Sign In / Up | Advertise | View Online

NerdNews

October 17, 2025

News & Trends
Microsoft Copilot Updates

Microsoft has introduced new features to its Copilot AI tool, including voice activation and screen analysis. The updates allow users to interact with their PCs using voice commands and enable Copilot to analyze the screen and provide helpful tips. The features are being rolled out to Windows 11 devices, with a gradual rollout expected. Microsoft has also announced plans to integrate Copilot into the taskbar, replacing the search box with an 'Ask Copilot' function.

The $100B memory war: Inside the battle for AI's future

The race to supply High Bandwidth Memory 4 (HBM4) is heating up, with SK Hynix, Micron, and Samsung competing to meet the demands of AI hardware. HBM4 promises unprecedented bandwidth per chip, with twice the bandwidth of current HBM3 chips. SK Hynix is currently the front-runner, having begun sampling HBM4 in March 2025 and announcing mass production readiness. Micron and Samsung are also pushing to deliver HBM4, with Micron shipping samples in June 2025 and Samsung aiming for mass production in the first half of 2026.

ChatGPT to Allow Explicit Content

OpenAI's ChatGPT will now allow adult users to access explicit interactive experiences, marking a shift in the company's practices. This move may attract paying subscribers but raises concerns about user safety and potential mental health issues.

AI Attacks Surge as Microsoft Process 100 Trillion Signals Daily

Microsoft's systems analyze over 100 trillion security signals daily, indicating a surge in AI-driven attacks. The company's Digital Defense Report 2025 warns that AI is central to both defending and compromising the digital world, with adversaries using generative AI to automate phishing and scale social engineering. Microsoft recommends urgent actions for leaders, including treating cybersecurity as a board-level risk and enforcing phishing-resistant MFA.

OpenAI Pauses Sora Video Generations of Martin Luther King Jr.

OpenAI has paused the ability for users to generate videos resembling Martin Luther King Jr. using its AI video model, Sora, after some users created 'disrespectful depictions' of his image. The company is adding safeguards at the request of Dr. King's estate.

Options & Tutorials
Why we should all be worried about the AI bubble

The AI investment bubble is growing, with warnings from the Bank of England and others. Experts argue that AI is not a superintelligence, but rather 'normal technology' with slow and uncertain adoption. The bubble is driven by narrative and investor hype, with companies like Meta seeking billions in private capital. A collapse of the bubble could have severe global economic consequences.

Anthropic brings mad Skills to Claude

Anthropic has introduced Skills to its AI model Claude, allowing it to interact with specific applications and perform tasks such as creating spreadsheets or presentations. Skills consist of a directory with a SKILL.md file and can be stored locally or uploaded to the cloud for use with the Claude API. The company has also integrated Claude with Microsoft 365, enabling enterprise search across connected data sources.

US Hyperscalers' Power Consumption to Rise by 22% by End of 2025

US hyperscale datacenters are expected to consume 22% more grid power by the end of 2025, driven by demand for AI and machine learning. This growth is forecast to continue, with power consumption nearly tripling by 2030. The increase is largely fueled by the construction of new datacenters, particularly in states like Virginia and Texas.

How ByteDance Made China's Most Popular AI Chatbot

ByteDance's Doubao app has become China's most popular AI chatbot, with over 157 million monthly active users. Its success can be attributed to its user-friendly design, integration with Douyin, and ability to generate images, spreadsheets, and other content. Doubao's popularity has surpassed that of DeepSeek, another Chinese AI chatbot, and it has become a major player in the country's AI market.

Why AI startups are taking data into their own hands

AI companies like Turing are taking data collection into their own hands, paying top dollar for carefully curated data to train their models, rather than relying on low-paid annotators or scraped web data. This shift is driven by the need for high-quality, proprietary training data as a competitive advantage.

Launches & Tools
Waymo plans to launch a robotaxi service in London in 2026

Waymo, an Alphabet-owned company, plans to launch a commercial robotaxi service in London in 2026. The service will use all-electric Jaguar I-Pace vehicles equipped with self-driving technology. Waymo will start with human safety drivers before transitioning to driverless testing and eventually inviting the public to hail its robotaxis.

Honor's new AI Robot phone

Honor has unveiled a concept for an AI-powered robot phone that integrates multi-modal intelligence, robotic functionality, and advanced imaging capabilities. The phone is set to be showcased at MWC in Barcelona next year.

General Intuition lands $134M seed to teach agents spatial reasoning

General Intuition, a startup spun out of Medal, has raised $133.7M in seed funding to develop AI agents that can understand spatial-temporal reasoning using video game clips. The company aims to apply this technology to gaming, search-and-rescue drones, and eventually, artificial general intelligence.

Amazon launches pay-per-visit virtual healthcare service for kids

Amazon One Medical has launched a pay-per-visit virtual healthcare service for children aged 2-11, with message-based visits starting at $29 and video visits at $49, covering issues like pink eye and skin problems

Microsoft Disrupts Ransomware Attacks Targeting Teams Users

Microsoft has disrupted a wave of Rhysida ransomware attacks targeting Teams users by revoking over 200 certificates used to sign malicious Teams installers. The attacks were carried out by the Vanilla Tempest threat group, which used fake Microsoft Teams installers to infect victims with the Oyster backdoor.

Quick Links
Microsoft patches critical ASP.NET Core vulnerability

Microsoft has patched a critical vulnerability in ASP.NET Core, rated 9.9 on the CVSS scale. The flaw, known as request smuggling, allows an attacker to hide an extra request inside another one, potentially bypassing security checks. The vulnerability affects all supported versions of ASP.NET Core and can be patched by updating the .NET SDK or the Kestrel.Core package.

UK Tech Grad Jobs

The UK tech sector is cutting graduate jobs dramatically, down 46% in the past year, due to AI performing entry-level tasks. Companies are hiring experienced workers instead of training newcomers, creating a vicious cycle for graduates.

Securing the AI era: Huawei’s cyber security strategy for the GCC

Huawei is positioning itself at the forefront of cyber security efforts in the Gulf Cooperation Council region, combining global expertise with local compliance to help enterprises and governments navigate cyber security and AI adoption. The company's approach begins with compliance and includes a 'secure by design' approach, leveraging AI to enhance security tools and detect cyber attacks.

F5 Hack Puts Thousands of Networks at Risk

A sophisticated threat group breached F5's network, gaining access to BIG-IP source code and customer configurations, posing an imminent threat to thousands of networks, including those operated by the US government and Fortune 500 companies.

Microsoft, AWS, and Google Reduce China's Supply Chain Role

Microsoft, AWS, and Google are moving production outside of China due to geopolitical tensions, aiming to reduce China's role in their supply chains by 2026, with Microsoft targeting 80% of its components to be manufactured outside of China

Share NerdNews

Share your affiliate link to get commission!

https://nerdnews.online/affiliate

Thanks for reading,
The NerdNews Team

Sign In / Up

If you dont want to receive future editions of NerdNews, Unsubscribe here.