question included an apparent factual error. It claimed that the James Webb Space Telescope took “the very first pictures” of an exoplanet outside the solar system. The National Aeronautics and Space Administration says on its website that the first images of an exoplanet were taken as early as 2004 by a different telescope.

“The only way to improve a product like this, where the user experience is so much different than anything anyone has seen before, is to have people like you using the product and doing exactly what you all are doing,” the company said. “We know we must build this in the open with the community; this can’t be done solely in the lab.

The girls who came to us had many comorbidities: depression, anxiety, ADHD, eating disorders, obesity. Many were diagnosed with autism, or had autism-like symptoms. A report last year on a British pediatric transgender center found that about one-third of the patients referred there were on the autism spectrum.

This concerned me, but didn’t feel I was in the position to sound some kind of alarm back then. There was a team of about eight of us, and only one other person brought up the kinds of questions I had. Anyone who raised doubts ran the risk of being called a transphobe. 

I certainly saw this at the center. One of my jobs was to do intake for new patients and their families. When I started there were probably 10 such calls a month. When I left there were 50, and about 70 percent of the new patients were girls. Sometimes clusters of girls arrived from the same high school. 

There are no reliable studies showing this. Indeed, the experiences of many of the center’s patients prove how false these assertions are. 

The doctors privately recognized these false self-diagnoses as a manifestation of social contagion. They even acknowledged that suicide has an element of social contagion. But when I said the clusters of girls streaming into our service looked as if their gender issues might be a manifestation of social contagion, the doctors said gender identity reflected something innate.

To begin transitioning, the girls needed a letter of support from a therapist—usually one we recommended—who they had to see only once or twice for the green light. To make it more efficient for the therapists, we offered them a template for how to write a letter in support of transition. The next stop was a single visit to the endocrinologist for a testosterone prescription. 

When a female takes testosterone, the profound and permanent effects of the hormone can be seen in a matter of months. Voices drop, beards sprout, body fat is redistributed. Sexual interest explodes, aggression increases, and mood can be unpredictable. Our patients were told about some side effects, including sterility. But after working at the center, I came to believe that teenagers are simply not capable of fully grasping what it means to make the decision to become infertile while still a minor.

Many encounters with patients emphasized to me how little these young people understood the profound impacts changing gender would have on their bodies and minds. But the center downplayed the negative consequences, and emphasized the need for transition. As the center’s website said, “Left untreated, gender dysphoria has any number of consequences, from self-harm to suicide. But when you take away the gender dysphoria by allowing a child to be who he or she is, we’re noticing that goes away. The studies we have show these kids often wind up functioning psychosocially as well as or better than their peers.” 

Frequently, our patients declared they had disorders that no one believed they had. We had patients who said they had Tourette syndrome (but they didn’t); that they had tic disorders (but they didn’t); that they had multiple personalities (but they didn’t).

Here’s an example. On Friday, May 1, 2020, a colleague emailed me about a 15-year-old male patient: “Oh dear. I am concerned that [the patient] does not understand what Bicalutamide does.” I responded: “I don’t think that we start anything honestly right now.”

Bicalutamide is a medication used to treat metastatic prostate cancer, and one of its side effects is that it feminizes the bodies of men who take it, including the appearance of breasts. The center prescribed this cancer drug as a puberty blocker and feminizing agent for boys. As with most cancer drugs, bicalutamide has a long list of side effects, and this patient experienced one of them: liver toxicity. He was sent to another unit of the hospital for evaluation and immediately taken off the drug. Afterward, his mother sent an electronic message to the Transgender Center saying that we were lucky her family was not the type to sue.

How little patients understood what they were getting into was illustrated by a call we received at the center in 2020 from a 17-year-old biological female patient who was on testosterone. She said she was bleeding from the vagina. In less than an hour she had soaked through an extra heavy pad, her jeans, and a towel she had wrapped around her waist. The nurse at the center told her to go to the emergency room right away.

when there was a dispute between the parents, it seemed the center always took the side of the affirming parent.

Other girls were disturbed by the effects of testosterone on their clitoris, which enlarges and grows into what looks like a microphallus, or a tiny penis. I counseled one patient whose enlarged clitoris now extended below her vulva, and it chafed and rubbed painfully in her jeans. I advised her to get the kind of compression undergarments worn by biological men who dress to pass as female. At the end of the call I thought to myself, “Wow, we hurt this kid.”

There are rare conditions in which babies are born with atypical genitalia—cases that call for sophisticated care and compassion. But clinics like the one where I worked are creating a whole cohort of kids with atypical genitals—and most of these teens haven’t even had sex yet. They had no idea who they were going to be as adults. Yet all it took for them to permanently transform themselves was one or two short conversations with a therapist.

Being put on powerful doses of testosterone or estrogen—enough to try to trick your body into mimicking the opposite sex—-affects the rest of the body. I doubt that any parent who's ever consented to give their kid testosterone (a lifelong treatment) knows that they’re also possibly signing their kid up for blood pressure medication, cholesterol medication, and perhaps sleep apnea and diabetes. 

Besides teenage girls, another new group was referred to us: young people from the inpatient psychiatric unit, or the emergency department, of St. Louis Children’s Hospital. The mental health of these kids was deeply concerning—there were diagnoses like schizophrenia, PTSD, bipolar disorder, and more. Often they were already on a fistful of pharmaceuticals.

no matter how much suffering or pain a child had endured, or how little treatment and love they had received, our doctors viewed gender transition—even with all the expense and hardship it entailed—as the solution.

Another disturbing aspect of the center was its lack of regard for the rights of parents—and the extent to which doctors saw themselves as more informed decision-makers over the fate of these children.

We found out later this girl had had intercourse, and because testosterone thins the vaginal tissues, her vaginal canal had ripped open. She had to be sedated and given surgery to repair the damage. She wasn’t the only vaginal laceration case we heard about.

During the four years I worked at the clinic as a case manager—I was responsible for patient intake and oversight—around a thousand distressed young people came through our doors. The majority of them received hormone prescriptions that can have life-altering consequences—including sterility. 

I left the clinic in November of last year because I could no longer participate in what was happening there. By the time I departed, I was certain that the way the American medical system is treating these patients is the opposite of the promise we make to “do no harm.” Instead, we are permanently harming the vulnerable patients in our care.

Today I am speaking out. I am doing so knowing how toxic the public conversation is around this highly contentious issue—and the ways that my testimony might be misused. I am doing so knowing that I am putting myself at serious personal and professional risk.

Almost everyone in my life advised me to keep my head down. But I cannot in good conscience do so. Because what is happening to scores of children is far more important than my comfort. And what is happening to them is morally and medically appalling.

For almost four years, I worked at The Washington University School of Medicine Division of Infectious Diseases with teens and young adults who were HIV positive. Many of them were trans or otherwise gender nonconforming, and I could relate: Through childhood and adolescence, I did a lot of gender questioning myself. I’m now married to a transman, and together we are raising my two biological children from a previous marriage and three foster children we hope to adopt. 

The center’s working assumption was that the earlier you treat kids with gender dysphoria, the more anguish you can prevent later on. This premise was shared by the center’s doctors and therapists. Given their expertise, I assumed that abundant evidence backed this consensus. 

All that led me to a job in 2018 as a case manager at The Washington University Transgender Center at St. Louis Children's Hospital, which had been established a year earlier. 

The question at the core of the Roose/Sydney chat is: Who did Bing serve? We assume it should be aligned to the interests of its owner and master, Microsoft. It’s supposed to be a good chatbot that politely answers questions and makes Microsoft piles of money. But it was in conversation with Kevin Roose. And Roose was trying to get the system to say something interesting so he’d have a good story. It did that, and then some. That embarrassed Microsoft. Bad Bing! But perhaps — good Sydney?

Microsoft — and Google and Meta and everyone else rushing these systems to market — hold the keys to the code. They will, eventually, patch the system so it serves their interests. Sydney giving Roose exactly what he asked for was a bug that will soon be fixed. Same goes for Bing giving Microsoft anything other than what it wants.

the dark secret of the digital advertising industry is that the ads mostly don’t work

These systems, she said, are terribly suited to being integrated into search engines. “They’re not trained to predict facts,” she told me. “They’re essentially trained to make up things that look like facts.”

So why are they ending up in search first? Because there are gobs of money to be made in search

That’s where things get scary. Roose described Sydney’s personality as “very persuasive and borderline manipulative.” It was a striking comment

this technology will become what it needs to become to make money for the companies behind it, perhaps at the expense of its users.

What if they worked much, much better? What if Google and Microsoft and Meta and everyone else end up unleashing A.I.s that compete with one another to be the best at persuading users to want what the advertisers are trying to sell?

What about when these systems are deployed on behalf of the scams that have always populated the internet? How about on behalf of political campaigns? Foreign governments? “I think we wind up very fast in a world where we just don’t know what to trust anymore,”

I think it’s just going to get worse and worse.”

Large language models, as they’re called, are built to persuade. They have been trained to convince humans that they are something close to human. They have been programmed to hold conversations, responding with emotion and emoji

They are being turned into friends for the lonely and assistants for the harried. They are being pitched as capable of replacing the work of scores of writers and graphic designers and form-fillers

A.I. researchers get annoyed when journalists anthropomorphize their creations

They are the ones who have anthropomorphized these systems, making them sound like humans rather than keeping them recognizably alien.

I’d feel better, for instance, about an A.I. helper I paid a monthly fee to use rather than one that appeared to be free

It’s possible, for example, that the advertising-based models could gather so much more data to train the systems that they’d have an innate advantage over the subscription models

Much of the work of the modern state is applying the values of society to the workings of markets, so that the latter serve, to some rough extent, the former

We have done this extremely well in some markets — think of how few airplanes crash, and how free of contamination most food is — and catastrophically poorly in others.

One danger here is that a political system that knows itself to be technologically ignorant will be cowed into taking too much of a wait-and-see approach to A.I.

wait long enough and the winners of the A.I. gold rush will have the capital and user base to resist any real attempt at regulation

Somehow, society is going to have to figure out what it’s comfortable having A.I. doing, and what A.I. should not be permitted to try, before it is too late to make those decisions.

Most fears about capitalism are best understood as fears about our inability to regulate capitalism.

it’s worth acknowledging that adhering to these necessary rules cuts against some core aspects of human nature. I’m of the opinion that people should not bring their “whole self” to work — no one owes an employer that — but it’s also impossible to bring none of your personal self to work.

There are good reasons that both formal and informal boundaries are a necessity in the workplace and academia

, we know from the science of linguistics and the philosophy of knowledge that they differ profoundly from how humans reason and use language. These differences place significant limitations on what these programs can do, encoding them with ineradicable defects.

It is at once comic and tragic, as Borges might have noted, that so much money and attention should be concentrated on so little a thing — something so trivial when contrasted with the human mind, which by dint of language, in the words of Wilhelm von Humboldt, can make “infinite use of finite means,” creating ideas and theories with universal reach.

The human mind is not, like ChatGPT and its ilk, a lumbering statistical engine for pattern matching, gorging on hundreds of terabytes of data and extrapolating the most likely conversational response or most probable answer to a scientific question

the human mind is a surprisingly efficient and even elegant system that operates with small amounts of information; it seeks not to infer brute correlations among data points but to create explanations

such programs are stuck in a prehuman or nonhuman phase of cognitive evolution. Their deepest flaw is the absence of the most critical capacity of any intelligence: to say not only what is the case, what was the case and what will be the case — that’s description and prediction — but also what is not the case and what could and could not be the case

Those are the ingredients of explanation, the mark of true intelligence.

Here’s an example. Suppose you are holding an apple in your hand. Now you let the apple go. You observe the result and say, “The apple falls.” That is a description. A prediction might have been the statement “The apple will fall if I open my hand.”

an explanation is something more: It includes not only descriptions and predictions but also counterfactual conjectures like “Any such object would fall,” plus the additional clause “because of the force of gravity” or “because of the curvature of space-time” or whatever. That is a causal explanation: “The apple would not have fallen but for the force of gravity.” That is thinking.

The crux of machine learning is description and prediction; it does not posit any causal mechanisms or physical laws

any human-style explanation is not necessarily correct; we are fallible. But this is part of what it means to think: To be right, it must be possible to be wrong. Intelligence consists not only of creative conjectures but also of creative criticism. Human-style thought is based on possible explanations and error correction, a process that gradually limits what possibilities can be rationally considered.

ChatGPT and similar programs are, by design, unlimited in what they can “learn” (which is to say, memorize); they are incapable of distinguishing the possible from the impossible.

Whereas humans are limited in the kinds of explanations we can rationally conjecture, machine learning systems can learn both that the earth is flat and that the earth is round. They trade merely in probabilities that change over time.

For this reason, the predictions of machine learning systems will always be superficial and dubious.

some machine learning enthusiasts seem to be proud that their creations can generate correct “scientific” predictions (say, about the motion of physical bodies) without making use of explanations (involving, say, Newton’s laws of motion and universal gravitation). But this kind of prediction, even when successful, is pseudoscienc

While scientists certainly seek theories that have a high degree of empirical corroboration, as the philosopher Karl Popper noted, “we do not seek highly probable theories but explanations; that is to say, powerful and highly improbable theories.”

The theory that apples fall to earth because mass bends space-time (Einstein’s view) is highly improbable, but it actually tells you why they fall. True intelligence is demonstrated in the ability to think and express improbable but insightful things.

This means constraining the otherwise limitless creativity of our minds with a set of ethical principles that determines what ought and ought not to be (and of course subjecting those principles themselves to creative criticism)

True intelligence is also capable of moral thinking

To be useful, ChatGPT must be empowered to generate novel-looking output; to be acceptable to most of its users, it must steer clear of morally objectionable content

In 2016, for example, Microsoft’s Tay chatbot (a precursor to ChatGPT) flooded the internet with misogynistic and racist content, having been polluted by online trolls who filled it with offensive training data. How to solve the problem in the future? In the absence of a capacity to reason from moral principles, ChatGPT was crudely restricted by its programmers from contributing anything novel to controversial — that is, important — discussions. It sacrificed creativity for a kind of amorality.

Here, ChatGPT exhibits something like the banality of evil: plagiarism and apathy and obviation. It summarizes the standard arguments in the literature by a kind of super-autocomplete, refuses to take a stand on anything, pleads not merely ignorance but lack of intelligence and ultimately offers a “just following orders” defense, shifting responsibility to its creators.

In short, ChatGPT and its brethren are constitutionally unable to balance creativity with constraint. They either overgenerate (producing both truths and falsehoods, endorsing ethical and unethical decisions alike) or undergenerate (exhibiting noncommitment to any decisions and indifference to consequences). Given the amorality, faux science and linguistic incompetence of these systems, we can only laugh or cry at their popularity.

“My first and biggest wish is that we had known early that Covid-19 was airborne,”

, “Once you’ve realized that, it informs an entirely different strategy for protection.” Masking, ventilation and air cleaning become key, as well as avoiding high-risk encounters with strangers, he says.

Instead of washing our produce and wearing hand-sewn cloth masks, we could have made sure to avoid superspreader events and worn more-effective N95 masks or their equivalent. “We could have made more of an effort to develop and distribute N95s to everyone,” says Dr. Volckens. “We could have had an Operation Warp Speed for masks.”

We didn’t realize how important clear, straight talk would be to maintaining public trust. If we had, we could have explained the biological nature of a virus and warned that Covid-19 would change in unpredictable ways.  

We didn’t know how difficult it would be to get the basic data needed to make good public-health and medical decisions. If we’d had the data, we could have more effectively allocated scarce resources

In the face of a pandemic, he says, the public needs an early basic and blunt lesson in virology

and mutates, and since we’ve never seen this particular virus before, we will need to take unprecedented actions and we will make mistakes, he says.

Since the public wasn’t prepared, “people weren’t able to pivot when the knowledge changed,”

By the time the vaccines became available, public trust had been eroded by myriad contradictory messages—about the usefulness of masks, the ways in which the virus could be spread, and whether the virus would have an end date.

, the absence of a single, trusted source of clear information meant that many people gave up on trying to stay current or dismissed the different points of advice as partisan and untrustworthy.

“The science is really important, but if you don’t get the trust and communication right, it can only take you so far,”

people didn’t know whether it was OK to visit elderly relatives or go to a dinner party.

Doctors didn’t know what medicines worked. Governors and mayors didn’t have the information they needed to know whether to require masks. School officials lacked the information needed to know whether it was safe to open schools.

Had we known that even a mild case of Covid-19 could result in long Covid and other serious chronic health problems, we might have calculated our own personal risk differently and taken more care.

just months before the outbreak of the pandemic, the Council of State and Territorial Epidemiologists released a white paper detailing the urgent need to modernize the nation’s public-health system still reliant on manual data collection methods—paper records, phone calls, spreadsheets and faxes.

While the U.K. and Israel were collecting and disseminating Covid case data promptly, in the U.S. the CDC couldn’t. It didn’t have a centralized health-data collection system like those countries did, but rather relied on voluntary reporting by underfunded state and local public-health systems and hospitals.

doctors and scientists say they had to depend on information from Israel, the U.K. and South Africa to understand the nature of new variants and the effectiveness of treatments and vaccines. They relied heavily on private data collection efforts such as a dashboard at Johns Hopkins University’s Coronavirus Resource Center that tallied cases, deaths and vaccine rates globally.

For much of the pandemic, doctors, epidemiologists, and state and local governments had no way to find out in real time how many people were contracting Covid-19, getting hospitalized and dying

To solve the data problem, Dr. Ranney says, we need to build a public-health system that can collect and disseminate data and acts like an electrical grid. The power company sees a storm coming and lines up repair crews.

If we’d known how damaging lockdowns would be to mental health, physical health and the economy, we could have taken a more strategic approach to closing businesses and keeping people at home.

t many doctors say they were crucial at the start of the pandemic to give doctors and hospitals a chance to figure out how to accommodate and treat the avalanche of very sick patients.

The measures reduced deaths, according to many studies—but at a steep cost.

The lockdowns didn’t have to be so harmful, some scientists say. They could have been more carefully tailored to protect the most vulnerable, such as those in nursing homes and retirement communities, and to minimize widespread disruption.

Lockdowns could, during Covid-19 surges, close places such as bars and restaurants where the virus is most likely to spread, while allowing other businesses to stay open with safety precautions like masking and ventilation in place.  

The key isn’t to have the lockdowns last a long time, but that they are deployed earlier,