Welcome, Busy Bees!
The race to build the world's most powerful supercomputer just got an incredible cash injection. Today, deep-tech startup Oratomic locked down a massive $300 million funding round led by legendary billionaire investor Vinod Khosla and his firm, Khosla Ventures. Elite firms ARCH Venture Partners and Spark Capital also joined the deal. Oratomic is building a next-generation quantum computer using trapped neutral atoms to process complex data at speeds that make today's standard silicon chips look like basic calculators. Khosla openly declared on X that this is one of his firm's largest foundational investments yet. For business leaders, it is a clear warning that the underlying hardware power backing the next wave of frontier AI models is scaling up much faster than anyone expected.
Simultaneously, the era of just "chatting" with AI is officially over. Today, OpenAI launched a massive upgrade called "ChatGPT Work" . Instead of just replying with paragraphs of text, this new system lets you turn simple text prompts directly into live, interactive files, like fully formatted budget spreadsheets, presentation decks, and mini web applications. It is a direct shot at Microsoft Office and Google Workspace. OpenAI is betting that busy professionals don't want a chatbot that gives them advice; they want an automated coworker that does the actual paperwork for them.
OpenAI drops "ChatGPT Work" to let you build live files instead of just text chats.
Oratomic lands $300M to build a new type of supercomputer using neutral atoms.
Chinese giant DeepSeek is building its own chips to avoid relying on Nvidia hardware.
Anthropic releases new research on how to safely "turn off" dangerous AI features
…and a whole lot more that you can read about below.
What's Happening in AI
OpenAI Launches "ChatGPT Work" to Automate Office Tasks
OpenAI officially launched a new workspace ecosystem called "ChatGPT Work". The big upgrade introduces a feature called "Sites," which lets users build functional web views, budget sheets, and presentation structures from a single chat prompt. Instead of copy-pasting answers into separate applications, users can now generate whole files inside the platform, marking a massive challenge to Microsoft Excel and PowerPoint.
DeepSeek Builds Its Own AI Chip to Lower Compute Costs
Industry reports released today reveal that Chinese AI leader DeepSeek has quietly started designing its own in-house computer chips [10.5]. The move is designed to help the company run its popular open-weight models without relying on expensive, supply-constrained hardware from Nvidia. If successful, it will allow DeepSeek to keep its low pricing advantage, which has already drawn a massive wave of Western software developers to its platform.
Anthropic Invents a Smart "Off-Switch" for Dangerous AI Habits
Anthropic published a milestone safety research paper today detailing a system called GRAM . This technology acts like a targeted software filter, allowing engineers to block specific high-risk model capabilities, such as giving instructions on how to build cyber weapons, without messing up the model's overall intelligence or general helpfulness.
Money Moves
Oratomic Raises $300M Series A for Next-Gen Supercomputing
Deep-tech vanguard Oratomic secured a massive $300 million funding injection today co-led by Khosla Ventures and ARCH Venture Partners. Oratomic uses a unique method involving laser-trapped atoms to process high-level data far faster than standard silicon chips. The company plans to use the massive runway to build an active, error-free quantum system, shifting the technology out of university labs and directly into commercial enterprise use cases.
Aukera Secures $10.7M to Scale Advanced Materials Tech
Industrial tech upstart Aukera closed a fresh capital funding round today led by Alteria Capital. The company specializes in manufacturing high-precision, lab-grown substrates. While the core tech is used for luxury goods, the underlying material process is increasingly critical for building high-performance heat-dissipation layers in data servers and advanced optical computer components.
Uvera Closes Seed Funding to Automate Food Tracking
Foodtech startup Uvera officially locked down its latest seed funding round today. The platform combines specialized computer vision cameras with predictive machine learning models to track food freshness in real-time. By automatically warning logistics managers about early food decay, the technology helps large supply chains prevent inventory waste before it hits the grocery shelves.
Today's Job Board
Fresh opportunities from leading AI companies and high-growth startups.
Senior ML Engineer - Neural Rendering | Torc Robotics
Remote - US, Ann Arbor, MI / Full-time
Developing advanced neural rendering models and simulation graphics architectures to train autonomous vehicle fleet cameras in unstructured spatial environments.
AI Engineer, Internal Enablement & Productivity | Air
Pittsburgh, Pennsylvania, United States / Full-time
Building customized generative AI tools and context layers to automate internal corporate operations, eliminating back-office developer and marketing bottlenecks.
Senior Engineer – AI Native Products | Honeycomb Insurance
Tel Aviv, Israel / Full-time
Architecting core machine learning pipelines and real-time computer vision models to automate commercial real estate risk underwriting and property data processing.
Senior ML Research Engineer, Marengo | TwelveLabs
Remote (Seoul, South Korea) / Full-time
Researching and optimizing foundational video-understanding multimodal architectures, scaling long-horizon sequence alignment models for complex multi-hour streaming video context arrays.
Senior Frontend Engineer | Fairmarkit
Poland - Warsaw / Full-time
Designing clean, component-driven user interfaces and automated dashboard layouts for enterprise supply chain procurement software systems.
Senior Software Engineer, Frontend | Afresh
Remote - U.S. / Full-time
Building high-performance data workflows and visual layout interfaces that store manager inputs to optimize fresh food inventory ordering across massive grocery networks.
Engineer / Developer – Freelance | SPACE44
Remote (Latin America) / Full-time Contract
Deploying full-stack software modules and custom database connectors to support cross-functional web application build outs for high-growth tech startups.
AI TOOL WORTH YOUR TIME
If you want to test it, I’ve got a link for a free month
Check it out!
Today's Learning Edge
Universal Tool Connections (MCP)
The Context: With companies like Omneky releasing open Model Context Protocol (MCP) tools today, the way AI models interact with software has completely changed. Instead of writing complicated, custom code to connect an AI to your inbox, database, or favorite design app, MCP acts like a universal plug. Learning how to set up these standard connections is a massive upskilling advantage, letting you build automated workflows where an AI can open, read, and edit files across your computer automatically.
The Resource: Check out the official Omneky API Integration Guides or review the Anthropic MCP Open Source Framework to learn how to link your daily tools to the web.
Actionable Takeaways:
Stop Writing Custom Code: Use universal MCP plugins to connect your business tools instead of wasting time building custom scripts for every single software tool.
Create Shared Style Guides: Set up simple data templates so any AI tool you use instantly understands your company’s brand colors, fonts, and goals.
Enable Smart Two-Way Editing: Build loops where the AI doesn't just read your files, but can actively update information inside your spreadsheets and design apps for you.
Prompt Drop
The Office Artifact Creator
Take full advantage of OpenAI's new file-generation focus by forcing the model to stop writing conversational text and start building a clean, production-ready document structure you can instantly use.
The Office Artifact Creator
Invite friends & get instant rewards
If you enjoy The Intelligent Edge, share it to unlock free resources when they subscribe.