Sign In / Up | Advertise | View Online

NerdNews

July 03, 2025

News & Trends
Amazon Deploys 1 Millionth Robot, Releases Generative AI Model

Amazon has deployed its 1 millionth robot in its warehouses, with 75% of global deliveries assisted by robots. The company also released a new generative AI model called DeepFleet to increase robotic fleet efficiency by 10%.

Everything that could go wrong with X’s new AI-written community notes

X is introducing AI-written community notes to its platform, which could potentially ruin trust in the fact-checking system. The AI agents may generate 'persuasive but inaccurate notes' that human raters might rate as helpful, disrupting the feedback loop and making the system less trustworthy. Experts warn that this could lead to the promotion of 'lies and conspiracy theories' on the platform.

OpenAI Condemns Robinhood's 'OpenAI Tokens'

OpenAI has condemned Robinhood's sale of 'OpenAI tokens', stating they do not represent equity in the company. Robinhood's tokens are tied to the price of OpenAI shares in a special purpose vehicle (SPV), but do not grant direct ownership. OpenAI emphasizes that any transfer of its equity requires approval, which was not given.

Google's Veo 3 and Playable World Models

Google's Veo 3, a video-generating model, could potentially be used for video games. Demis Hassabis, CEO of DeepMind, suggested this in a post. Veo 3 creates realistic video and audio, but it's not yet a world model. Google plans to turn its multimodal foundation model into a world model, simulating aspects of the human brain.

Confidence in Agentic AI

As AI agents enter real-world deployment, organizations face pressure to define their role, build, and operationalize them at scale. Evaluating infrastructure is crucial for confidence in agentic AI reliability. Leaders discuss transforming businesses with agents, tackling complexity, and preparing for growing complexity. They emphasize the need for eval infrastructure, unit testing, and simulating conversations at scale to ensure reliable agent behavior.

Options & Tutorials
Meta's $1,000-Ish AR Glasses

Meta's AR glasses, codenamed 'Meta Celeste', may be the company's first pair of smart glasses capable of replacing phones. The glasses have a built-in display, an EMG wristband controller, and a touch-sensitive panel. They are expected to be priced around $1,000.

NYT to start searching deleted ChatGPT logs after beating OpenAI in court

A US district judge has denied OpenAI's objections to a court order requiring the company to retain all ChatGPT logs, including deleted and temporary chats. The New York Times and other news organizations will soon start searching these logs as part of a copyright case, sparking concerns over user privacy and potential market implications.

A guide to using Edits, Meta's new CapCut rival

Meta's Edits is a new video editing app designed to rival CapCut, offering features like green screen, timeline, captions, and audio library. The app is currently free and mobile-only, with plans to add paid features in the future.

Grok 4 Spotted Ahead of Launch with Special Coding Features

xAI is releasing Grok 4 after Independence Day, skipping Grok 3.5. Grok 4 will have 'Thinking—Bigger and Smarter' features, including text modality and coding capabilities. It will be available on the xAI Console, allowing developers to access the Grok API.

Elon Musk’s X Is Turning Community Notes Over to AI

Elon Musk's X is planning to deploy AI agents to help with Community Notes, a fact-checking service on the platform. The AI-generated notes will be reviewed by human moderators before being published. This move comes as human-generated notes have decreased significantly, with the number of notes published on the platform cut in half from 120,000 in January to 60,000 in May of 2025.

Launches & Tools
Genesis AI launches with $105M seed funding

Genesis AI, a startup building AI models for robots, has launched with $105M in seed funding from Eclipse Ventures and Khosla Ventures. The company aims to develop a general-purpose model that enables robots to automate repetitive tasks using synthetic data generated by a proprietary physics engine.

Racist AI Videos Created with Google's Veo 3 Proliferating on TikTok

Google's Veo 3 AI video generator is being used to create racist and antisemitic content on TikTok, despite both companies having policies against such content. The videos are spreading quickly, with some gaining millions of views, and highlight the challenges of enforcing moderation policies on social media platforms.

NimDoor macOS Malware

Researchers discovered a new family of macOS malware called NimDoor, used by North Korean state-backed hackers to target web3 and cryptocurrency organizations. The malware uses unusual techniques and a previously unseen signal-based persistence mechanism, making it resilient to basic defensive actions.

Cisco Patch for Critical Flaw

Cisco has issued an urgent patch for a critical flaw in its Unified Communications Manager and Session Management Edition, which could allow an unauthenticated, remote attacker to gain full root control of a system. The flaw, rated 10.0 on the CVSS scale, is due to hardcoded credentials that cannot be changed or deleted. Admins are advised to upgrade to the newest code to fix the issue.

The One Big Beautiful Bill is one big disaster for AI

The One Big Beautiful Bill cuts funding for clean energy, making it harder for AI companies to power their data centers, and gives China an advantage in the AI race. The bill also cuts safety net programs, which could worsen the impact of job loss due to automation.

Quick Links
Asia Tech News Roundup

Canada has ordered Chinese CCTV vendor Hikvision to cease operations due to national security concerns. Meanwhile, Japan's SoftBank will fly broadband blimps to expand 5G coverage, and Starbucks China has been criticized for prioritizing ads over user privacy.

Tailor Raises $22M Series A

Tailor, a San Francisco- and Tokyo-based enterprise resource planning (ERP) platform, has raised $22 million in a Series A funding round. The 'headless' ERP system separates the front end from the back end, allowing for independent selection or development of the front end and enabling AI agents to securely access its ERP system via API to automate tasks.

Google bets on nuclear fusion

Google has agreed to purchase 200 megawatts of carbon-free power from a nuclear fusion power plant being developed by Commonwealth Fusion Systems. This deal marks Google's first offtake agreement for fusion and demonstrates the company's confidence in the technology. However, the timeline for nuclear fusion is uncertain, and it may not be commercially viable in time to help Google meet its 2030 emissions goals.

Amazon Now Has 1 Million Robots

Amazon has deployed over 1 million robots in its warehouse facilities, with the goal of increasing productivity and efficiency. However, this shift towards automation has raised concerns about job security for human employees, who may be replaced by robots in the future. Despite the potential benefits of automation, data shows that Amazon warehouse workers are still experiencing high injury rates and poor working conditions.

Microsoft cancels Perfect Dark and Everwild Xbox games

Microsoft is canceling its Perfect Dark and Everwild games as part of sweeping cuts to its Xbox division. The company will be closing The Initiative, the studio behind Perfect Dark. The cuts are part of layoffs affecting around 9,000 employees across Microsoft.

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.