google, openai and anthropic Updates

Google maintains its position as the world's most popular internet service, a spot it has held since 2022. However, the landscape of online activity is rapidly evolving with the swift rise of generative AI services. OpenAI's ChatGPT, for instance, has quickly climbed into the top 40 most visited sites, boasting 700 million active users by September. This growth signals a significant shift, with other major players like Anthropic's Claude and Google Gemini intensifying the competition in the AI space, alongside established tech giants such as Apple and Microsoft. As AI becomes more integrated into global industries, new security challenges emerge, demanding robust protection for AI systems and their complex supply chains. Attackers can exploit vulnerabilities through methods like data poisoning or prompt injection, potentially leading to intellectual property theft or system disruption. Recognizing these threats, the National Institute of Standards and Technology (NIST) recently released a draft guide to help organizations manage AI cybersecurity risks and strengthen their defenses. This guide also addresses how AI can be used to improve security and counter AI-powered attacks, with public comments open until January 30. The legal and ethical implications of AI are also coming into focus. In Hawaii, courts are addressing instances where lawyers used AI to generate fake legal citations, with one attorney, Mark Valencia, found to have violated Rule 11. A Hawaii Supreme Court committee believes current rules are adequate, emphasizing the need for proper training and human oversight when using AI tools. Beyond legal frameworks, the U.S. National Science Foundation, with partners like the Walton Family Foundation, launched the $9 million CAMEL program to integrate AI into K-12 math education, preparing students for future tech careers. AI's influence extends into consumer behavior and philosophical discussions. Many holiday shoppers are turning to AI agents for gift recommendations, with a survey showing 74% trust AI advice as much as friends. On a broader scale, Elon Musk predicts AI and robotics will usher in a "universal high income," making traditional jobs optional and eliminating poverty, though he notes finding meaning in life could become the primary challenge. Meanwhile, pop music critics like Jon Caramanica observe that AI-generated music often lacks emotional depth, raising questions about creativity and the ethical sourcing of its learning data. An ancient Jewish debate even mirrors today's discussions about AI superintelligence, questioning if an all-knowing AI might diminish human choice.

Key Takeaways

• Google remains the world's most popular internet service, while generative AI services like OpenAI's ChatGPT are rapidly growing, with ChatGPT reaching 700 million active users by September.
• The generative AI market sees intense competition from OpenAI's ChatGPT, Anthropic's Claude, and Google Gemini, with other tech giants like Apple and Microsoft also maintaining top online service positions.
• AI systems face new cybersecurity threats, including data poisoning and prompt injection, which target the entire AI supply chain from data to deployment.
• NIST released a draft guide to help organizations manage AI cybersecurity risks, use AI to enhance defenses, and combat AI-powered attacks, inviting public comments until January 30.
• Hawaii courts confirm that existing legal rules, such as Rule 11, are sufficient to address AI misuse by lawyers, emphasizing the necessity of human oversight and content verification.
• The U.S. National Science Foundation, in partnership with the Walton Family Foundation, launched the $9 million CAMEL program to integrate AI, learning science, and data science into K-12 math education.
• A significant number of holiday shoppers, 83%, plan to use AI shopping agents, with 74% trusting AI recommendations as much as advice from friends.
• Elon Musk predicts that AI and robotics will lead to a "universal high income," making traditional jobs optional and eliminating poverty, with an 80% chance of this positive outcome.
• AI-generated music often lacks the subtle emotional changes of human voices, raising questions about its creative depth and the ethical implications of its training data.
• Philosophical discussions highlight concerns that creating an all-knowing AI, even if aligned with human values, could potentially diminish human choice and the meaning of life.

Businesses Must Secure AI Systems From New Cyber Threats

Artificial intelligence is now a key part of many global industries, but it brings new security challenges. Attackers can exploit AI systems through data poisoning, prompt injection, or by manipulating and stealing models. These attacks can lead to intellectual property theft, biased results, or system disruption. Experts like Dr. Vrizlynn L. H. Thing warn that attackers target the AI supply chain. Organizations need to harden their training pipelines and use verified datasets to protect their AI.

Protecting the AI Supply Chain Is Now Crucial

AI supply chain security is vital to protect models and data from tampering and unauthorized access. This supply chain includes everything from raw data to a live AI system, covering data, development, model artifacts, deployment, and third-party components. Weaknesses in one area can affect the entire system. Many organizations now use AI in production with sensitive data, but often do not know the origin or full access of their AI components. Securing the AI supply chain means understanding these complex connections, not just individual parts.

Hawaii Court Rules Can Handle AI Misuse by Lawyers

Lawyers in Hawaii have used AI to create fake legal citations in court documents. One lawyer, Mark Valencia, was found to have violated Rule 11 by Circuit Judge Jordon Kimura after submitting a brief with non-existent cases. A Hawaii Supreme Court committee report states that current rules, like Rule 11, are strong enough to prevent AI abuses. The report suggests lawyers use AI tools with proper training and human oversight. While some judges have not sanctioned lawyers for these mistakes, federal courts in Hawaii require disclosure of AI use and verification of all content.

AI Shopping Agents Help Find Holiday Gifts

Many holiday shoppers are using personal AI agents to help them find gifts this year. A recent survey shows that 74% of shoppers trust AI recommendations as much as advice from friends. With consumer confidence lower, these AI tools offer value, convenience, and inspiration. Retailers are investing in AI bots to assist the 83% of shoppers planning to use artificial intelligence for their purchases. Forbes contributor Kristen Classi-Zummo notes that AI can deliver all three benefits.

Ancient Debate Shows AI's Challenge to Human Choice

A 2,000-year-old Jewish debate between Rabbi Eliezer and Rabbi Yoshua highlights a major problem with artificial intelligence. Rabbi Yoshua argued that humans, not a divine voice, should decide the law, which mirrors today's discussion about AI superintelligence. The article questions if creating an all-knowing AI, even if aligned with human values, would take away human choice and meaning in life. Philosopher John Hicks' idea of "epistemic distance" suggests that allowing human agency is important. AI companies face challenges in defining what makes AI "good" or "helpful," often leading to superficial ethical guidelines.

NIST Releases New AI Cybersecurity Guide

The National Institute of Standards and Technology, NIST, has released a new guide for AI security. This draft document helps organizations manage cybersecurity risks with AI, use AI to improve their defenses, and stop AI-powered attacks. It connects these AI concerns to NIST's well-known Cybersecurity Framework. Barbara Cuthill, an author, noted that all organizations will face these three areas. NIST gathered ideas from over 6,500 people and is now asking for public comments until January 30. This guide is part of NIST's ongoing efforts to provide security standards for AI.

Pop Critic Explores Sadness in AI Music Hits

Jon Caramanica, a pop music critic for the New York Times, discusses why AI-generated music often sounds sad and emotionally empty. He points out that songs like Solomon Ray's "Find Your Rest" lack the small emotional changes found in human voices. This music often targets feelings of sadness and uses styles from Black music, raising questions about who creates it and what it learns from. Caramanica notes that human vocals have been filtered for years, making it easier for AI to replicate this smooth sound. He hopes that future AI music, with better ethics and technology, will find more creativity.

NSF Invests 9 Million in AI Math Education

The U.S. National Science Foundation, with partners like the Walton Family Foundation, launched a new program called CAMEL. This $9 million investment aims to improve K-12 math education by combining artificial intelligence, learning science, and data science. Kaye Husbands Fealing from NSF SBE stated that this program will strengthen math education and prepare students for future jobs in AI and other technologies. CAMEL has two phases, first creating research networks for AI-ready data, then building a national collaboratory to support ongoing collaboration. This effort supports the National AI Research and Development Strategic Plan.

Elon Musk Predicts AI Will Bring Universal High Income

Elon Musk believes that artificial intelligence and robotics will create a "universal high income," making traditional jobs optional and saving money unnecessary. He shared this view in response to investor Ray Dalio's post about new savings accounts. Musk predicts a future with no poverty and plenty of goods and services. He stated at the Viva Technology conference in May 2024 that there is an 80% chance of this positive outcome. In this future, the biggest challenge for people will be finding meaning in their lives, as work could become more like a hobby.

Google Stays Top Online While AI Tools Grow Fast

A new report from Cloudflare shows that Google remains the world's most popular internet service, holding the top spot since 2022. However, generative AI services are growing very quickly, with OpenAI's ChatGPT now ranking among the top 40 most visited sites. ChatGPT had 700 million active users by September, but faces competition from tools like Anthropic's Claude and Google Gemini. Other top services include Facebook, Apple, and Microsoft, while TikTok dropped to eighth place overall. The report highlights the rapid rise of AI in online usage.

Sources

• The Need for AI Security as Organizations Adopt AI Models and Technology
• AI Supply Chain Security: Why It's Becoming Harder to Ignore
• Hawaiʻi Supreme Court Says Current Rules Are Enough To Prevent AI Abuses
• Let Your Personal AI Shopping Agent Pick The Perfect Holiday Gift
• The 2,000-year-old debate that reveals AI’s biggest problem
• NIST adds to AI security guidance with Cybersecurity Framework profile
• Video: Why Are A.I. Hits So Sad?
• NSF and partners invest $9M in AI-focused math education program
• Musk predicts AI will create 'universal high income' and make saving money unnecessary
• Google retains spot as world's most popular internet service, but AI is growing fast