The artificial intelligence landscape continues to evolve rapidly, with significant developments impacting various sectors. In education technology, Chegg is laying off 45% of its staff, attributing the decision to the rise of AI tools like ChatGPT and a decrease in search engine traffic, while also announcing Dan Rosensweig is returning as CEO. Meanwhile, the infrastructure supporting AI is facing its own challenges; AI data centers are experiencing a bottleneck due to a shortage of jet engines for backup generators, with lead times extending to 2028-2030. In finance, BlackRock Systematic, managing $336 billion, is leveraging AI and alternative data, analyzing over 1,000 market signals for investment strategies. Microsoft is addressing privacy concerns regarding its Gaming Copilot feature, clarifying that user screenshots are not used for AI model training, though text or voice data may be used to improve AI services. The autonomous driving sector is seeing growth, with Pony.ai launching IPOs in Hong Kong and Nasdaq to fund its research and development. On the hardware front, French researchers have developed a new AI chip capable of learning and inferring on a single device, reducing energy consumption. However, the integration of AI also presents risks, as demonstrated by a security system that mistook chips for a gun, leading to a student being handcuffed, and by the potential security vulnerabilities when AI agents manage cryptocurrency wallets, despite tools like Coinbase's Payments MCP aiming to streamline these transactions.
Key Takeaways
- Chegg is laying off 45% of its staff due to advancements in AI tools like ChatGPT and reduced search engine traffic.
- Dan Rosensweig is returning as CEO of Chegg.
- AI data centers face a significant bottleneck due to a shortage of jet engines for backup generators, with delivery times extending to 2028-2030.
- BlackRock Systematic, managing $336 billion, utilizes AI and alternative data for investment strategies, analyzing over 1,000 market signals.
- Microsoft has stated that screenshots from its Gaming Copilot feature are not used for AI model training, though text or voice data may be used for AI improvement.
- Autonomous mobility company Pony.ai is pursuing IPOs in Hong Kong and Nasdaq.
- A new AI chip developed by French researchers can learn and infer on a single device, reducing energy consumption.
- An AI security system mistakenly identified a bag of chips as a gun, leading to a student being handcuffed.
- AI agents can manage crypto wallets, but security risks such as prompt injection remain a concern.
Chegg lays off 45% of staff due to AI advancements
Chegg, an online education company, is cutting 45% of its workforce, affecting 388 employees. The company cited the rise of artificial intelligence tools like ChatGPT and decreased traffic from search engines as reasons for the layoffs. Chegg also announced that Dan Rosensweig is returning as CEO. The company's stock has significantly dropped since its peak in 2021. Chegg offers services like homework help and tutoring, alongside its own AI tools.
AI data centers face jet engine shortage, delaying expansion
The rapid expansion of AI data centers is facing a major bottleneck: a shortage of jet engines used to power backup generators. Manufacturers are quoting lead times stretching into 2028-2030 for these aeroderivative gas turbines. Companies like GE Vernova and Siemens Energy report high demand, with over 60% of Siemens' US gas turbine orders linked to AI data centers. While these turbines offer fast power, they also have environmental concerns due to emissions. The manufacturing of these complex machines cannot be easily scaled to meet the sudden surge in demand.
Pony.ai launches IPO in Hong Kong and Nasdaq
Autonomous mobility company Pony.ai has launched its initial public offering in Hong Kong, offering 41,955,700 Class A ordinary shares. The shares are expected to list on the Hong Kong Stock Exchange under the code '1234' on November 10, 2025. Pony.ai is also pursuing a concurrent offering in the United States, listing on the Nasdaq under the ticker symbol 'PONY'. The company plans to use the proceeds for general corporate purposes, including research and development and potential acquisitions. Pony.ai is a global leader in commercializing autonomous driving technology.
Microsoft addresses Gaming Copilot data privacy concerns
Microsoft has responded to backlash over its Gaming Copilot feature, which faced criticism for potentially training AI models on user data. Users reported that the tool took screenshots and used OCR to capture in-game text, which was then sent to Microsoft. While Microsoft stated that these screenshots are not used for AI model training and that the feature is optional, they did mention that text or voice conversations might be used to improve AI. Users can disable this feature in the Xbox Game Bar's privacy settings.
BlackRock Systematic uses AI and alternative data for investing
BlackRock Systematic, a division managing $336 billion, employs a rules-based approach using AI and alternative data to generate investment returns. Co-head Jeff Shen explained that their team uses machine learning and AI to analyze over 1,000 market signals, focusing on a three to four-month investment horizon. Examples of alternative data include job posting data to gauge economic health and social media sentiment analysis using natural language processing. This approach aims to deliver consistent alpha, with about 90% of their funds outperforming peers over five years.
AI security system mistakes chips for gun, handcuffs teen
Baltimore County police handcuffed a high school student after an AI security system incorrectly identified a bag of chips as a gun. The incident occurred at a school where the AI system is used for security purposes. Video footage showed officers surrounding and detaining the student. The school is reportedly addressing the situation following the mistaken identification by the artificial intelligence system.
AI agents can manage crypto wallets but risks remain
Coinbase has introduced Payments MCP, a tool allowing AI agents to access crypto wallets and conduct transactions autonomously. While this technology can streamline interactions with decentralized finance, experts caution about security risks. Giving AI access to wallets requires careful prompting and secure trading credentials to prevent issues like prompt injection or man-in-the-middle attacks. Although AI agents can help manage portfolios and offer guidance, users must remain vigilant and review all transactions, as safer systems do not guarantee foolproof security.
New AI chip learns and infers on a single device
French researchers have developed a novel AI chip that can both learn and infer, addressing the hardware challenges of current AI systems. The chip combines ferroelectric capacitors (FeCAPs) and memristors, allowing for efficient learning updates and inference computations on a single device. This integration reduces energy consumption and the need for constant cloud updates, making it suitable for applications like autonomous vehicles and industrial sensors. The technology uses standard 130 nm CMOS processes, suggesting it can be scaled for manufacturing.
Sources
- Chegg slashes 45% of workforce, blames 'new realities of AI'
- Jet engine shortages threaten AI data center expansion as wait times stretch into 2030 — the rush to power AI buildout continues
- PONY AI Inc. Launches Hong Kong Initial Public Offering
- Gaming Copilot AI under fire, but Microsoft says it’s not training AI models on your data
- How BlackRock Systematic Uses AI and Alternative Data
- Teen handcuffed after AI system mistakes bag of chips for gun
- AI agents want to handle your crypto wallet, but is it safe?
- First AI Chip That Learns and Infers