分类： technology

Silicon Valley, the global epicenter of technological innovation, is once again at the forefront of a heated debate: Is the artificial intelligence (AI) sector heading toward a bubble? At OpenAI’s DevDay event this week, CEO Sam Altman addressed concerns head-on, acknowledging the ‘bubbly’ aspects of AI while defending the industry’s potential. ‘There’s something real happening here,’ he asserted, flanked by his top executives. However, skepticism is growing, with critics questioning whether the skyrocketing valuations of AI companies are driven by ‘financial engineering’ rather than sustainable growth. The Bank of England, the International Monetary Fund, and JP Morgan CEO Jamie Dimon have all issued warnings, with Dimon emphasizing that ‘the level of uncertainty should be higher in most people’s minds.’ Early AI entrepreneur Jerry Kaplan, who has lived through four tech bubbles, expressed grave concerns about the current AI frenzy. ‘When [the bubble] breaks, it’s going to be really bad, and not just for people in AI,’ he warned at a panel discussion at Silicon Valley’s Computer History Museum. ‘It’s going to drag down the rest of the economy.’ Despite these warnings, the AI sector continues to attract unprecedented investments. OpenAI, the company behind ChatGPT, recently secured a $100 billion deal with chipmaker Nvidia and announced plans to purchase billions of dollars worth of equipment from Nvidia rival AMD. These complex financial arrangements, coupled with OpenAI’s ambitious Stargate project in Texas, have raised eyebrows. Some experts have labeled these deals as ‘circular financing,’ where companies invest in their own customers to sustain demand. Nvidia CEO Jensen Huang defended the arrangement, stating that OpenAI is not obligated to use Nvidia’s technology exclusively. ‘Our primary goal is just really to support them and help them grow – and grow the ecosystem,’ he said. Meanwhile, the surge in AMD stock suggests that retail investors are eager to capitalize on the AI boom. However, concerns about overinvestment and environmental impact persist. Kaplan warned of ‘enormous data centers in remote places like deserts, that will be rusting away and leaching bad things into the environment.’ Despite the risks, some remain optimistic. Jeff Boudier of Hugging Face likened the current AI infrastructure investments to the overinvestment in telecom infrastructure during the dot-com era. ‘If there is overinvestment into infrastructure for AI workloads, there may be financial risks tied to it,’ he said. ‘But it’s going to enable lots of great new products and experiences.’ As the AI sector continues to evolve, the question remains: Are we witnessing the birth of a transformative technology or the buildup of another financial bubble?

OpenAI’s latest text-to-video AI tool, Sora, has achieved a groundbreaking milestone, surpassing one million downloads in less than five days—a feat that outpaced ChatGPT’s initial launch. The app, which has dominated the Apple App Store charts in the US, transforms simple text prompts into ten-second, hyper-realistic videos. Sora’s rapid adoption was highlighted in an X post by Bill Peebles, the head of Sora, who noted the ‘surging growth’ despite the app being invitation-only and limited to North America. However, the tool’s handling of copyrighted material and depictions of deceased public figures has ignited widespread criticism. Social media platforms have been flooded with user-generated Sora videos, including those featuring late celebrities like Michael Jackson and Tupac Shakur. Zelda Williams, daughter of the late Robin Williams, recently urged people to stop sharing AI-generated videos of her father, linking the trend to Sora’s popularity. OpenAI defended its stance, citing ‘strong free speech interests’ in allowing historical figure depictions but acknowledged that authorized individuals could request the removal of recently deceased public figures’ likenesses. The app has also been used to create deepfake videos, such as one featuring OpenAI CEO Sam Altman alongside Pokémon characters, raising concerns about intellectual property rights. Legal battles loom as companies like Anthropic face lawsuits over AI model training. OpenAI has pledged to adapt its approach, offering rights holders greater control and exploring revenue-sharing models. However, whether these measures will satisfy stakeholders or lead to further legal challenges remains uncertain.

In recent years, a growing trend among Silicon Valley’s elite has sparked widespread curiosity and speculation: the construction of luxury bunkers and underground shelters. Mark Zuckerberg, the founder of Facebook, has reportedly been developing a 1,400-acre compound on Kauai, Hawaii, since 2014, complete with a 5,000-square-foot underground shelter. Despite Zuckerberg’s insistence that it’s merely a basement, rumors persist that it’s a doomsday bunker. Similarly, tech moguls like Reid Hoffman, co-founder of LinkedIn, have openly discussed ‘apocalypse insurance,’ with New Zealand emerging as a popular destination for such investments. The motivations behind these projects remain unclear, but they coincide with rising concerns about artificial intelligence (AI) and its potential to disrupt society. Ilya Sutskever, chief scientist at OpenAI, has even suggested building a bunker for top scientists before releasing artificial general intelligence (AGI), a technology that could match human intelligence. While some, like Elon Musk, envision AGI as a path to ‘universal high income’ and sustainable abundance, others fear its potential to surpass human control. Governments are taking steps to regulate AI, but the debate over its risks and benefits continues to divide experts. As AI advances, the line between innovation and existential threat grows increasingly blurred, leaving humanity to grapple with the implications of its own creations.

Tesla is under scrutiny by the US government following reports that its self-driving vehicles have allegedly violated traffic laws, including instances of driving on the wrong side of the road and failing to stop at red lights. The National Highway Traffic Safety Administration (NHTSA) has identified 58 such incidents in a recent filing, prompting a preliminary evaluation of Tesla’s ‘Full Self-Driving (Supervised)’ mode. This advanced feature, which requires an additional fee, allows vehicles to perform lane changes and turns autonomously, though drivers are expected to remain vigilant and ready to intervene. Approximately 2.9 million Tesla vehicles equipped with this technology are now part of the investigation. The NHTSA report highlights six crashes caused by vehicles stopping at red lights before proceeding prematurely, four of which resulted in injuries. Tesla has reportedly taken corrective measures to address recurring issues at a specific intersection in Maryland. Additionally, the agency is examining reports of vehicles entering opposite lanes during turns, with some incidents offering drivers minimal time to react. This probe adds to Tesla’s existing investigation concerning door-locking mechanisms in Model Y vehicles, where children were allegedly trapped inside, leading some owners to break windows for rescue. Amid these challenges, Tesla has introduced more affordable versions of its popular models to compete with cost-effective electric vehicles, particularly those manufactured by Chinese companies. Elon Musk, Tesla’s CEO and former ally of President Donald Trump, recently launched the America Party, aiming to challenge the dominance of the Republican and Democratic parties. For more updates on global tech developments, subscribe to the Tech Decoded newsletter.