In this week's edition of the News Roundup, the US is tightening restrictions on chip sales to China, particularly those used in AI models. At the same time, during a recent summit in Silicon Valley, allied spy chiefs have cautioned about Chinese efforts to steal advanced technology such as AI. In Washington DC, Congress has initiated a new round of hearings focusing on AI technology and data privacy, while US Senator Michael Bennet is pressing tech giants for answers on combating the spread of false content related to the Israel-Hamas conflict. Furthermore, due to Hamas’s usage of crypto currency, the US is considering treating crypto anonymity services as suspected money launderers. On a different note, UC Berkeley unveils ambitious plans for a $2 billion space research center in Silicon Valley.
Industrial Policy & International Security
US tightens China’s access to advanced chips for artificial intelligence | The New York Times
The Biden administration has imposed additional restrictions on the sale of advanced semiconductors to China, with a focus on chips used for artificial intelligence models. The new rules are expected to halt most shipments of advanced semiconductors to Chinese data centers. The administration aims to prevent the risk of these chips reaching China via third countries and is requiring chip makers to obtain licenses for shipping to several countries subject to US arms embargoes. US officials argue that China's access to such technology poses national security risks and could enhance the country's military capabilities. The restrictions are set to impact chip makers and Chinese companies working on AI applications.
Intelligence leaders from the United States, the United Kingdom (MI5), Australia, Canada, and New Zealand gathered to address the growing threat of Chinese espionage, particularly focusing on stealing technology. They discussed collaborating with private industry to protect new technologies and maintain an advantage over China. The meeting took place at Stanford University in Silicon Valley, emphasizing the shift of Chinese espionage toward Silicon Valley offices. The intelligence chiefs warned of China's intense interest in Western artificial intelligence and its potential impact on intelligence collection and economic gains. They highlighted China's use of hacking, pressure on Chinese students, informants in Western companies, and joint ventures with Western firms for technology theft.
The House Energy and Commerce subcommittee is starting a series of hearings on artificial intelligence (AI) technology. The first hearing focuses on concerns around how AI systems collect and use data, and lawmakers will hear from a range of experts and witnesses. Amba Kak, executive director of the AI Now Institute, recommends prioritizing data privacy, particularly passing “strong, legally enforceable data minimization mandates.” A federal privacy law could create new requirements for companies collecting and using consumer data, aligning with the growth of AI and its impact on data use.
The US Department of the Treasury has proposed new rules that would designate foreign cryptocurrency “mixers” as money laundering tools that pose a threat to national security. The move comes as a response to the use of cryptocurrency by groups such as Hamas and Palestinian Islamic Jihad. These groups have used cryptocurrency for fundraising, raising millions of dollars, although it's important to note that the majority of cryptocurrency illicit activity is linked to other actors such as North Korean state-sponsored cyber criminals and Russian ransomware gangs. These proposed rules may lead to increased regulatory scrutiny of cryptocurrency mixing services and could impact cryptocurrency transactions' privacy.
OpenAI is hesitant to release a tool that identifies images created with DALL-E 3, their generative AI art model, as it hasn't met their quality threshold. The classifier is highly accurate, with a 99% reliability rate, but OpenAI's aim may be even higher. OpenAI is also grappling with the philosophical question of what constitutes an AI-generated image, particularly in cases where images have undergone extensive editing. The organization is seeking input from artists and those impacted by such tools. Other organizations are exploring watermarking and detection techniques for generative media as deep fakes proliferate, but industry standards have yet to emerge. OpenAI is not committed to supporting images created with non-OpenAI generative tools but may consider it depending on the reception of their current classifier tool.
Famed venture capitalist Marc Andreessen recently published “The Techno-Optimist Manifesto,” championing technology's potential for human progress. He emphasized the need for rapid innovation to solve global issues, highlighting technology’s historical contributions like vaccines and artificial intelligence. Andreessen challenged critics, including regulators and those focused on “tech ethics,” arguing against AI regulation and stating it could save lives. However, his manifesto drew mixed reactions, failing to alleviate concerns about unregulated tech, which has previously led to issues like fraud in the cryptocurrency industry. The need for responsible development in AI and nuclear fusion remains high, and regulators are unlikely to be deterred.
State & Local Tech Ecosystems
UC Berkeley unveils plan for $2B space research center just 7 miles from Stanford | San Francisco Chronicle
UC Berkeley has announced plans to build a $2 billion space research center at NASA Research Park in Mountain View. The Berkeley Space Center will focus on breaking boundaries on Earth and beyond. The project aims to create thousands of new science jobs, a million square feet of office space for private partners, and acres of classrooms and laboratories. UC Berkeley expects the center to generate about $40 million per year in income and to become a major hub for space technology innovation, strengthening California's position in the sector. The center could welcome tenants by 2027 pending environmental approvals.
Silicon Valley Ditches News, Shaking an Unstable Industry | The New York Times
Major online platforms, including Facebook, Twitter (X), and Google, are reducing their support for news content. Social media referral traffic to top news sites in the US declined from 11.5% in September 2020 to 6.5% in September 2021. This shift affects the news industry, which had relied on these platforms for traffic and advertising. The decline is attributed to concerns about polarized debates and changes in algorithms. Google, while still a significant traffic source, has laid off news employees, and some publishers have observed recent declines in Google referral traffic, raising questions about its reliability. Publishers are exploring alternative ways to connect with readers, such as branded newsletters and print magazines.
US Senator Michael Bennet is seeking information from tech giants, including Meta, TikTok, and Google, on how they are combating the spread of false and misleading content related to the Israel-Hamas conflict on their platforms. Deceptive content, including visuals from older conflicts and altered documents, has been widely disseminated on social media platforms. Senator Bennet criticized the companies for not doing enough to address the issue and questioned their content moderation practices. This comes after European Union industry chief Thierry Breton demanded stricter measures to combat disinformation amid the conflict. The companies have recently taken some actions, but Senator Bennet believes their current policies and protocols are inadequate.