Welcome back! The center of power in AI is shifting from models to machinery. Governments are digging up old lab data for state-run moonshots. Tech giants are planting company towns to feed their hunger for electricity. Startups are bracing as Google’s scale threatens to reshape the economics of the entire sector. Regulators are warning that the industry’s momentum now rivals geopolitical forces. The new AI frontier is territory, influence, and control over the physical systems.
In today’s Generative AI Newsletter:
• Genesis Mission turns hidden lab data into fuel for state-run AI
• Amazon builds a new AI company town with $65B to $900 data center footprint
• OpenAI faces “rough vibes” as Google gains momentum
Latest Developments
On November 24 in Washington, Trump signed something that sounds like a sci-fi movie pitch: the “Genesis Mission.” Its official goal is to use AI for scientific breakthroughs. A lot of hidden lab data from decades ago is being used by the federal government to train an elaborated AI system. Whoever runs Genesis will have a say in the future of energy, weapons and maybe even climate technology.
What Genesis is actually doing:
Mandate: Energy Department develops “Genesis Mission” to solve AI concerns nationwide.
Machine: Supercomputers, quantum equipment, and robotics labs are all linked by a new platform called "American Science and Security."
Upside: For clean energy, medicine, and smart chips, supporters want to build the Manhattan Project.
Risk: Critics worry about privacy, using a lot of power, corporate control, and AI in defense.
Genesis is not the first science story linked to patriotic branding. It's similar to the Human Genome Project and other "moonshots," but this time it's run by security agencies and private AI companies. Within the next ten years, AI may be less focused on chatbots and more on who controls the data centers that run trials. Genesis can either lead to new discoveries or give scientists a powerful place to work alone.
Special highlight from our network
How teams are using AI to cut costs and be productive
Pecan AI makes predictive analytics actually accessible. Business teams build trust-worthy models in hours. No data scientist required. Pecan handles the messy parts: data prep, feature engineering, and optimization, while predictions flow straight into your CRM or marketing tools. Predict churn, forecast demand, score leads, ROAS and more. From eCommerce to fintech, teams are cutting costs and driving real impact with real ROI in weeks.
Special highlight from our network
Transform Enterprise Knowledge into Trusted, Verifiable Answers
Most teams still rely on search.
The next step is turning scattered enterprise knowledge into instant, trustworthy answers.
Progress Agentic RAG moves you from “here’s a document” to “here’s the next best action.”
With traceable sources, governed data, and modular RAG pipelines that you can scale.
Join us on Nov 27 at 12:00 PM CET to see:
What Progress Agentic RAG is and where it fits
Why LLMs alone cannot fix data fragmentation or governance
How to structure enterprise data for quality, traceability and auditability
Want your AI to give trusted, verifiable answers instead of guesswork?
Register for the live session here
Amazon made two major announcements on November 24 that will significantly impact the AI landscape. It plans to invest $15 billion into new data centers in northern Indiana and up to $50 billion into AI supercomputing that will benefit U.S. government departments. The media focuses on "jobs," "innovation," and "national security" but if you look more closely, you'll notice a different picture. The project sits on former farmland requiring plants and power lines, based on a deal where Amazon uses 2.4 gigawatts of electricity.
Key details of the deal:
Power build: Amazon is investing in new power lines and plants for the required power.
Bills and grid: The power company says that customers will save money while it changes the grid to fit this load.
Local benefits: The deal includes 1,100 skilled jobs and STEM training places.
Pattern elsewhere: This is a lot like how chip plants have grown in Ohio and data centers have grown in Virginia.
All of these things make Indiana look like a new "AI company town" experiment. It is easy to see the benefits like a tax base, construction, technical job opportunities and a blooming spot in the upcoming era of computing. These same data centers can also strain local power and water, push up demand on the grid and turn large pieces of land into locked off industrial sites that residents cannot really use. As a result of this deal, Amazon will be in charge of the computing resources for the local companies and the government. The true verdict on this "massive win" will come years later, when contracts are renewed and residents realize what they truly pay and how much influence they have.
A leaked internal company memo shows Sam Altman warning OpenAI employees that Google’s recent progress could create “temporary economic headwinds” for the company. The note was circulated just before Gemini 3 launched, and it describes a period where morale, revenue expectations, and competitive footing may tighten. Altman also referenced a new model in development but made clear that OpenAI “has work to do” as Google gains speed across hardware, distribution, and product reach.
What was in the memo?
Google’s advantage: Altman cited Google’s control of Search, YouTube, Gmail, Maps, Android, and its own TPU chips, calling the combination difficult to match.
Economic warning: He told staff that Google’s momentum could slow OpenAI’s revenue growth and create short-term strain across the business.
Internal admission: Altman wrote that researchers had seen Google build a model that “appears to have leapfrogged OpenAI’s in the way it was developed.”
Shrinking lead: The memo noted that OpenAI’s edge over Google and Anthropic “has narrowed,” even as ChatGPT still leads in usage and revenue.
Altman's message comes at a time when the AI economy is shaped more by distribution, talent, and infrastructure than by model performance alone. Google controls its own chips, its global user base, a vast financial cushion, and the platforms that feed its models. OpenAI carries high expectations, a steep burn rate, and an investor base that assumes long arcs of uninterrupted growth. The rumored upcoming model might prove to be a test of whether OpenAI can bend the economics of this industry back in its favor.
Claude Opus 4.5 is Anthropic’s new flagship model designed for users who work with long documents, evolving projects, and multi-step tasks. It introduces a stronger memory system that decides what to keep and what to compress, so conversations and coding sessions can continue without hitting a context wall. The model handles ambiguity well and feels steady when exploring large codebases, redesigning interfaces, or reviewing complex material.
Core functions:
Endless memory: Compresses older context to keep long chats and long tasks running.
Engineering strength: Fixes cross file bugs, writes tests, and handles large codebases.
Frontend and UI: Generates layouts, component trees, and style implementations.
Computer use: Operates across Chrome and Excel for tab work, data cleaning, formulas, and charts.
Multi agent work: Directs subagents for research, exploration, and document tasks.`
Try this yourself:
Give Opus 4.5 a task you encounter in real life. For example, upload a long design spec and ask it to: extract the requirements, sketch a UI flow, generate components, and create a working draft. Or load a multi file bug and ask it to diagnose the issue, propose a fix, write tests, and summarize the change. Watch how it preserves important details and keeps the thread intact over time.
ToneUp has been one of the engines powering GenAI Works’ growth.
It gives creators, founders and teams one place to plan, write, design and refine content with a steady, recognisable voice.
Idea generation, long-form drafts, social-ready posts, voice alignment, images and cross-platform repurposing, everything flows in a single system.
A system like this rarely emerges early in a category.
Reg CF offering via DealMaker Securities. Investing is risky.
