United News of Bangladesh

Meta is defending its efforts to protect children online as it faces a growing number of lawsuits accusing the social media company of harming young users. In the latest case, Meta is fighting allegations in a California federal court that it helped fuel the youth mental health crisis by deliberately designing Instagram and Facebook features to keep children engaged for long periods. Four US states are seeking billions of dollars in damages and major changes to the way Meta operates the platforms. "Meta takes seriously" issues involving children's mental health and what they see and do on social media, the company's lawyer Paul Schmidt said in opening arguments this week. Critics, however, say Meta's safety measures have not gone far enough and that many of them do not work properly. Marc Berkman, executive director of the Organization for Social Media Safety, said Meta introduced many of its safeguards mainly in response to public pressure rather than genuine concern for children's safety. Measures targeting harmful content Many of Meta's safety changes have focused on bullying, harassment and offensive content. Instagram has long allowed users to block others. Over the years, both Instagram and Facebook have added tools to report problematic accounts and inappropriate content. In 2016, Instagram introduced options to delete or filter offensive comments. It later added warnings encouraging users to think twice before posting potentially harmful content, as well as controls for limiting comments and tags. Meta says adults who are not already connected with teenagers are prevented from starting private conversations with them. The company has also increased efforts to prevent children under 13 — the minimum age for Facebook and Instagram accounts — from using the platforms. Schmidt acknowledged that some users still provide false birth dates when signing up but said Meta takes its age restrictions seriously. The company also displays warnings and support information when users search Instagram for topics such as suicide, self-harm and eating disorders. Meta points to its "Teen Accounts", introduced in 2024, as another major safety measure. The accounts limit the type of photos and videos teenagers can view, broadly following restrictions similar to those of a PG-13 movie unless parents give permission for wider access. Critics say such measures are more about public relations than effective protection. A September 2025 report by former Meta employee and whistleblower Arturo Béjar and four online safety groups accused the company of failing to take meaningful action on child safety. Haley Hinkle, policy counsel at Fairplay, which helped prepare the report, said researchers reviewed 47 of Meta's 53 teen safety features on Instagram and found that 60% were either unavailable or did not work as advertised. Meta rejected the report's conclusions as "misleading" and "dangerously speculative." Concerns over endless scrolling Researchers also question whether Meta has done enough to prevent children from becoming overly dependent on social media. Ashley Shea, a Cornell University Ph.D. candidate who studies social media and online spaces, said features such as endless scrolling, pull-to-refresh designs, frequent notifications and algorithm-based recommendations can keep users engaged. She said these features can take advantage of vulnerabilities in young people's developing brains, particularly their desire for rewards. Shea acknowledged that Meta has taken important steps to deal with harmful content, cyberbullying and sexual predators. But she said she has not seen similar efforts to reduce the addictive effects of features designed to capture users' attention. Time limits and parental controls Meta also points to parental controls and other time-management tools as evidence that it is addressing concerns over excessive social media use. Schmidt said some teenagers struggle to control how much time they spend on social media and highlighted tools such as parental supervision and in-app reminders that encourage teenagers to take a break after repeatedly viewing similar content or using the platforms late at night. Hinkle said these tools are voluntary and difficult to set up, making them inadequate. Béjar also criticized Meta's "Take a Break" feature during testimony Wednesday, saying it is "designed to fail" because users must activate it themselves and few choose to do so. Berkman said technology companies are built to increase revenue, which for social media platforms often means encouraging users, including children, to spend as much time as possible on their services. Shea said the lawsuits are significant because they focus on whether social media platforms are harmful because of the way they are designed, rather than only because of the content users encounter. She added that growing pressure from lawmakers and lawsuits could force broader changes across the technology industry. "The momentum only continues to grow," Shea said, adding that companies driven by profit are likely to respond more quickly as they face greater financial pressure from legal action.  

Russia disabled thousands of satellite modems in Ukraine and other European countries on the day it invaded Ukraine in 2022, disrupting communications. Now, an artificial intelligence-assisted tool is being used to help protect satellite networks from similar attacks amid growing concerns over Russian aggression and cyber threats from Iran. Cybersecurity company Atalanta has launched Argo, a product that includes technology being used to test the resilience of Viasat’s satellite communications network, which was targeted in the 2022 cyberattack. Cybersecurity experts say the technology offers a new way to protect critical infrastructure from increasingly sophisticated cyberattacks. Atalanta says Argo uses what it calls “software understanding,” combining advanced mathematics and AI to examine software and internet-connected systems and identify vulnerabilities more thoroughly. Recent conflicts in Ukraine and Iran have highlighted the growing risks posed by cyberattacks on essential services. Several US government agencies have warned this summer that Iranian hackers have targeted water and wastewater systems and other critical infrastructure. “The tools and techniques of yesterday do not scale for today,” Atalanta CEO and co-founder Anjana Rajan told The Associated Press, saying software understanding would be increasingly important for systems critical to the United States. Nick Saunders, Viasat’s chief of cyber strategy, said the technology is helping the company strengthen its network against future attacks while also providing mathematical evidence to support its security claims. “We have built over the past year, working with Atalanta ... a capability where we can start to develop an understanding of how our systems are more resilient against attacks,” Saunders said. Atalanta is also working on the Genesis Mission, a Department of Energy initiative aimed at developing autonomous nuclear reactors. Greg Shannon, chief cybersecurity scientist at the government’s Idaho National Laboratory, said such technology could eventually become a standard part of testing new systems. “In a decade or two, I would expect it to be just part of the standard development tool kits,” Shannon said. He noted that the Defense Advanced Research Projects Agency, or DARPA, has invested billions of dollars in related technology over the past two decades. That research helped create demand for systems such as Argo. DARPA, the US military’s advanced research agency, has played a key role in developing technologies including the internet and stealth aircraft. “They've probably invested well over $2 billion in various programs to advance formal methods ... and we're finally seeing them start to come to fruition,” Shannon said. Last year, Pentagon Chief Technology Officer Emil Michael said in a prerecorded address to a DARPA software conference that he viewed the mathematics behind Atalanta’s technology as an important part of the military’s cybersecurity future. He said he wanted it to become the Department of Defense’s “gold standard” for cybersecurity.

China has successfully recovered a rocket’s first stage on land for the first time, state media reported Wednesday, marking another step forward in the country’s efforts to develop reusable space technology. The achievement marks China’s second successful recovery of a rocket stage. In July, the country recovered a rocket stage on a sea platform. China’s state-run Xinhua News Agency reported that the reusable Zhuque-3 rocket was launched Wednesday morning and its first stage was successfully recovered afterward. SpaceX, founded by Elon Musk, and Jeff Bezos’ Blue Origin have been recovering rockets since 2015. Reusing rocket hardware that would otherwise be discarded has helped lower launch costs for satellites and other payloads. Xinhua said Wednesday’s mission was also the first time China had used deployable landing legs during a rocket recovery. It described the successful landing as “a major breakthrough” in the country’s reusable rocket technology. On July 10, the first stage of a Long March-10B rocket separated from its second stage after liftoff and returned to a platform at sea. The stage was recovered using a net-capture system. The following day, July 11, Japan conducted the first test flight of an experimental reusable rocket, which successfully launched and landed. Japan is also working to develop reusable rocket technology as it seeks to reduce launch costs and strengthen its competitiveness in the global space market.

China is drawing on its nationwide capabilities to develop a locally tailored yet nationally coordinated artificial intelligence industry, aiming to accelerate the commercial application of AI and create new economic momentum. The latest economic data from 31 provincial-level regions shows rapidly growing demand for AI terminal applications, alongside a tiered development pattern based on the strengths and resources of individual regions. “The driving role of the AI wave is becoming increasingly pronounced in emerging industries such as integrated circuits and industrial robots across regions,” said Xiao Hongwei, a researcher at the State Information Center under the National Development and Reform Commission, China’s top economic planning agency. “AI is accelerating its transformation from a technical concept into tangible productivity gains in China,” Xiao said. First-tier ai hubs China’s intelligent computing capacity reached 2,185 EFLOPS by the end of June, up 177 percent year on year, reflecting the rapid expansion of the country’s AI infrastructure. The nationwide utilization rate of computing facilities stood at 71.4 percent, according to the Ministry of Industry and Information Technology. EFLOPS, or exaFLOPS, is a measure of computing performance, with one unit representing 1 quintillion floating-point operations per second. Xiao said regions across China are developing their AI sectors according to local conditions and leveraging their respective strengths, resulting in a pattern of complementary advantages and coordinated development. “Benefiting from their strengths in technology, talent and industrial chains, Beijing, Shanghai and Guangdong Province are taking the lead in developing indigenous AI industrial ecosystems while continuing to advance the full industrial chain in areas such as integrated circuit manufacturing and industrial robots,” he said. All three regions have identified “global AI competitiveness” as a goal in their five-year plans for 2026-2030, seeking leadership in AI technology, industrial ecosystems and application scenarios. Beijing is working to establish itself as a global AI innovation hub. As of July 23, the capital had completed filings for 259 large models and registrations for 195 AI products, including general-purpose large models, industry-specific models and on-device mobile models, according to the municipal cyberspace administration. By the end of June, 169 large AI models in Shanghai had received filing approval. The city’s 394 above-scale AI enterprises generated more than 637 billion yuan ($93.85 billion) in industrial output in 2025. In Guangdong, the core AI industry exceeded 300 billion yuan in scale last year. Diversified tracks with unique advantages Other regions are also using their distinctive resources to develop specialized AI sectors and pursue differentiated growth paths. East China’s Anhui Province is stepping up efforts in the AI chip industry and has achieved a series of breakthroughs. Nexchip, the province’s first 12-inch wafer foundry, debuted on the Hong Kong Stock Exchange in July, giving it dual listings in Hong Kong and on the A-share market. Its fourth-phase project, involving an investment of 35.5 billion yuan, is scheduled to begin production in the fourth quarter of this year. ChangXin Memory Technologies, a memory chipmaker based in Hefei, the capital of Anhui, also went public in late July and became the most valuable company on China’s A-share market. Central China’s Hubei Province is leveraging its strong optoelectronics sector. Its capital, Wuhan, is home to the Optics Valley of China, and the province has built a globally competitive full industrial chain covering optical communications, chips, displays, terminals and networks to meet growing demand for AI computing power. Henan Province, traditionally a major transportation hub, is using AI to transform itself from a “railway heart” into a “computing power hub.” In April, the country’s largest scientific intelligent computing cluster began operations at the national supercomputing network’s core node in Zhengzhou, the capital of Henan. The project marked a major advance in China’s computing infrastructure for AI-driven scientific research and industrial applications. Meanwhile, regions including Xinjiang, Inner Mongolia and Guizhou are leveraging their abundant clean energy resources to develop intelligent computing centers. Computing center in the mountains Located in the Wuling Mountains in Lichuan City, Hubei Province, the Lichuan AI Computing Center demonstrates how geographical and natural advantages, including a cool climate and clean air, can support AI development. Although Lichuan lacks abundant scientific and educational resources, its relatively cool weather and rich hydropower resources make it a suitable location for a computing center with comparatively low electricity costs. The facility houses rows of black server cabinets, each about two meters high, with rapidly flashing indicator lights. It provides around 150 PFLOPS of computing power, with utilization exceeding 80 percent. PFLOPS, or petaflops, measures computing speed for AI and data-intensive applications, with one PFLOP equivalent to 1 quadrillion calculations per second. The center’s services now support sectors including culture and tourism, government affairs and aerospace information. In Liangwu Township of Lichuan, GEOVIS Raywin Technology Co. Ltd. has established a satellite ground receiving station that can directly access computing resources from the nearby AI center to process satellite data in real time. “The site is surrounded by mountains yet offers expansive views, with a pristine electromagnetic environment and a mild, stable climate. These natural endowments make it an ideal location,” said Gao Kui, the company’s general manager of operations. He said the station’s proximity to the computing center also improves operational efficiency and convenience while reducing costs. Beyond meeting external demand for computing power, the center’s surplus capacity is supporting local industries. More than 20 million tourists visit Lichuan each year for summer retreats and wellness, generating huge amounts of data related to accommodation, food, transportation, sightseeing, shopping and entertainment. These data have supported the development of a domestic large language model for culture and tourism services, improving visitors’ experiences through AI-powered smart guides and other services. The Ministry of Industry and Information Technology, working with other government departments, has been promoting the development of high-quality computing infrastructure. It is accelerating the coordinated deployment of hub, regional and edge computing facilities while building a national integrated automated monitoring system to track computing power and its use.

Berkshire Hathaway significantly increased its investment in Google parent Alphabet and expanded its holdings in US homebuilders during the second quarter, according to the conglomerate’s latest regulatory filing. The Omaha-based company also reduced its positions in several financial firms and other companies during the April-June period, the filing showed. Berkshire bought about 48.1 million additional Alphabet shares in the second quarter, bringing its total holding in the technology giant to roughly 106 million shares. The stake was valued at about $37.76 billion as of June 30. At the end of last year, Berkshire held only around 17.8 million Alphabet shares worth about $5.6 billion. Berkshire CEO Greg Abel, who succeeded Warren Buffett at the start of the year, agreed in June to make a $10 billion investment in Alphabet, adding to the position Berkshire began building last year. Alphabet has announced plans to spend $80 billion on computing infrastructure to support its artificial intelligence services. Beyond the technology sector, Berkshire continued to increase its exposure to the US housing market. Its investment in homebuilder Lennar rose by nearly 30% during the quarter. Berkshire also took a new, small position in D.R. Horton, which was worth about $580,504 at the end of June. In July, Berkshire completed its $6.8 billion acquisition of homebuilder Taylor Morrison. The conglomerate also substantially increased its holdings in Delta Air Lines and Macy’s during the quarter. Its Delta stake was valued at about $5.37 billion at the end of June, while its Macy’s investment was worth around $173 million. At the same time, Berkshire reduced its positions in several companies, including supermarket operator Kroger, steelmaker Nucor and dialysis services provider DaVita. It completely exited beverage company Constellation Brands, selling all 632,890 shares it held. Berkshire also cut several financial-sector holdings. Its stakes in Bank of America and Ally Financial fell by about 6% and 6.9%, respectively, while its holding in Capital One Financial was reduced by 58%. Investors closely monitor Berkshire’s portfolio because many have historically tried to replicate Buffett’s investment decisions. Buffett remains Berkshire’s chairman and largest shareholder despite handing over the chief executive role to Abel at the beginning of the year. Berkshire, which owns a wide range of businesses including insurer Geico and railroad operator BNSF, does not typically comment on its quarterly stock transactions, allowing the regulatory filings to provide the main public insight into its investment decisions.

Facing acute labour shortages, an ageing population and long-standing productivity problems, Japan would appear to be an ideal environment for the adoption of artificial intelligence (AI). Yet AI use in Japanese workplaces remains relatively slow and cautious compared with the United States and the United Kingdom. Japan’s response to AI is beginning to resemble one of the country’s traditional, slow-moving Noh plays. Everyone on stage agrees that urgent action is needed, and the message is repeated with solemnity, but the actors remain largely frozen in place. While that may build dramatic tension, it does little to address the country’s immediate economic challenges. Data on workplace AI use suggests Japan is lagging behind. Only 8.4% of Japanese workers use AI in their jobs, according to a report published late last year by the Organisation for Economic Co-operation and Development (OECD). The figure compares with 50% in the US and 32% in the UK. In Singapore, where AI adoption is reportedly among the highest in the world and the highest in Asia, 56% of workers are said to use AI several times a week. So why is Japan moving so cautiously? Austin Xu, co-founder of US start-up Kuse AI, attributes the slow adoption to the conservative and risk-averse nature of many Japanese companies. His company recently opened an office in Japan to market its AI systems to local businesses. “Some organisations genuinely take a long time to make decisions because of their process and consensus-driven culture,” he says. “Tolerance for AI mistakes is close to zero, particularly in client-facing roles. Some would rather leave a position vacant than allow a machine to handle it.” Xu says the approach is very different in the US, where some companies are already giving AI agents greater freedom to improve productivity. In many American workplaces, he says, AI systems are introduced as assistants and gradually become part of normal workflows, with managers willing to test them and correct mistakes along the way. Japanese companies, by contrast, often require more evidence before trusting AI, he adds. Parrisa Haghirian, professor of international management at Kyoto University of Advanced Science, also believes risk aversion is a major factor. “The challenges of adopting AI are similar to those involved in introducing any kind of change in Japanese firms,” she says. “This is why AI use remains limited and cautious, particularly in workplaces.” Haghirian contrasts Japan’s approach with the more entrepreneurial culture in the US, saying Japanese companies are particularly sensitive to errors, uncertainty and reputational damage. Because generative AI is still not completely reliable, she says, Japanese companies mainly use it for relatively low-risk activities such as writing, summarising documents and gathering information. Its use remains much more limited in core operations, decision-making and broader business-process improvements. Japan’s healthcare sector is considered especially slow to embrace AI. Some hospitals have yet to fully digitise patient records. “Paper documents accumulate at a staggering scale,” an employee at a Japanese hospital, who requested anonymity, says. “It’s like the Stone Age.” The Japanese government is aware of the broader problem and is seeking to encourage businesses to adopt AI. Last year, parliament passed the country’s first AI Promotion Act, which uses a light-touch regulatory approach aimed at encouraging greater investment in AI. Tokyo has pledged to make Japan “the world’s most-friendly country for developing and utilizing AI.” The government sees AI as a potential tool for improving productivity, an area where Japan continues to perform poorly compared with other G7 economies. Japan’s Ministry of Finance says business adoption of AI has increased substantially, with 75% of companies now using the technology, compared with just 11% five years ago. Critics, however, say the headline figure masks the limited use of AI within many companies. Only a small proportion of employees may actually use the technology, and even those users often rely on it for only a narrow range of tasks. Prof Yasushi Ogasawara, an expert on Japan’s social system and technology at Meiji University, says the country also lacks enough workers with strong technology skills. “Although Japanese people like using gadgets such as smartphones, digital literacy is low here,” he says. The problem is underscored by a report earlier this year estimating that Japan could face a shortage of nearly 800,000 IT professionals by 2030. Many Japanese companies also continue to rely on outdated computer systems, with around 60% reportedly more than two decades old. Ogasawara says companies also face pressure to avoid contributing to unemployment, making them more cautious about developing AI that could replace human workers. “The government is talking about AI, reskilling and digital skills, but in reality it is difficult to adapt human resources to digital skills because the top priority is maintaining full employment,” he says. There are signs of change, however. Some companies are beginning to prioritise “AI literacy” when recruiting new graduates. Larger private-sector firms such as major trading company Kanematsu are hiring graduates with strong AI skills. Uta Yamaguchi, who joined the company in April, says she frequently uses AI at work, as do many of her colleagues. She uses it to draft emails for her boss, summarise complex documents and record and summarise meetings. However, she says AI systems capable of independently handling multi-step tasks remain virtually unheard of in Japan compared with the US. Her employer, known for being forward-looking, has now begun introducing such systems. Ogasawara says the pace of change in Japan remains limited. “Japanese society expects painless reform, so I think it can be said that drastic reform is difficult, to say the least,” he says. Disruption may remain an uncomfortable concept in Japan, but without greater willingness to embrace change, the productivity gains the country urgently needs from AI could remain out of reach. #Source: BBC

Brazil’s data protection authority has ordered messaging platform Discord to suspend its livestreaming feature after a 13-year-old girl allegedly was encouraged to take her own life during a broadcast on the platform. The National Data Protection Authority (ANPD) launched an investigation on Aug. 7 following the death of the teenager, who was reportedly pressured by other users in a Discord server, or group, to end her life. Discord allows users to create and join interest-based communities and communicate through real-time voice and video calls using its “Go Live” livestreaming feature. Discord said it was “thoughtfully reviewing” the ANPD’s decision and reaffirmed its commitment to user safety. “Groups or individuals who promote or encourage violence have no place on Discord,” a company spokesperson said. The spokesperson said Discord quickly investigated the private, invite-only server involved and shut it down shortly after it was created. The company also said it continues to cooperate with law enforcement investigating the case. According to Reuters, the ANPD found that Discord does not have real-time access to livestream content, limiting its ability to automatically detect potential violations. The platform, available on computers and mobile devices, has more than 200 million monthly active users and is particularly popular among gamers. Discord said its investigation found evidence that the criminal activity linked to the girl’s death had been coordinated on other platforms before the Discord server was created and continued after those involved were banned from Discord. Five teenagers have reportedly been arrested in connection with the girl’s death. Discord has three days to comply with the suspension and 10 business days to appeal the decision. The platform could face fines of up to 50 million reais ($9.67 million) for each violation. The ANPD said the suspension will remain in effect until Discord demonstrates that it has introduced “adequate protective measures for minors,” which could include age-verification systems. #Source: BBC

Vietnamese aerospace company VinSpace has signed an agreement with SpaceX to launch its first satellites into orbit in 2027, as Vietnam steps up efforts to develop its emerging space industry. VinSpace said Tuesday that its satellites will be sent into space on a SpaceX Transporter rideshare mission. The service allows satellites from multiple customers to share a single rocket, reducing launch costs. The company said it will develop and operate the satellites, which will be used to test its technology in orbit and support future commercial applications. VinSpace is part of Vietnam's Vingroup conglomerate and was established in November 2025. It announced in April that it planned to develop and launch its first satellites in 2027. Founded by billionaire Pham Nhat Vuong, Vingroup is Vietnam's largest privately owned company, with businesses ranging from real estate and retail to healthcare, education and tourism. The conglomerate is expanding into high-tech industries, including electric vehicles, space, artificial intelligence and robotics, as Vietnam seeks to strengthen its economy and become one of Asia's emerging economic powers. Vietnam has been working to develop its space capabilities for decades but remains a relatively small player in the global space sector. The country launched its first telecommunications satellite in 2008 and a second one in 2012, according to official documents. In March, Vietnam opened a major space science and technology centre at the Hoa Lac High-Tech Park in Hanoi. The facility is intended to boost satellite development and the use of space-based data. Vietnam aims to become a mid-level space power in Southeast Asia by 2030. In February, Vietnam approved SpaceX's Starlink satellite internet service in the country. Some analysts viewed the move as part of Vietnam's efforts to avoid US tariffs. Starlink has since started accepting orders in Vietnam. VinSpace said the SpaceX agreement is part of its plan to become a full-service aerospace company, with capabilities covering satellite design and manufacturing, launch management, satellite operations and space-based data services. “Reliable access to space is fundamental to turning satellite innovation into operational missions,” VinSpace CEO Thu Vu said in a statement. VinSpace did not disclose the value of the contract or the number and specifications of the satellites.

President Donald Trump's media company reported a massive loss in the second quarter and announced plans to abandon several new business ventures and refocus on its core mission as a social media platform. Trump Media & Technology, which operates the Truth Social platform, said Monday it lost $238 million in the three months through June after expanding into businesses outside traditional media, including cryptocurrency. The loss was more than 10 times the amount recorded a year earlier. The company's loss per share widened to 86 cents from 8 cents. During a conference call following the earnings release, new CEO Kevin McGurn said the company would largely abandon a yearlong effort to expand into industries such as online betting and cryptocurrency and instead concentrate on its social media business. "We made the disciplined choice to pivot in order to invest more time and resources in our most important initiatives," McGurn said. "We will say no to things or change course as warranted." Trump Media shares fell slightly in after-hours trading following the earnings announcement after declining 8% during regular trading. A key part of McGurn's strategy is Truth API, a service that provides Wall Street trading firms with early access to posts from prominent Truth Social users. Among those users is Trump, the platform's most-followed account, who frequently influences financial markets by announcing major U.S. policy changes on the site. Trump Media has faced criticism from government watchdog groups since the beginning of Trump's second term, with critics accusing the company of providing him a way to profit from the presidency. The new service has intensified those concerns, while Democrats have vowed to investigate the paid service if they regain control of Congress in the midterm elections. McGurn dismissed the concerns, saying other companies also provide traders with specialized and faster access to information. "Providing licensed real-time public data through commercial APIs is a well-established business practice across the technology, financial information and media industries," he said. "This is no different." One previously announced venture, nuclear fusion, will continue. McGurn said Trump Media expects to complete its previously announced merger with energy company TAE Technologies by the end of the year. "We continue to believe it's the single most important driver of long-term value for this company," McGurn said of the fusion business. Much of the company's quarterly loss resulted from unrealized losses tied to declines in the value of its bitcoin holdings and a cryptocurrency token called Cronos. Even after excluding those paper losses, along with taxes, interest and other items, the company's operating loss increased, though by a much smaller amount. The adjusted loss was $164 million, compared with $44 million a year earlier. McGurn said Truth API charges customers between $60,000 and $100,000 a month and has already signed 10 customers, most of them high-frequency trading firms that execute transactions within milliseconds. "We're in the early innings," McGurn said, adding that the potential market extends beyond trading firms to data center operators, news organizations and developers of large language models. With 10 customers, the service could generate between $7 million and $12 million in annual revenue, potentially two to three times Trump Media's total revenue last year. Trump Media reported $1.7 million in revenue in the second quarter, more than double the figure from a year earlier. The company has $1 billion in debt through special convertible notes that are not due until 2028. However, lenders have an option to demand repayment in November, which could put pressure on the company's finances, although Trump Media appeared to have sufficient cash. At the end of the quarter, the company held more than $400 million in cash and short-term investments, along with $1.2 billion in bitcoin and bitcoin-related assets.

As the technology world races toward artificial intelligence, a growing number of enthusiasts are looking back to the early days of computing by collecting and restoring vintage computers. Software engineer Josh Dersch is one of them. He spends hours in his garage trying to bring decades-old machines back to life and estimates that he has more than 200 vintage computers and related devices at his home in Seattle. “I’ve always found the origins of things fascinating,” Dersch said. “It’s a fun challenge to figure out what’s wrong and repair it. You can understand the whole system from top to bottom.” Dersch is part of a growing community dedicated to preserving the computers that helped start the digital revolution. While Silicon Valley is now focused on developing AI and other advanced technologies, vintage computer enthusiasts are keeping machines made by Atari, Commodore and IBM alive — systems that paved the way for today’s laptops, smartphones and wearable devices. For collectors, these machines are not outdated junk but important pieces of computer history. Many enjoy hearing the familiar sounds of old machines and playing simple games such as “Pong” and “Space Invaders.” “Interest in the history of computers and everything connected to it is growing,” said Erik Klein, president of the Vintage Computer Federation, which holds an annual festival at the Computer History Museum in Mountain View. Vintage computer festivals draw thousandsMore than 3,500 people attended a two-day Vintage Computer Festival in early August to mark the 50th anniversary of Apple, founded in nearby Cupertino in 1976. The event included a panel discussion with Apple co-founders Steve Wozniak and Ronald Wayne, who recalled the company’s early days as a small startup led by the late Steve Jobs. More than a dozen similar events are now held in cities across North America, South America and Europe. Klein said the growing interest is also driving demand for vintage computers on online marketplaces. Rare and iconic machines, such as the Apple-1 — Apple’s first product — can sell for hundreds of thousands of dollars. At the festival, collectors displayed computers they had carefully restored to working condition, giving visitors a chance to use machines developed long before the internet, social media and AI became part of everyday life. Restoring the ‘grandfather’ of modern computersDersch traveled from Seattle to showcase the prized machine in his collection — a Xerox Alto. He spent years restoring the computer, which is often described as the “grandfather of the modern computer.” It was developed at Xerox’s Palo Alto Research Center in the 1970s. The Alto was among the first computers to move beyond text-based displays and introduce technologies such as the mouse and graphical user interface, including windows, icons and cursors. Those features later became standard in personal computers and are widely credited with influencing early products developed by Jobs and Bill Gates. Dersch’s exhibit included a fully working Xerox Alto workstation with its large vertical monitor, bulky keyboard and mouse. It also featured large disk drives that stored only tiny amounts of data compared with modern storage devices. “It’s fun to show people what I’ve done and see what others have done,” Dersch said. “There’s also a section where I can buy more old machines to take home, so that’s always fun.” Klein, a Silicon Valley software engineering manager, was once an avid collector himself. At one point, he had about 150 computer systems, along with accessories, manuals and original packaging, packed into his closets and garage. But the collection eventually became too much to manage, and Klein worried that it could become a burden for his family. He gradually sold or gave away most of his machines, keeping only a few favorites. Hobby spreading around the worldKlein now puts his passion for vintage computers into organizing festivals, repair workshops, swap meets and other events through the Vintage Computer Federation. “It’s really about sharing the history and encouraging people to maintain, collect and enjoy this vintage technology,” he said. Sellam Ismail, who organized the first vintage computer festival in 1997, said he is pleased to see how much the collector community has grown. The first festival, held at a venue east of San Francisco, attracted about 150 hobbyists. Ismail organized the events for about a decade before helping establish the Vintage Computer Federation, which now runs festivals in the United States and Canada. He has also helped groups organize similar events in Argentina, Germany and Switzerland. “The community grew. More and more people started getting old computers and collecting them,” said Ismail, who once had a huge collection stored in a warehouse. With the festival approaching its 30th anniversary next year, Ismail expects the hobby to continue expanding. “Next year is the 30th anniversary, and I expect this to continue and grow even bigger,” he said.