anthropic unveils new tools as google ships new models

The Los Angeles Unified School District's ambitious "Ed" AI chatbot project, a $6 million initiative led by Superintendent Alberto Carvalho, has collapsed and is now under federal investigation. Launched in March 2024 with startup AllHere, the project failed, and AllHere's founder, Joanna Smith-Griffin, faces fraud charges. Federal agents have raided Carvalho's home and office, as well as a consultant's residence, though Carvalho himself has not been charged. Critics point to a lack of transparency and efficacy in the deal, despite Carvalho's recent reappointment.

Ethical considerations surrounding AI's military applications are also in the spotlight, with AI firm Anthropic rejecting the Pentagon's demand for unrestricted access to its AI tool, Claude. Anthropic explicitly stated it would not permit its technology for fully autonomous weapons or mass domestic surveillance. This decision has led to a deepening standoff, intensified by a dispute over a discussion concerning a hypothetical nuclear strike, underscoring the critical ethical boundaries in defense AI.

Meanwhile, AI continues to integrate into various sectors. A Google executive reportedly dedicates over 20 hours weekly to exploring AI tools like Gemini, using them as a sounding board and for rapid software prototyping to enhance productivity. Microsoft researchers have also introduced On-Policy Context Distillation (OPCD), a new training method that enables AI models to internalize lengthy system prompts, improving performance and reducing computational overhead by training models with their own generated outputs.

In the financial world, Pictet Asset Management launched the Pictet AI Enhanced US Equity ETF (PQUS), an actively managed fund that uses AI for stock selection in large-cap US equities. However, the rapid expansion of AI also presents challenges and misuses. A West Point cadet, Cayden Cork, was expelled from the Army for using generative AI to create fake nude images for extortion. Additionally, a Washington state hotline inadvertently began using an AI voice speaking English with a Spanish accent for Spanish-language services, highlighting accessibility issues.

The fashion industry, too, is grappling with AI's impact, with Prada co-designers Miuccia Prada and Raf Simons citing AI as a major challenge, even as tech companies like Meta explore collaborations. Furthermore, the high demand for AI data center infrastructure is driving a severe global RAM shortage, with memory chip costs soaring by up to 90%. Analysts predict this shortage will persist through 2026 and potentially into 2028, affecting various consumer products.

Key Takeaways

  • LAUSD's $6 million "Ed" AI chatbot project, developed with AllHere, failed and is now under FBI investigation, with the startup's founder facing fraud charges.
  • AI firm Anthropic rejected the Pentagon's demand for unrestricted access to its Claude AI tool, refusing its use for fully autonomous weapons or mass domestic surveillance.
  • A Google executive dedicates over 20 hours weekly to experimenting with AI tools like Gemini to enhance productivity and prototype software.
  • Microsoft researchers developed On-Policy Context Distillation (OPCD), a new AI training method that allows models to internalize lengthy prompts, improving efficiency.
  • Pictet Asset Management launched the Pictet AI Enhanced US Equity ETF (PQUS), an actively managed fund using AI for large-cap US equity selection.
  • A West Point cadet was expelled from the Army after being convicted of extortion for using generative AI to create fake nude images.
  • The high demand for AI data center infrastructure is causing global RAM prices to surge by up to 90%, with shortages projected to continue through 2026, possibly into 2028.
  • Prada co-designers Miuccia Prada and Raf Simons identified artificial intelligence as a primary challenge facing the fashion industry, alongside inequality.
  • A Washington state hotline for Spanish-speaking callers mistakenly uses an AI voice speaking English with a Spanish accent, highlighting unexpected AI implementation issues.
  • The standoff between Anthropic and the Pentagon includes a dispute over a discussion regarding a hypothetical nuclear strike, emphasizing the sensitive nature of military AI integration.

LAUSD Superintendent's AI Bet Fails, Sparks FBI Raids

Los Angeles Unified Superintendent Alberto Carvalho's ambitious AI chatbot project, "Ed," launched in March 2024, collapsed within months. The multimillion-dollar deal with startup AllHere, which later went bankrupt, is now under FBI investigation. Federal agents raided Carvalho's home and office, as well as a home owned by consultant Debra Kerr, who helped broker the deal. AllHere's founder, Joanna Smith-Griffin, faces charges of defrauding investors. While Carvalho and Kerr have not been charged, sources indicate the investigation focuses on financial issues related to the failed chatbot contract.

Failed $6M AI Chatbot at LAUSD Now Under Federal Investigation

The "Ed" AI chatbot, intended to be an educational tool for Los Angeles students, failed to deliver on its promises and is now the focus of a federal probe. The $6 million project with startup AllHere collapsed after the company's founder, Joanna Smith-Griffin, was charged with fraud. Federal agents recently raided the homes of Superintendent Alberto Carvalho and Smith-Griffin. Despite the chatbot's failure and the ongoing investigation, Carvalho was reappointed to his position. Critics argue the AI deal lacked transparency and efficacy.

AI Firm Anthropic Refuses Pentagon's Demand for Lethal Weapon Use

AI company Anthropic has rejected the Pentagon's demand for unrestricted access to its AI tool, Claude. Anthropic stated it will not allow its technology to be used for fully autonomous weapons or mass domestic surveillance. This decision comes amid a deepening standoff between the AI firm and the Department of Defense over the terms of access to advanced AI capabilities. The disagreement highlights the ethical considerations and potential risks associated with military applications of artificial intelligence.

Google Exec Spends 20+ Hours Weekly Exploring AI Tools

A Google executive reveals he dedicates over 20 hours each week to experimenting with AI tools like Gemini. He uses AI as a sounding board for ideas and to help prototype new software rapidly. The executive emphasizes that AI enhances his productivity as a software engineer, allowing him to focus on complex tasks like technical architecture. He believes embracing AI is crucial for staying relevant and productive in the evolving tech landscape.

Pictet Launches New AI-Driven US Equity Exchange Traded Fund

Pictet Asset Management has introduced the Pictet AI Enhanced US Equity ETF (PQUS), an exchange-traded fund utilizing artificial intelligence for stock selection. This actively managed fund focuses on large-cap US equities, aiming to generate alpha with minimal tracking error against a major US equity benchmark. The ETF is designed for investors seeking AI-enhanced exposure to both US and international developed markets.

Microsoft Develops New AI Training Method to Reduce Prompt Length

Microsoft researchers have created a new AI training method called On-Policy Context Distillation (OPCD). This technique allows AI models to internalize information from lengthy system prompts, eliminating the need for them during inference. OPCD trains models using their own generated outputs, improving performance and reducing computational overhead and latency. This method addresses issues like "exposure bias" and "forward hunting" found in older distillation techniques, leading to more efficient and accurate AI applications.

Pentagon Showdown with AI Firm Anthropic Over Hypothetical Nuclear Attack

A conflict between AI company Anthropic and the Pentagon has intensified over the use of AI in military applications. The dispute centers on differing accounts of a discussion regarding a hypothetical nuclear strike against the United States. This disagreement highlights the critical and sensitive nature of integrating advanced AI technologies into defense strategies and the ethical boundaries involved.

West Point Cadet Expelled for Extortion Using AI Nudes

A West Point cadet, Cayden Cork, has been dismissed from the Army after being convicted of extortion and indecent conduct. Cork used generative AI to create fake nude images of a woman and threatened to release them if she didn't send him actual nude photos. He pleaded guilty to the charges and was sentenced to be reprimanded and dismissed from the Army. The incident highlights the misuse of AI technology for criminal purposes and the military justice system's response to such crimes.

Prada Designers Cite Inequality and AI as Fashion's Biggest Challenges

Miuccia Prada and Raf Simons, co-designers of Prada, identified inequality and artificial intelligence as the primary challenges facing the fashion industry. Speaking at Milan fashion week, they presented a more focused collection reflecting everyday life. The designers' comments come amidst discussions about AI's growing influence and potential impact on the industry, including collaborations with tech companies like Meta.

AI Drives RAM Prices Up Dramatically, Shortage Expected

The global RAM market is facing a severe price crisis, with memory chip costs soaring by up to 90%. This surge is driven by high demand for AI data center infrastructure, causing manufacturers to shift production away from consumer products. Analysts predict this shortage will continue through 2026 and possibly into 2028, significantly impacting the availability and cost of memory for various devices.

Washington State Hotline Accidentally Uses AI with Spanish Accent

Callers to a Washington state hotline seeking service in Spanish are unexpectedly hearing an AI voice speaking English with a Spanish accent. This issue arose after the Department of Licensing updated its automated system to include 10 languages using AI technology. While the agency is working to fix the error, the incident highlights accessibility challenges and the unexpected outcomes of AI implementation. The specific AI vendor involved has not yet been disclosed.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI Chatbot FBI Investigation Startup Failure Fraud Charges Educational Technology Autonomous Weapons AI Ethics Military AI AI Productivity Software Development AI Investment Exchange Traded Fund AI Training Methods Prompt Engineering Generative AI AI Misuse Fashion Industry RAM Prices AI Data Centers AI Implementation Accessibility

Comments

Loading...