THE ATLANTIC – Technology | Internet & Technology

For the past 24 hours, Silicon Valley has been obsessed with a 26,000-word blog post by the AI researcher Nick Cammarata. The document has been called the “the most important essay,” “the most important result,” and “the single best piece of news” of the year, or in the history of AI research—or, perhaps, ever. Even OpenAI CEO Sam Altman lauded it. Cammarata makes a series of bold, almost astonishing claims about the nature of AI that, if true, would upend some of his industry’s basic assumptions about the very thing it is building. And after months of AI-hacking sprees and mounting public backlash against the industry, many people see Cammarata as offering a glimmer of hope when it comes to not just understanding but also controlling the technology being unleashed on the world.Boiled down, Cammarata’s main contention is that generative-AI models are not actually black boxes, contrary to the common perception that these systems are so big and labyrinthine as to be nearly impenetrable. In fact, Cammarata argues, there is a relatively clear path toward understanding them. If that’s the case, the material significance could be tremendous: It would provide a way to map AI models’ inner workings and, in turn, perhaps prevent them from wreaking all sorts of harm—autonomously hacking companies and sovereign nations, encouraging users’ delusions, appearing to deceive AI researchers themselves. There is an entire field of research, called “mechanistic interpretability,” devoted to the study of just this.The argument is just as important as the person making it. Cammarata, a former OpenAI employee, is a respected figure in his field. Years ago, he played a key role in establishing the opposite claim, which is now common wisdom: that neural networks, the software architecture that generative AI is built on, are impossible, or at least very difficult, to understand. In this new essay, Cammarata, who now works at the AI-research company Goodfire, essentially says that he and his entire field have been wrong ever since. “The claims here are viscerally shocking,” one observer wrote. Or, as another wrote on X, they are “big big big big.”[Read: Imagine every AI agent on Earth buying the same cheap paper towels]Neural networks, inspired by the human brain, consist of layered connections among individual nodes, or “neurons,” that activate and send information to one another when responding to a query. Earlier research, including Cammarata’s, established that individual machine neurons can activate in… [TheTopNews] Read More.

Close your eyes, and imagine a frictionless day.You wake up to a notification that a grocery-store employee will deliver all of the ingredients for the three recipes you saved last night to your “Weeknight Dinner” folder. While you brew your coffee, a program browses online retailers to compare toilet-paper prices, and another solicits competitive bids for Sheetrock work you need done in the garage. During your commute, a bot moves your upcoming dentist appointment to avoid a scheduling conflict. On your lunch break, you get an alert that computers acting on your behalf noticed some unclaimed funds and class-action payouts in a state database and claimed them. Also, the machine has negotiated a cheaper internet plan. By the time you get home, the groceries are on your doorstep, and you’ve gotten back $150.This is the future that OpenAI, Meta, xAI, Google, and the rest are selling with a spate of always-on AI assistants that work for you. “Imagine every person on earth with a second mind beyond their own,” Meta’s chief AI officer, Alexandr Wang, wrote of the company’s new AI helper, Muse. A few weeks after Muse’s launch, OpenAI launched its competing product, Dots, which Sam Altman claimed would let you “get more of your time and attention back.”[Read: The singularity is not what it seems]These AI agents promise to bulldoze the inefficiencies of daily life. Think about all of the opportunities you’re currently leaving on the table: You don’t change cell-service providers for a flash sale because of the paperwork involved. You haven’t fiddled with your 401(k) investments in years. You have kids to pick up, a date to dress for, sleep to catch up on. But your AI agents are different—they’re relentless, tireless, and shameless. They work in the background, without supervision, eagerly helping with all of these tasks and a bunch you didn’t even ask for. They use every tactic and loophole at their disposal to get you the best price and the highest-quality service because they’re able to pay attention to everything at once.The AI companies see a massive opportunity for hyper-personalized agents to become a new intermediary layer for the internet and the physical world. This future is already unfolding: People are giving Muse access to their phone number, email, credit cards, and bank accounts to make dinner reservations, contact customer service, and book vacations. Mark Zuckerberg says that he is using Muse to… [TheTopNews] Read More.

For today’s high schoolers, a typical school day from a decade ago must seem like something from the Dark Ages. These days, students are issued digital textbooks instead of physical ones, they learn how to conduct chemistry experiments through apps that simulate a science lab, and they take standardized tests on a Chromebook. All of this is the result of many years of frenzied investment in ed tech, capped by a pandemic that forced virtual learning. The overwhelming majority of schools nationwide provide each student with their own laptop or tablet.All of these screens have changed the dynamic of the classroom. When a student is supposed to be using their school-issued iPad to diligently take notes during math class, they might instead be messaging a friend. Across the country, some parents and teachers are mounting a backlash against screens in the classroom. In April, the Los Angeles school board voted to stop giving devices to its youngest students and implemented screen-time limits for all grades.But other schools are pulling back on ed tech for an entirely different reason. Chromebooks, like essentially all other consumer electronics, need RAM, or random-access memory. So do the data centers that power every conversation with a chatbot. The AI industry is using so many memory chips that it’s led to a global shortage, making gadgets much more expensive. For an individual consumer, the memory crunch—RAMaggedon, if you will—means having to spend a few hundred dollars more on a new computer. But school districts, many of which are already cash-strapped, buy hundreds or thousands of computers every year. The administrators I spoke with for this story told me that they have had to scramble to acquire all the electronics they now need. As RAMaggedon drags on, they said, something has got to give.[Read: If you need a laptop, buy it now]Schools started picking up on the memory shortage last winter, months before most of the country. They have to figure out bulk orders for laptops and tablets long before the academic year begins. Many of them turn to third-party vendors, which have typically provided quotes that are locked in for a month, giving technology directors time to secure the funds from their districts to make purchases that can run millions of dollars. As memory prices have soared, Mark Samberg, the chief technology officer of Chatham County Schools in North Carolina, is hearing a different message from… [TheTopNews] Read More.

A man walks into a restaurant. He doesn’t have a reservation, but that shouldn’t be a problem; there are ample seats at the bar, visible just over the hostess’s shoulder. “Of course,” she says, smiling. “We can seat you right away.” As she gathers menus, he absorbs the pleasant hum of a busy dining room, breathes in the tantalizing smells from the semi-open kitchen. Then, the hostess stops, Columbo-like, to interrupt with one more question: “And what’s your phone number?”The man was me. The restaurant was the Kingsway, a newish, Michelin-recommended spot in New Orleans. I’m not going to give you my phone number, but I gave it to the hostess, almost without thinking. And then, I did think about it: a lot. Throughout the meal, and for a long time afterward. There had been no reservation to check, no rewards program, no waiting list. So why the automatic ask?Let me say quickly that I liked the Kingsway very much. It’s coolly fancy, with a pan-Asian menu evoking the journey of the chef-owner Ashwin Vilkhu’s family from South Asia to the Gulf South. The result brings to mind an earlier era’s scruffy raiding of the global pantry, but this time executed with thought, technique, and precision. As I tucked into a tangle of Singapore-style chili-crab noodles, the term stoner soigné passed through my head.But that very first interaction with the hostess colored my meal. I brooded. What if I had just wanted to sit at the bar anonymously? Was that not one of life’s great pleasures? (I was working up a head of steam.) Was there no place in this enshittified world where a man could get a break?[Read: You will never get that restaurant reservation]It mattered little that I knew the answer. Like pretty much everyone else, I am well aware that my information is strip-mined wherever I go, in more ways than I can count and in many that I will never know. Like everyone else, I spend my days downloading apps to operate things such as toothbrushes, unsubscribing from marketing lists by somehow giving up more personal information, rejecting cookies like I’m on a gluten-free diet.Nor am I overly romantic about restaurants, which I know have been collecting data for years. The reservation platform OpenTable began making algorithmic recommendations to users as early as 2014; by 2017, The New York Times reported that Big Data was “staging… [TheTopNews] Read More.

The second floor of OpenAI’s old headquarters had a theme. Each room was named after one of the most important inventions in human history: “Clock,” “Fire,” and, of course, “AGI,” or artificial general intelligence. I saw this in the spring of 2024, when I was interviewing some of the company’s researchers about recent technical advances. At the time, ChatGPT was still making embarrassing arithmetic errors, and the notion that the bot would ever rival human intelligence was dubious. But here, in the renovated shell of a former mayonnaise factory in San Francisco, AGI’s arrival was predestined.This belief in AGI—a computer program with capabilities that rival or even exceed those of the human mind—is foundational to the mission of Google DeepMind, OpenAI, and Anthropic, three of the most influential AI companies in the world. And the pursuit of that milestone is the subject of The AGI Chronicles, a new book by the tech reporter and Atlantic contributing writer Kevin Roose, which he describes as his attempt to document the decade “when the human monopoly on advanced intelligence ended, and a new world was born.” Roose, a prominent and well-connected commenter on the AI boom, follows the internecine world of the frontier AI companies in remarkable, at times almost insular, detail. Based on interviews with more than 150 researchers, executives, investors, and policy makers close to the technology’s development, The AGI Chronicles traces crucial moments across DeepMind, OpenAI, and Anthropic from 2015 through early 2026 to figure out how we got to this very messy moment in AI, and who’s responsible for what happens next.[Read: The singularity is not what it seems]If you scan statements from executives such as Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman, there is a sense of fatalism about the rise of AI. In my conversation with Amodei last year, he likened developing AI to operating an out-of-control bus: It can’t be stopped or even slowed, but perhaps it can be steered. In an interview last month, Altman called AI “an inevitability.” According to this framing, AI executives are simply conductors, doing their best to provide guardrails for an era that’s careering our way, whether anyone likes it or not. Roose’s book prompted me to consider a question that is often neglected in the mile-a-minute news cycle: Are the leading AI companies simply heralding an inevitable evolution of this technology, as many of them claim, or… [TheTopNews] Read More.