विज्ञान

By Stephen Nellis SAN FRANCISCO (Reuters) -IBM said on Friday it is able to run a key quantum computing algorithm on commonly available chips from Advanced Micro Devices, in a step toward commercializing super-powerful computers. The U.S. stalwart is racing to develop quantum computing against Microsoft and Alphabet's Google, which announced a breakthrough algorithm this week. Quantum computers use what are known as qubits to tackle problems that would take conventional computers thousands of years to crack – problems such as how trillions of atoms react over time. However, qubits are prone to errors that can quickly overwhelm the useful computing work of a quantum chip. In June, IBM said it had developed an algorithm to run alongside quantum chips that can address such errors. In a research paper seen by Reuters to be published Monday, IBM will show that it can run those algorithms in real time on a type of chip called a field programmable gate array manufactured by AMD. Jay Gambetta, the IBM vice president running the firm's quantum efforts, said the work showed that IBM's algorithm not only works in the real world, but can operate on a readily available AMD chip that is not "ridiculously expensive."  "Implementing it, and showing that the implementation is actually 10 times faster than what is needed, is a big deal," Gambetta said in an interview. IBM has a multi-year plan to build a quantum computer called Starling by 2029. Gambetta said the algorithm work disclosed Friday was completed a year ahead of schedule. (Reporting by Stephen Nellis in San Francisco; Editing by Christopher Cushing)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Stephen Nellis SAN FRANCISCO (Reuters) -IBM said on Friday it is able to run a key quantum computing algorithm on commonly available chips from Advanced Micro Devices, in a step toward commercializing super-powerful computers. The U.S. stalwart is racing to develop quantum computing against Microsoft and Alphabet's Google, which announced a breakthrough algorithm this week. Quantum computers use what are known as qubits to tackle problems that would take conventional computers thousands of years to crack – problems such as how trillions of atoms react over time. However, qubits are prone to errors that can quickly overwhelm the useful computing work of a quantum chip. In June, IBM said it had developed an algorithm to run alongside quantum chips that can address such errors. In a research paper seen by Reuters to be published Monday, IBM will show that it can run those algorithms in real time on a type of chip called a field programmable gate array manufactured by AMD. Jay Gambetta, the IBM vice president running the firm's quantum efforts, said the work showed that IBM's algorithm not only works in the real world, but can operate on a readily available AMD chip that is not "ridiculously expensive."  "Implementing it, and showing that the implementation is actually 10 times faster than what is needed, is a big deal," Gambetta said in an interview. IBM has a multi-year plan to build a quantum computer called Starling by 2029. Gambetta said the algorithm work disclosed Friday was completed a year ahead of schedule. (Reporting by Stephen Nellis in San Francisco; Editing by Christopher Cushing)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

(Corrects formatting of company ticker symbol in first paragraph) By Rashika Singh (Reuters) -Intel shares surged nearly 9% in premarket trading on Friday as investors rallied behind CEO Lip-Bu Tan's aggressive cost-cutting measures that helped the chipmaker surpass quarterly profit estimates and regain stability amid a flurry of high-stakes bets on future growth. The results mark a turning point for Intel, which has struggled to maintain relevance in the face of fierce competition and manufacturing setbacks. After a bruising 2024 that saw its first annual loss in nearly four decades, the company is now leaning on strategic investments and operational discipline to rebuild investor confidence. STEADYING THE SHIP Intel also drew support during the quarter from multi-billion-dollar investments by Nvidia and Japan's SoftBank as well as a U.S. government stake, moves that offered a financial cushion as it works to revive growth. These investments, along with Tan's turnaround efforts, have offered a lifeline to the stock, which has rebounded by more than 90% in 2025, outperforming AI chip leaders Nvidia and AMD. Intel trades at a 12-month forward price-to-earnings ratio of 71.51 versus 30.49 for Nvidia and 40.14 for AMD. "Intel has turned a corner and is steadying the ship," said Ben Bajarin, CEO of Creative Strategies. "It feels like a strong setup for 2026." TURNAROUND FAR FROM OVER Intel said demand for its chips was outpacing supply, particularly in data centers where operators are upgrading central processing units (CPUs) to support AI workloads. However, finance chief Dave Zinsner cautioned that yields for its advanced 18A manufacturing process will remain below industry standards and won't reach "acceptable levels" until 2027. Tan has also sold a majority stake in Altera and shifted Intel's capital strategy to rely more on external commitments, following criticism of his predecessor's spending-heavy approach. He has pared back Intel's manufacturing ambitions and cut over 20% of the workforce. "We understand the desire to claim victory for the embattled company, but this fight is far from over; perhaps it's better to call it a draw for now," analysts at Bernstein said. (Reporting by Rashika Singh and Arsheeya Bajwa in Bengaluru; additional reporting by Alun John in London; Editing by Joice Alves and Saumyadeb Chakrabarty)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Aditya Kalra NEW DELHI (Reuters) -An Apple fee of up to 30% in India will over time stifle revenues for Match, the owner of dating app Tinder, it said in its submission to the Indian antitrust body, adding that hefty fines were needed to protect its business. Tinder is one of the biggest players in the Indian dating apps market, which will be worth an estimated $1.42 billion by 2030, according to advisory firm MarketNtel. Since 2022, Match has been locked in an antitrust battle with Apple at the Competition Commission of India, where investigators last year issued a report saying the U.S. smartphone company had engaged in "abusive conduct" on the apps market of its iPhone Operating System, iOS. Apple has denied wrongdoing and said it is a small player in India where phones mostly use Google's Android system. It did not respond to a request on Friday for comment.  Match and the CCI also did not respond to Reuters queries.  In a response dated October 13 to the CCI investigators' findings, which Reuters is the first to report on, Match urged the watchdog to issue final directives asking Apple to discontinue its anti-competitive practices. "Apple's policies adversely affect the return on capital and revenue streams of Match’s portfolio brands," Match said in its submissions, reviewed by Reuters. "In the long term, the constraints and excessive fee imposed by Apple will stifle the growth and scalability of Match’s portfolio brands," it added, without quantifying the impact. As Apple faces antitrust scrutiny across the world, on Thursday, a London tribunal ruled it abused its dominant position by charging app developers unfair commissions. 'TINDER AND UBER BOTH OFFER MATCHMAKING' Apple's smartphone base has become four times larger in the last five years in India. Its iOS powered an estimated 4.5% of 735 million smartphones in India as of mid-2025, with the rest using Android, Counterpoint Research says.  The CCI has powers to fine Apple up to 10% of its global average turnover for the last three financial years.  Such a fine based on global turnover, not Indian, could "act as a significant deterrent against recidivism," Match said in its submission. The CCI investigation team found last year that Apple was not permitting any third-party payment processor to provide the services for in-app purchases, where the fee could be up to 30%. Match's latest submission cited discriminatory treatment, saying Uber's ride-hailing app was classified by Apple as providing physical goods, and was charged lower commission than Tinder, which is treated as providing digital services. "Tinder and Uber both offer matchmaking services in essence," Match said in its filing. In April, the EU watchdog said Apple must remove restrictions that prevent app developers from steering users to cheaper deals outside the App Store, and fined the company 500 million euros ($583 million). Apple in June changed its App Store rules to comply with the EU order. (Reporting by Aditya Kalra; additional reporting by Munsif Vengattil; editing by Barbara Lewis)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

BRUSSELS (Reuters) -The European Commission said on Friday that U.S. Big Tech giant Meta and Chinese-owned social media app TikTok breach their obligation to grant researchers adequate access to public data under the Digital Services Act (DSA) according to its preliminary findings. In a statement, the Commission also said that Meta's Facebook and Instagram do not appear to provide a user-friendly and easily accessible mechanism for users to flag illegal content, such as child sexual abuse material and terrorist content. The EU has cracked down on Big Tech companies with the Digital Services Act, which requires large platforms such as social media sites and search engines to have robust measures to mitigate the spread of illegal and harmful content. "The Commission's preliminary findings show that Facebook, Instagram and TikTok may have put in place burdensome procedures and tools for researchers to request access to public data", the EU executive said about the transparency issue. "Allowing researchers access to platforms' data is an essential transparency obligation under the DSA, as it provides public scrutiny into the potential impact of platforms on our physical and mental health." On Meta hindering the reporting of illegal content, the Commission said the company currently has mechanisms that impose several unnecessary steps and additional demands on users and use "deceptive interface designs". "Such practices can be confusing and dissuading. Meta's mechanisms to flag and remove illegal content may therefore be ineffective. Under the DSA, 'Notice and Action' mechanisms are key to allowing EU users … to inform online platforms that certain content does not comply with EU or national laws", it said. A Meta spokesperson told Reuters the company disagreed with any suggestion it breached the DSA, and that it would continue to negotiate with the European Commission on this. "In the European Union, we have introduced changes to our content reporting options, appeals process, and data access tools since the DSA came into force and are confident that these solutions match what is required under the law in the EU," the spokesperson said. TikTok was not immediately available to respond to a Reuters request for comment. The Commission added the companies now have the possibility to examine its findings and take measures to remedy the breaches, adding that the preliminary findings do not prejudge the outcome of the investigation. But if the findings of the Commission are confirmed by relevant consultations, it may impose a fine on the companies of as much as 6% of their annual global sales. (Reporting by Benoit Van Overstraeten and Supantha Mukherjee ; Editing by Alessandro Parodi)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Jessica DiNapoli NEW YORK (Reuters) -Snack maker Mondelez is using a new generative AI tool to cut costs for the production of marketing content by 30% to 50%, a senior executive told Reuters. The packaged food manufacturer began developing the tool last year with IT firm Accenture and expects that it will be capable of making short TV ads that would be ready to air as soon as next year's holiday season, and potentially for the 2027 Super Bowl, said Jon Halvorson, Mondelez’s global senior vice president of consumer experience. The Cadbury chocolate producer has invested more than $40 million in the tool, Halvorson said, adding that savings would grow if the tool is able to make more elaborate videos. Faced with tariffs and shrinking shopper budgets, Mondelez, like other consumer goods companies, is looking to adopt AI to slash fees paid to advertising agencies, and speed up how long it takes to develop and sell new products.  Rivals such as macaroni-and-cheese maker Kraft Heinz and Coca-Cola have also been trying out AI for ads. Coke in 2024 ran AI-created holiday ads, though the computer-created people in them were ridiculed by some consumers for lacking real emotion. Mondelez is not yet putting human likenesses in its AI-created content. It is using content generated by the new tool on social media for its Chips Ahoy cookies in the U.S. and Milka chocolate in Germany. An eight-second Milka video shows waves of chocolate rippling over a wafer, along with different backgrounds depending on which consumer Mondelez is targeting. The cost to do animations "is in the hundreds of thousands," Halvorson said. "This type of set-up is orders of magnitude smaller." In the U.S., Oreo will use the tool for product pages on Amazon and Walmart in November. Mondelez plans to use the tool in the coming months for Lacta chocolate and Oreo in Brazil, and Cadbury in the UK, Halvorson said.  Tina Vaswani, vice president of digital enablement and data for the company, said humans will always check what the tool produces to avoid any mishaps. Mondelez has rules prohibiting highlighting unhealthy eating habits, vaping, over-consumption, emotionally manipulative language and the use of offensive stereotypes, according to a document shared by the Chicago-based company. (Reporting by Jessica DiNapoli in New York; Editing by Aurora Ellis)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Naomi Rovnick LONDON (Reuters) -Major investors, spooked by AI exuberance yet wary of betting against it, are shifting from hyped-up stocks into potential next-in-line winners, reviving a strategy from the 1990s dotcom era that helped some sidestep the crash. As U.S. stocks have hit successive records and AI chipmaker Nvidia's valuation has surged beyond $4 trillion, professional investors have been trying to find ways to make money from the bull market while avoiding excessive risk. Some are looking back to the 1990s internet boom, which spread from startups to telecoms and tech, and where hedge funds rode the wave by flipping out of highly-valued stocks before they peaked and picking others that had room to rise. "What we are doing is what worked from 1998 to 2000," said Francesco Sandrini, multi-asset head and Italy CIO at Europe’s largest asset manager Amundi. He highlighted signs of irrational exuberance on Wall Street, such as frenzied trading in risky options pegged to the share prices of big AI stocks. But he said he expected the new tech enthusiasm to continue and hoped to bank gains via bets on reasonably valued assets that might rally next. Sandrini said this involved trying to find "the highest growth opportunities that so far the market had failed to spot", with moves into software groups, robotics and Asian tech. Other investors also expected to edge out of Wall Street's Magnificent Seven stocks after shares in Nvidia more than tripled in two years, but want to keep their diversification within the AI sphere. ASSET MANAGERS NEED TO BE NIMBLE TO RIDE THE WAVE "The odds of this (AI boom) being a bust are very high because you've got companies spending trillions and all fighting for the same market that does not yet exist," said Goshawk Asset Management CIO Simon Edelsten, who worked on telecom IPOs at stockbroker Dresdner Kleinwort Benson in London in 1999. He expected the next phase of AI fever to spread from Nvidia and others like Microsoft and Alphabet into related sectors. Timing the phases of a bubble has historically been a way to play it without the risk of trying to call the peak too early. A study by economists Markus Brunnermeir and Stefan Nagel showed that hedge funds mostly did not bet against the dotcom bubble, but rode it skillfully enough to beat the market by about 4.5% per quarter from 1998-2000 and avoid the worst of the downturn. They shed high-priced internet stocks in time to recycle profits into others before they caught the attention of less sophisticated investors. "There were good profits to be made for the fleet of foot even during 2000 when the top came," Edelsten at Goshawk said, adding the current market environment was similar to 1999. He favoured IT consultants and Japanese robotics groups that can potentially pick up revenues from AI heavyweights, in what he said was the typical chronology of a market gold rush. "When someone strikes gold, (you) buy the local hardware store where the prospectors will buy all their shovels." INVESTORS TRY TO STAY IN AI WITHOUT EXCESSIVE RISK Investors are also attempting to benefit from the trillions of dollars so-called hyperscalers such as Amazon, Microsoft and Alphabet are committing to AI data centres and advanced chips without taking on more direct exposure to these companies. Fidelity International multi-asset manager Becky Qin said uranium was her favoured new AI trade because power-hungry AI data centres could gobble up nuclear energy. Kevin Thozet, investment committee member at asset manager Carmignac, was taking profits on Magnificent Seven stocks and building up a position in Taiwan's Gudeng Precision, which makes delivery boxes for AI chipmakers including TSMC. Asset managers are also concerned that the rush to build data centres could result in overcapacity, as in the fiber-optic cable boom in the telecoms industry. "In any new technological paradigm we don't get from A to B without excesses along the way," said Pictet Asset Management senior multi-asset strategist Arun Sai. Even though top AI stocks like Microsoft, Amazon, and Alphabet are being powered by strong earnings, he still sees "the building blocks of a bubble" and favours Chinese stocks as a hedge if rapid AI advancements in China sap Wall Street's AI enthusiasm. Some investors, though, do not favour this relative value approach to AI investing as a way to mitigate future losses. Oliver Blackbourn, portfolio manager at Janus Henderson, said he was hedging his U.S. tech positions with European and healthcare assets lest an AI stock crash takes the U.S. economy down with it. He said it was impossible to forecast how long the AI craze would roll on because calling the peak was usually only possible with hindsight. "We're in 1999 until the bubble pops." (Reporting by Naomi Rovnick; editing by Dhara Ranasinghe and Jane Merriman)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

More and more young people are struggling with mental health problems, including eating disorders. A medic explains what causes them, whether they are curable and how family and friends can help. Berlin (dpa) – Food. We can't live without it, and eating is one of life's greatest pleasures. Yet for many people, preoccupied with their weight and body shape, it's a source of emotional stress that can seriously damage their physical and mental health. Long-term observational studies in various countries show a clear increase in eating disorders, in particular anorexia nervosa, since the Covid-19 pandemic, says Germany's Federal Institute of Public Health (BIÖG). They typically emerge in adolescence and young adulthood, mostly among girls, says the World Health Organization. What's behind the increase in this mental health condition and, more importantly, what can relatives of people with an eating disorder do to help? Here are answers from Dr Stephan Zipfel, medical director of the Department of Psychosomatic Medicine and Psychotherapy at Tübingen University Hospital in Germany: Why do so many young people have an eating disorder? For about the past 20 years, mental illness in general has been rising worldwide among young people, especially in industrialized countries. Possible reasons include megatrends, societal change, social media, war and climate change, write researchers from The Lancet Psychiatry Commission on youth mental health. "These developments lead to rising distress, alienation and loneliness," Zipfel says. "The overall situation, particularly for adolescents, was growing more stressful even earlier, and the coronavirus pandemic, along with accompanying preventive measures such as lockdowns, have caused a marked further rise." Being spatially and socially isolated within one's immediate family during the pandemic led to more stress and conflicts, especially for people who were already stressed, he says. Above all, he goes on, "there were no – or very limited – opportunities to associate with other adolescents, and a lack of in-person contacts with persons of trust outside the home as well as socio-pedagogical workers and teachers." What's more, counselling and treatment services that could have addressed mental health issues early were unavailable. "This created an extremely toxic situation in which the affected persons developed a pronounced eating disorder," Zipfel says. Why do people develop eating disorders? Roughly speaking, it's about control over your sense of control, as Zipfel puts it. "There's a group of people who, if they feel their life is slipping out of their hands, try to get a grip on it again by controlling their eating behaviour, their weight." Also playing a role, he says, is that appetite, in conjunction with body weight regulation, is, via evolution, a very deeply rooted basic need. "Only a small group of people are able to circumvent this basic need. For them, experiencing this control feels good at first." What then happens is that "the person's self-esteem, self-image, increasingly depends on controlling their appetite." Zipfel describes this dependence as "I-syntony, that is, it's connected to – actually almost baked into – the person." At this stage, there's a risk of the disorder becoming part of their personality. How do you overcome an eating disorder? Even severe cases of eating disorders are treatable, Zipfel says. Referencing a follow-up study of a clinical trial on psychotherapy for anorexia (trying to keep one's weight as low as possible) in which he participated, he says 41% of the patients – all female – were classified as recovered five years after therapy ended. "This means they displayed no symptom shift either. Neither were there abnormalities as regards eating behaviour, nor other mental illnesses." The recovery process usually isn't a smooth path, however. Zipfel therefore recommends patients and medical practitioners jointly devise an overall treatment plan early, by discussing the steps necessary to overcome the eating disorder. Techniques that strengthen patients' therapy motivation are especially helpful. But these initial steps, too, can be difficult and painful. "A big reason that eating disorder sufferers are ambivalent about undergoing therapy is knowing it's aimed at making them part with their ability – which stabilizes them – to control these very basic needs to a certain extent." Consequently, they relatively rarely have the desire to address their anorexia. "People with an eating disorder don't necessarily seek help, but rather tend to isolate themselves," says Zipfel. "Dealing with this is a challenge for family and friends." What can family members and friends do for someone with an eating disorder? "It's important they know that they aren't solely responsible," Zipfel says. "If you're a parent, you're naturally in a different position to a teacher or friend. So you've always got to see who can assume what responsibility." Sometimes, he says, friends feel that they're responsible, since the affected person doesn't speak with their parents. But "this would be an excessive burden," he says. What you could do, though, is to say, "Let's go to Ms X or Y, our school social worker, and talk with her." He also recommends counselling centres: "They offer relatively low-threshold services and provide a professional framework within which affected persons can receive initial support." How should you best broach the subject of professional help for the person affected? It's very important to not be reproachful or scolding. Even well-intentioned advice is often rebuffed, Zipfel says. A solicitous approach by someone who has some sort of relationship with the person affected is more likely to be effective. "If you have a relationship – of whatever kind – with them, it's important to first say something like, 'I've noticed you don't seem to be doing well. This pains me, and I'd gladly help you.'" It's not unusual for the person to then brush you off because their eating disorder is often a desperate attempt to solve a problem they think they have, and not something akin to a next-top-model aspiration run amok, Zipfel says. With this in mind, "you can better understand the person's plight and suffering, and then better deal with a nasty reaction." The following information is not intended for publication dpa/tmn lue xxde sw ob arw

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

By Jessica DiNapoli NEW YORK (Reuters) -Snack maker Mondelez is using a new generative AI tool to cut costs for the production of marketing content by 30% to 50%, a senior executive told Reuters. The packaged food manufacturer began developing the tool last year with IT firm Accenture and expects that it will be capable of making short TV ads that would be ready to air as soon as next year's holiday season, and potentially for the 2027 Super Bowl, said Jon Halvorson, Mondelez’s global senior vice president of consumer experience. The Cadbury chocolate producer has invested more than $40 million in the tool, Halvorson said, adding that savings would grow if the tool is able to make more elaborate videos. Faced with tariffs and shrinking shopper budgets, Mondelez, like other consumer goods companies, is looking to adopt AI to slash fees paid to advertising agencies, and speed up how long it takes to develop and sell new products.  Rivals such as macaroni-and-cheese maker Kraft Heinz and Coca-Cola have also been trying out AI for ads. Coke in 2024 ran AI-created holiday ads, though the computer-created people in them were ridiculed by some consumers for lacking real emotion. Mondelez is not yet putting human likenesses in its AI-created content. It is using content generated by the new tool on social media for its Chips Ahoy cookies in the U.S. and Milka chocolate in Germany. An eight-second Milka video shows waves of chocolate rippling over a wafer, along with different backgrounds depending on which consumer Mondelez is targeting. The cost to do animations "is in the hundreds of thousands," Halvorson said. "This type of set-up is orders of magnitude smaller." In the U.S., Oreo will use the tool for product pages on Amazon and Walmart in November. Mondelez plans to use the tool in the coming months for Lacta chocolate and Oreo in Brazil, and Cadbury in the UK, Halvorson said.  Tina Vaswani, vice president of digital enablement and data for the company, said humans will always check what the tool produces to avoid any mishaps. Mondelez has rules prohibiting highlighting unhealthy eating habits, vaping, over-consumption, emotionally manipulative language and the use of offensive stereotypes, according to a document shared by the Chicago-based company. (Reporting by Jessica DiNapoli in New York; Editing by Aurora Ellis)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)

BRUSSELS (Reuters) -The European Commission said on Friday that U.S. Big Tech giant Meta and Chinese-owned social media app TikTok breach their obligation to grant researchers adequate access to public data under the Digital Services Act (DSA) according to its preliminary findings. In a statement, the Commission also said that Meta's Facebook and Instagram do not appear to provide a user-friendly and easily accessible mechanism for users to flag illegal content, such as child sexual abuse material and terrorist content. The EU has cracked down on Big Tech companies with the Digital Services Act, which requires large platforms such as social media sites and search engines to have robust measures to mitigate the spread of illegal and harmful content. "The Commission's preliminary findings show that Facebook, Instagram and TikTok may have put in place burdensome procedures and tools for researchers to request access to public data", the EU executive said about the transparency issue. "Allowing researchers access to platforms' data is an essential transparency obligation under the DSA, as it provides public scrutiny into the potential impact of platforms on our physical and mental health." On Meta hindering the reporting of illegal content, the Commission said the company currently has mechanisms that impose several unnecessary steps and additional demands on users and use "deceptive interface designs". "Such practices can be confusing and dissuading. Meta's mechanisms to flag and remove illegal content may therefore be ineffective. Under the DSA, 'Notice and Action' mechanisms are key to allowing EU users … to inform online platforms that certain content does not comply with EU or national laws", it said. The Commission added the companies now have the possibility to examine its findings and take measures to remedy the breaches, adding that the preliminary findings do not prejudge the outcome of the investigation. But if the findings of the Commission are confirmed by relevant consultations, it may impose a fine on the companies of as much as 6% of their annual global sales. (Reporting by Benoit Van Overstraeten; Editing by Alessandro Parodi)

(The article has been published through a syndicated feed. Except for the headline, the content has been published verbatim. Liability lies with original publisher.)