More stories

  • in

    Can Math Help AI Chatbots Stop Making Stuff Up?

    Chatbots like ChatGPT get stuff wrong. But researchers are building new A.I. systems that can verify their own math — and maybe more.On a recent afternoon, Tudor Achim gave a brain teaser to an A.I. bot called Aristotle.The question involved a 10-by-10 table filled with a hundred numbers. If you collected the smallest number in each row and the largest number in each column, he asked, could the largest of the small numbers ever be greater than the smallest of the large numbers?The bot correctly answered “No.” But that was not surprising. Popular chatbots like ChatGPT may give the right answer, too. The difference was that Aristotle had proven that its answer was right. The bot generated a detailed computer program that verified “No” was the correct response.Chatbots like ChatGPT from OpenAI and Gemini from Google can answer questions, write poetry, summarize news articles and generate images. But they also make mistakes that defy common sense. Sometimes, they make stuff up — a phenomenon called hallucination.Mr. Achim, the chief executive and co-founder of a Silicon Valley start-up called Harmonic, is part of growing effort to build a new kind of A.I. that never hallucinates. Today, this technology is focused on mathematics. But many leading researchers believe they can extend the same techniques into computer programming and other areas.Because math is a rigid discipline with formal ways of proving whether an answer is right or wrong, companies like Harmonic can build A.I. technologies that check their own answers and learn to produce reliable information.Google DeepMind, the tech giant’s central A.I. lab, recently unveiled a system called AlphaProof that operates in this way. Competing in the International Mathematical Olympiad, the premier math competition for high schoolers, the system achieved “silver medal” performance, solving four of the competition’s six problems. It was the first time a machine had reached that level.We are having trouble retrieving the article content.Please enable JavaScript in your browser settings.Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.Thank you for your patience while we verify access.Already a subscriber? Log in.Want all of The Times? Subscribe. More

  • in

    Will A.I. Be a Bust? A Wall Street Skeptic Rings the Alarm.

    Jim Covello, Goldman Sachs’s head of stock research, warned that building too much of what the world doesn’t need “typically ends badly.”As Jim Covello’s car barreled up highway 101 from San Jose to San Francisco this month, he counted the billboards about artificial intelligence. The nearly 40 signs he passed, including one that promoted something called Writer Enterprise AI and another for Speech AI, were fresh evidence, he thought, of an economic bubble.“Not that long ago, they were all crypto,” Mr. Covello said of the billboards. “And now they’re all A.I.”Mr. Covello, the head of stock research at Goldman Sachs, has become Wall Street’s leading A.I. skeptic. Three months ago, he jolted markets with a research paper that challenged whether businesses would see a sufficient return on what by some estimates could be $1 trillion in A.I. spending in the coming years. He said that generative artificial intelligence, which can summarize text and write software code, makes so many mistakes that it was questionable whether it would ever reliably solve complex problems.The Goldman paper landed days after a partner at Sequoia Capital, a venture firm, raised similar questions in a blog post about A.I. Their skepticism marked a turning point for A.I.-related stocks, leading to a reassessment of Wall Street’s hottest trade.Goldman’s basket of A.I. stocks, which is managed by a separate arm of the firm and includes Nvidia, Microsoft, Apple, Alphabet, Amazon, Meta and Oracle, has declined 7 percent from its peak on July 10, as investors and business leaders debate whether A.I. can justify its staggering costs.The pause has come early in the A.I. arms race. The tech industry has a history of spending big to deliver technology transitions, as it did during the personal computer and internet revolutions. Those build outs spanned five years or more before there was a reckoning.We are having trouble retrieving the article content.Please enable JavaScript in your browser settings.Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.Thank you for your patience while we verify access.Already a subscriber? Log in.Want all of The Times? Subscribe. More

  • in

    MAGA Wants Transgression, and This Is What Comes With It

    The North Carolina Republican Party is facing one of the most predictable crises in the history of party politics.Its primary voters enthusiastically supported a candidate for governor named Mark Robinson — voting for him by a more than 45-point margin over his closest rival (he won by 64.8 percent to 19.2 percent) — even though he had a remarkable record of deeply inflammatory and even unhinged statements.Last week, a comprehensive CNN report unearthed compelling evidence that Robinson had posted on a porn site called Nude Africa. I cannot possibly repeat the worst posts, but the less graphically obscene ones included statements like this: “I’m a Black Nazi,” and “Slavery is not bad. Some people need to be slaves. I wish they would bring it back. I would certainly buy a few.”That’s not all. “I’m not in the K.K.K.,” he also said, according to the CNN report. “They don’t let Blacks join. If I was in the K.K.K. I would have called him Martin Lucifer Koon!” He said he’d prefer Hitler to what he sees in Washington today.No one, however, should be surprised. Even before the primary, Robinson’s horrific character was on display. Among other things, he had called school shooting survivors who advocated gun control “media prosti-tots,” accused Michelle Obama of being a man, and trafficked in so many antisemitic tropes that his election as lieutenant governor in 2020 was an alarm bell for Jewish leaders in the state.In other words, Republican voters knew he was a bad man when they chose him. Now they know he is a very bad man.We are having trouble retrieving the article content.Please enable JavaScript in your browser settings.Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.Thank you for your patience while we verify access.Already a subscriber? Log in.Want all of The Times? Subscribe. More

  • in

    Qualcomm Asked Rival Intel if It Would Consider Sale

    While Intel has struggled in recent years, other chipmakers are thriving because of a boom in demand.The chipmaker Qualcomm has approached its rival Intel in recent days about the possibility of acquiring the slumping Silicon Valley giant, two people familiar with the matter said Friday, requesting anonymity because the talks were confidential.Qualcomm has not yet made an official offer for Intel, one of the people said, and the obstacles to a deal remain steep. Any deal would likely draw significant regulatory scrutiny, given the mammoth size and national security importance of both chip companies. It is unclear whether regulators would allow Qualcomm to buy Intel without taking on its struggling foundry business, and it remains equally unclear whether Qualcomm would want to take on that complex endeavor. A deal would also be costly. Intel, which has seen its shares fall nearly 40 percent over the last year, has a market capitalization of $93 billion. Qualcomm, which has seen its shares rise 55 percent, has a market value of $169 billion. Qualcomm and Intel, through spokeswomen, both declined to comment. The Wall Street Journal earlier reported Qualcomm’s approach. That any chip-making rival would consider trying to buy Intel would have been inconceivable a decade ago. But years of management issues and whiffs on technology transitions have weakened what was once one of Silicon Valley’s most powerful companies.Intel missed out on selling chips for mobile phones and has failed to capitalize on the boom in artificial intelligence, a field rival Nvidia now dominates with specialized chips used in data centers. Intel’s chip manufacturing operations, once the most advanced, also lost a technology lead to Taiwan Semiconductor Manufacturing Company.Intel’s problems were underscored in early August, when it announced a $1.6 billion quarterly loss and plans to cut 15,000 jobs. The company, the largest planned recipient of federal financing under legislation called the CHIPS Act, on Monday announced other moves that include plans to pause the setting up of new plants in Germany and Poland.Qualcomm, based in San Diego, is a leader in cellular technology and provides chips used in flagship smartphones from companies such as Apple and Samsung Electronics. Unlike Intel, Qualcomm has never operated factories, a costly business that most chip designers avoid. So it would seem more likely to be interested in the Intel operations that design chips, as well as its broad expertise in PC software and channels for selling those systems, said Patrick Little, a former Qualcomm executive who now is chief executive of SiFive, a Silicon Valley start-up that sells rival microprocessor designs.“Those are things Qualcomm would have to mature on their own over time,” Mr. Little said. “If they worked with or somehow had a piece of Intel that could accelerate that part of their strategy.”Any effort to buy Intel would likely face a tough antitrust review and would be scrutinized closely on national security grounds, since its design and manufacturing operations are important for defense applications and overall U.S. competitiveness in semiconductors.Lauren Hirsch reported from New York and Don Clark from San Francisco. More

  • in

    Telegram Founder Pavel Durov Defends App in First Comments Since Arrest

    Pavel Durov, held in France since last month, blamed “growing pains” for the company’s problems and pledged to make improvements.Pavel Durov, the founder of the online communications tool Telegram, said on Thursday that it was a “misguided approach” to hold him personally responsible for the spread of illicit content on the platform.Mr. Durov’s comments, made on his Telegram account, were his first public remarks since he was arrested at an airport outside Paris and charged last month by the French authorities for failing to prevent illegal activity on the app. The crimes on Telegram included the spread of child sexual abuse material, fraud and drug sales, French prosecutors have said.“No innovator will ever build new tools if they know they can be personally held responsible for potential abuse of those tools,” Mr. Durov wrote. He said “growing pains” on Telegram, which has 950 million users, had made it easy for criminals to abuse the platform.“That’s why I made it my personal goal to ensure we significantly improve things in this regard,” he said.Mr. Durov’s case has become a point of contention in the politically charged debate over the limits of free speech on the internet. Telegram is committed to light supervision of what people say or do on the platform. The app has helped people living under authoritarian governments communicate and organize. But it has also become a hothouse for disinformation, extremism and other harmful content.Telegram has long been in the cross hairs of global law enforcement agencies because it has refused to cooperate with the authorities. French prosecutors said Mr. Durov had been arrested in part because of Telegram’s “almost total lack of response” to requests related to criminal investigations.We are having trouble retrieving the article content.Please enable JavaScript in your browser settings.Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.Thank you for your patience while we verify access.Already a subscriber? Log in.Want all of The Times? Subscribe. More

  • in

    Start-Up Investors Push Back Against Venture Capital’s Bigger-Is-Better Mantra

    A small but vocal group is forming new funds and taking new approaches to counter the swell of money into venture capital in recent years.After nearly 10 years running his own venture capital firm, Nick Chirls decided to call it quits this year.His firm, Notation Capital, had raised three funds and invested in more than 100 companies. But Mr. Chirls said he had become disillusioned as venture capital grew from a collection of small partnerships into an industry dominated by firms that managed enormous sums.The focus on accumulating and deploying as much money as possible “completely dehumanized the entire business,” he said.Instead, Mr. Chirls is starting a new kind of firm. From the outside, the endeavor, Asylum Ventures, looks like his old firm, with a $55 million venture fund that will invest in very young tech companies. But the approach is set to be very different, making fewer investments over a longer period in companies that will not need to raise increasingly large funding rounds, he said.Mr. Chirls and his partners, Jonathan Wu and Mackenzie Regent, are part of a small but vocal group of start-up investors who are pushing back against venture capital’s changing scope and priorities. Venture capital investing has traditionally involved small groups of financiers who backed very young, very risky companies that couldn’t obtain traditional loans. The sums invested were often small.But that changed in recent years as investors poured billions of dollars into unproven start-ups with little diligence and investment firms expanded rapidly into new strategies and geographies. Last year, venture capital managed $1.1 trillion, up from $297 billion in 2013, according to PitchBook, which tracks start-ups.We are having trouble retrieving the article content.Please enable JavaScript in your browser settings.Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.Thank you for your patience while we verify access.Already a subscriber? Log in.Want all of The Times? Subscribe. More

  • in

    Brazil Threatens to Ban Elon Musk’s X

    The country’s Supreme Court gave the service 24 hours to name a legal representative in Brazil or face suspension.Brazil’s Supreme Court on Wednesday gave Elon Musk 24 hours to name a legal representative for X in Brazil or face a ban of his social network across the nation of 200 million.Mr. Musk closed X’s office in Brazil last week in protest of orders from a Brazilian Supreme Court justice to suspend certain accounts. If X refuses to comply, it could lose access to one of its largest markets outside the United States — a blow as the company struggles to regain revenue after Mr. Musk battled with advertisers and told them not to spend on the platform.The court posted its order on X on Wednesday night, suggesting that Mr. Musk had until about 8 p.m. local time Thursday to respond.The moment is one of the biggest tests yet for Mr. Musk’s efforts to mold X to his personal ideology, and how he responds will shed light on how far he is willing to take his stated commitment to protecting his social network from what he calls censorship.X and Mr. Musk did not immediately respond to requests for comment on the possible ban.Mr. Musk has been enmeshed in a monthslong feud with Alexandre de Moraes, a Brazilian Supreme Court justice whom he has accused of censoring conservative voices online. Justice Moraes has ordered the suspension of more than 100 X accounts in what he says is a battle against misinformation, hate speech and attacks on democracy.Most of the accounts that Justice Moraes has targeted belong to right-wing supporters of Jair Bolsonaro, Brazil’s former president. Some of them questioned Mr. Bolsonaro’s 2022 election loss and sympathized with protesters who raided Brazil’s halls of power, hoping to invoke a military takeover.We are having trouble retrieving the article content.Please enable JavaScript in your browser settings.Thank you for your patience while we verify access. If you are in Reader mode please exit and log into your Times account, or subscribe for all of The Times.Thank you for your patience while we verify access.Already a subscriber? Log in.Want all of The Times? Subscribe. More