The United States is set for one of its most significant nuclear energy expansions in decades, as three companies — Westinghouse Electric, Cameco, and Brookfield Asset Management — announced a sweeping $80 billion plan to build a new fleet of nuclear reactors across the country in partnership with the U.S. government.

The initiative marks a bold step under President Donald Trump’s “energy dominance” agenda, which emphasizes a renewed focus on oil, gas, coal, and nuclear energy.

The plan comes amid surging electricity demand in the United States, the first such sustained increase in two decades, driven largely by the explosion of artificial intelligence data centers that require massive, constant power supplies to sustain machine learning and cloud computing operations. Officials and some business leaders believe nuclear energy, long neglected and controversial, could play a vital role in powering the next wave of AI-driven digital infrastructure while maintaining low-carbon output.

Under the agreement, the U.S. government will help finance the construction of reactors using Westinghouse’s advanced nuclear technology and will also assist with permits and regulatory clearances. In return, Washington will obtain a participation interest that allows it to collect 20 percent of any cash distributions exceeding $17.5 billion once the projects are operational. However, the government’s equity interest will only take effect after a final investment decision and completion of key construction milestones.

While this partnership represents a major policy shift toward atomic power, it also revives long-standing concerns about the economics of new nuclear projects. The industry’s recent record has been marred by delays and massive cost overruns. The Vogtle nuclear expansion in Georgia, completed between 2023 and 2024, took seven years longer than expected and cost roughly $35 billion — more than double its initial $14 billion estimate. No other large-scale nuclear reactor is currently under construction in the United States, underscoring the financial and logistical hurdles the new initiative faces.

Critics have also renewed concerns about nuclear waste management, a persistent issue that remains unresolved. The U.S. still lacks a permanent repository for radioactive waste, forcing plant operators to store spent fuel on-site in cooling pools before transferring it into hardened casks. Environmental groups have noted that without a clear disposal plan, expanding the nation’s reactor fleet could pose long-term safety and storage challenges.

However, optimism is returning to the nuclear sector, largely fueled by the technological needs of the AI and cloud computing industries. Major technology companies, including Google, Microsoft, and Amazon, have begun securing long-term power contracts generated by next-generation nuclear systems, such as small modular reactors and even fusion-based facilities. These energy partnerships are seen as essential to maintaining consistent, carbon-free power for hyperscale data centers that operate around the clock.

Just this week, NextEra Energy and Google reached an agreement to restart an idle nuclear plant in Iowa, marking a symbolic shift toward nuclear’s reemergence as a cornerstone of digital-era energy planning. Similarly, Microsoft has partnered with Constellation Energy to revive one unit of the Three Mile Island nuclear station in Pennsylvania, which will directly power the company’s data centers.

The deal between Westinghouse, Cameco, and Brookfield also includes a clause that could see Westinghouse go public in the coming years. If the U.S. government’s participation interest vests and the company’s valuation exceeds $30 billion by January 2029, Westinghouse would be required to make a public share offering — a move that could reshape the global nuclear supply chain and attract institutional investors into the sector.

Cameco and Brookfield Renewable Partners acquired Westinghouse in 2023 for $7.9 billion, including debt, from Brookfield Business Partners, which had rescued the company from Toshiba’s bankruptcy in 2018 for $4.6 billion. Following Tuesday’s announcement, Cameco’s shares surged more than 15% in premarket trading, indicating strong investor confidence in the renewed U.S. nuclear drive.

The plan also aligns with President Trump’s May executive order promoting nuclear energy, which targets an additional 5 gigawatts of reactor capacity by 2030 — part of a broader push to secure energy independence and sustain industrial competitiveness amid growing AI-related power consumption.

Good People, the age of intelligent systems is here, and Tekedia is teaching how to build them. At Tekedia AI Technical Lab, we take learners deep into the fabrics of AI agent construction, from code-level architecture to system integration. You will design AI agents for customer support (WinSupport.zenvus.com), a learning companion (WinLearn.zenvus.com), career and hiring portal (WinJob.zenvus.com), and more. You will also learn how to summarize webpages, classify data, and determine web SEO and keywords, all right from your personal computer.

And you can have your own chatbot on your laptop (see video) [I used Gemma 3:1B model but you can edit the Python codes which I share with co-learners to use higher AI models and those will deliver better results]

We have made this simple and accessible. To reduce costs, we use open-source foundational models like DeepSeek, Google Gemma 3, etc, so you pay zero license fees. Yet, for those who prefer a no-code, vibecoding experience, Tekedia AI in Business Masterclass comes at no extra cost if you enroll for Tekedia AI Lab. There, you’ll create AI agents using drag-and-drop workflows, business prompts, and no programming. Tekedia offers a full-cycle path, from zero code to deep code, from beginner to builder.

You do not need a university degree; a secondary school education is enough. This is how we democratize AI capability, one learner, one builder, one dreamer at a time.

Join us at Tekedia AI Lab as the next edition begins on Nov 15. Let’s build the future, one AI agent at a time. Register here

Elon Musk has unveiled an early version of Grokipedia, an artificial intelligence–driven encyclopedia built on his xAI’s large language model, Grok.

The launch, which took place on Monday, was marked by a temporary website crash due to heavy traffic before the platform came back online later in the day.

Musk, who leads Tesla, SpaceX, and xAI, has promoted Grokipedia as a superior and less “biased” alternative to the long-established, volunteer-written Wikipedia. The new service, which Musk said is currently at “version 0.1,” is designed to generate articles using AI rather than human editors. He promised that “version 1.0 will be 10X better,” claiming that even its current version “already outperforms Wikipedia.”

The billionaire’s push into the AI encyclopedia space follows a suggestion from tech investor and U.S. presidential aide David Sacks, who now serves as President Donald Trump’s AI and crypto czar.

Musk’s creation of Grokipedia marks another extension of his ongoing campaign against what he calls “woke bias” in mainstream media and digital platforms. He has repeatedly accused Wikipedia of promoting political and cultural bias, criticizing it for relying on what he calls “left-leaning sources” such as The New York Times and NPR.

Grokipedia’s interface is minimalist — a dark-themed homepage with a search bar and a counter showing over 885,000 AI-generated articles. While it mirrors Wikipedia’s format, it is far smaller in scale compared to Wikipedia’s 7 million English-language articles maintained by millions of volunteer editors.

Yet, despite Musk’s framing of Grokipedia as a rival, early users and observers noticed that the new AI platform frequently cites Wikipedia as one of its main data sources. This paradox — promoting itself as a Wikipedia replacement while drawing from Wikipedia’s database — has fueled debate over whether Grokipedia is innovating or merely repackaging open data.

Wikimedia Responds: “Knowledge Must Remain Human”

In a statement to CNBC, a spokesperson for the Wikimedia Foundation, which operates Wikipedia, said the organization was “still assessing how Grokipedia works.” The spokesperson added that “alternative encyclopedias have appeared before without undermining Wikipedia’s mission.”

They also underscored Wikipedia’s unique advantage: “Wikipedia’s knowledge is — and always will be — human. Through open collaboration and consensus, people from all backgrounds build a neutral, living record of human understanding.”

The statement added that AI companies, including Musk’s xAI, “rely on Wikipedia’s open data to generate content; even Grokipedia needs Wikipedia to exist.”

Wikipedia co-founder Jimmy Wales echoed similar sentiments in an interview with The Washington Post, saying he did not expect Grokipedia to meet its lofty claims because “AI language models simply aren’t accurate enough yet — there will be a lot of errors.”

On the other hand, Larry Sanger, Wikipedia’s other co-founder, who left the platform in 2002 after disagreements over editorial direction, has long supported alternatives to Wikipedia. But following Grokipedia’s debut, Sanger posted a detailed thread highlighting factual errors about himself found on the AI-generated site, suggesting even skeptics of Wikipedia’s current structure are not entirely sold on Musk’s version.

A Broader Push for AI-Created Knowledge

The launch of Grokipedia comes amid Musk’s effort to weave AI more deeply into his business empire. His AI model, Grok, has already been integrated into the social media platform X (formerly Twitter), providing users with an AI assistant that he describes as “humorous, insightful, and politically neutral.”

Musk has marketed Grok — and now Grokipedia — as “anti-woke” alternatives to OpenAI’s ChatGPT and other mainstream AI products, which he claims are trained to censor controversial viewpoints. However, xAI’s reliance on existing public data sources like Wikipedia and news outlets means Grokipedia’s promise of “unbiased knowledge” may be difficult to realize in practice.

Grokipedia’s emergence is believed to signal Musk’s ambition to turn xAI into an ecosystem of autonomous information tools that could eventually rival both traditional media and AI-based knowledge systems.

The Challenge of Replacing Human Collaboration with AI

While Grokipedia’s debut grabbed attention, its long-term viability may hinge on the balance between speed and credibility. AI-generated encyclopedic content offers massive scale, potentially producing millions of entries at low cost, but lacks the human oversight and editorial debate that have helped Wikipedia build trust over two decades.

The project also deepens the growing competition between AI-powered knowledge systems — from OpenAI’s retrieval-based ChatGPT tools to Google’s search-integrated Gemini.

In the end, Grokipedia’s biggest test may not be whether it can generate millions of articles, but whether it can win the same trust that Wikipedia has earned — not through algorithms, but through the collective work of human editors.