This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
A disaster is brewing on dark-web forums, in messaging apps, and in schools around the world: Generative AI is being used to create sexually explicit images and videos of children, likely thousands a day.
When the Three Mile Island power plant in Pennsylvania was decommissioned in 2019, it heralded the symbolic end of America’s nuclear industry. In 1979, the facility was the site of the worst nuclear disaster in the nation’s history: a partial reactor meltdown that didn’t release enough radiation to cause detectable harm to people nearby, but still turned Americans against nuclear power and prompted a host of regulations that functionally killed most nuclear build-out for decades.
After weeks of speculation about a new and more powerful AI product in the works, OpenAI today announced its first “reasoning model.” The program, known as o1, may in many respects be OpenAI’s most powerful AI offering yet, with problem-solving capacities that resemble those of a human mind more than any software before. Or, at least, that’s how the company is selling it.
As with most OpenAI research and product announcements, o1 is, for now, somewhat of a tease.
This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
Today’s generative-AI tools can concoct stunning designs and playful prose with the push of a few buttons. That, in turn, has bred fears about how the technology could hurt human artists and writers, and led many, in their defense of humanity, to a well-intentioned but confused claim.
This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
Year three of AI college is about to begin, and instructors across the country still seem to have no clue how to handle the technology: no good way to stop students from using ChatGPT to write essays, and no clear way to instruct students on how AI might enhance their work.
At a rally this past April in Michigan, surrounded by a cadre of law-enforcement officials, Donald Trump suddenly began railing against electric cars. President Joe Biden’s decision to support EVs, he decried, “is one of the dumbest I’ve ever heard.” Minutes later, he was back to praising the sheriffs behind him: “We have to get law and order back. These are the best people in the world,” he said to a smattering of applause.
This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
For the past two years, the biggest tech firms have begun spending historic amounts of money to develop generative-AI products and spread them across the web. The build-out may demand a trillion dollars or more of investment across the economy this decade—more than the Apollo missions or the interstate-highway system.
Every second of every day, people across the world type tens of thousands of queries into Google, adding up to trillions of searches a year. Google and a few other search engines are the portal through which several billion people navigate the internet. Many of the world’s most powerful tech companies, including Google, Microsoft, and OpenAI, have recently spotted an opportunity to remake that gateway with generative AI, and they are racing to seize it.
This is Atlantic Intelligence, a newsletter in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
Yesterday OpenAI made what should have been a triumphant entry into the AI-search wars: The start-up announced SearchGPT, a prototype tool that can use the internet to answer questions of all kinds. But there was a problem, as I reported: Even the demo got something wrong.
Whenever AI companies present a vision for the role of artificial intelligence in the future of searching the internet, they tend to underscore the same points: instantaneous summaries of relevant information; ready-made lists tailored to a searcher’s needs. They tend not to point out that generative-AI models are prone to providing incorrect, and at times fully made-up, information—and yet it keeps happening.
Overnight, much of the world ground to a halt: Tens of thousands of flights and trains were canceled or delayed, hospitals stopped elective surgeries, doctors couldn’t book appointments, banks struggled to process transactions, television networks stopped broadcasting. The culprit was not a war, an earthquake, a mounting heat wave, or a terrorist attack, but some faulty computer code. It was likely the largest IT failure in history.
For more than 40 years in a row, Ford’s F-150 and its family of pickup trucks have been the best-selling vehicles in America. So when Ford released an electric version in 2022, the F-150 Lightning, it should have been a turning point for electric cars in the country—if not, that is, for the price tag.
Today, at Apple’s annual developers conference—where new software products are previewed in slick video presentations—the company finally joined the generative-AI race. The company introduced Apple Intelligence, a suite of AI features that will be rolled out to the tech giant’s latest operating systems starting this fall. New generative-AI models will help Apple users write work memos and highly personalized text; create images and emoji; connect and organize photos, calendar events, and emails.
In April, a group of people in a red Tesla driving through the Moroccan desert were glued to the odometer on the car’s giant touch screen. “Two million, Hans! Two million,” exclaimed the front-seat passenger to the owner and driver, Hansjörg von Gemmingen-Hornberg. His 2014 Model S had become likely the first electric vehicle to drive 2 million kilometers, or more than 1.2 million miles. The car could have traveled from the Earth to the moon and back, twice, then circled the equator 11 times.
OpenAI appears to be in the midst of a months-long revolt from within. The latest flash point came yesterday, when a group of 11 current and former employees—plus two from other firms—issued a public letter declaring that leading AI companies are not to be trusted. “The companies are behaving in a way that is really not in the public interest,” William Saunders, a signatory who, like several others on the letter, left OpenAI earlier this year, told me.
This is Atlantic Intelligence, a limited-run series in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
This week has felt like the early days of the generative-AI boom, filled with dazzling events concerning the future of the technology.
On Monday, OpenAI held a last-minute “Spring Update” event in which the company announced its newest AI model, GPT-4o, in an impressive live demo.
Earlier today, Google presented a new vision for its flagship search engine, one that is uniquely tailored to the generative-AI moment. With advanced technology at its disposal, “Google will do the Googling for you,” Liz Reid, the company’s head of search, declared onstage at the company’s annual software conference.
Googling something rarely yields an immediate, definitive answer.
Earlier today, OpenAI announced its newest product: GPT-4o, a faster, cheaper, more powerful version of its most advanced large language model, and one that the company has deliberately positioned as the next step in “natural human-computer interaction.
Moments after receiving my lunch order, the robots whirred to life. A clawlike contraption lurched forward, like a bird pecking at feed, to snatch dishes holding a faux-chicken cutlet and potatoes, then inserted them onto a metal track that snakes through a 650-degree-Fahrenheit oven. Seven minutes, some automatic food dispensers, and two conveyor belts later (with a healthy assist from human hands), my meal was sitting on a shelf of mint-green cubbies.
Recently, Bonaventure Dossou learned of an alarming tendency in a popular AI model. The program described Fon—a language spoken by Dossou’s mother and millions of others in Benin and neighboring countries—as “a fictional language.”
This result, which I replicated, is not unusual. Dossou is accustomed to the feeling that his culture is unseen by technology that so easily serves other people.
This is Atlantic Intelligence, a limited-run series in which our writers help you wrap your mind around artificial intelligence and a new machine age. Sign up here.
In the first few months after the release of ChatGPT, AI chatbots felt, to many, like magic: They conjured poems and cocktail recipes, and secretly did at least one writer’s job. These programs appeared to be the first nonhuman entity to master human language, and many people ascribed them with intelligence, even sentience.
Yesterday afternoon, Elon Musk fired the latest shot in his feud with OpenAI: His new AI venture, xAI, now allows anyone to download and use the computer code for its flagship software. No fees, no restrictions, just Grok, a large language model that Musk has positioned against OpenAI’s GPT-4, the model powering the most advanced version of ChatGPT.
Sharing Grok’s code is a thinly veiled provocation. Musk was one of OpenAI’s original backers.
Artificial intelligence can appear to be many different things—a whole host of programs with seemingly little common ground. Sometimes AI is a conversation partner, an illustrator, a math tutor, a facial-recognition tool. But in every incarnation, it is always, always a machine, demanding almost unfathomable amounts of data and energy to function.
AI systems such as ChatGPT operate out of buildings stuffed with silicon computer chips.
Apple has long been a company shrouded in mystery—and perhaps its most mysterious recent project of all has been the Apple Car. For a decade, reports have continually emerged about the company building an electric, autonomous car as smart and beautiful as an iPhone. The company never confirmed the project’s existence, even as it hired executives from Lamborghini and BMW. MotorTrend created a speculative Apple Car in 2016, treating it like an object of science fiction, then another in 2022.
The first few steps on the path toward living forever alongside the longevity enthusiast Bryan Johnson are straightforward: “Go to bed on time, eat healthy food, and exercise,” he told a crowd in Brooklyn on Saturday morning. “But to start, you guys are now going to do a breathing exercise.” He directed the 100-plus people gathered around him to put their hands on the shoulders of their neighbors, forming a series of concentric circles; he then counted as we inhaled and exhaled in unison.
Yesterday afternoon, OpenAI teased Sora, a video-generation model that promises to convert written text prompts into highly realistic videos. Footage released by the company depicts such examples as “a Shiba Inu dog wearing a beret and black turtleneck” and “in an ornate, historical hall, a massive tidal wave peaks and begins to crash.
Science fiction, when revisited years later, sometimes doesn’t come across as all that fictional. Speculative novels have an impressive track record at prophesying what innovations are to come, and how they might upend the world: H. G. Wells wrote about an atomic bomb decades before World War II, and Ray Bradbury’s 1953 novel, Fahrenheit 451, features devices we’d describe today as Bluetooth earbuds.
If you squint and tilt your head, you can see some similarities in the blurry shapes that are Harvard and OpenAI. Each is a leading institution for building minds, whether real or artificial—Harvard educates smart humans, while OpenAI engineers smart machines—and each has been forced in recent days to stare down a common allegation. Namely, that they are represented by intellectual thieves.
At the turn of the century, when the modern web was just emerging and Microsoft was king, a small but growing technology movement posed an existential threat to the company. Steve Ballmer, Microsoft’s CEO at the time, called one of its core elements “a cancer that attaches itself” to “everything it touches.
This is Atlantic Intelligence, an eight-week series in which The Atlantic’s leading thinkers on AI will help you understand the complexity and opportunities of this groundbreaking technology. Sign up here.The bedrock of the AI revolution is the internet, or more specifically, the ever-expanding bounty of data that the web makes available to train algorithms.