Introduction: Elon Musk’s Colossus—built by xAI in Memphis—is currently considered the largest AI supercomputer in the world in terms of GPU count, computing power, and rapid build-out, but it is not the largest data center globally when measured by physical size or total server capacity. Competitive megacenter projects from China, the U.S., and Europe rival or exceed it in site area and may soon overtake xAI both in scale and ambition.wikipedia+5
Introduction: The hardware for each is optimized for vastly different workloads. The core difference is that desktop GPUs (Graphics Processing Units) are designed for high frame rates in a single application (a game) on a single machine, while data center GPUs (like those in xAI’s cluster) are designed for high throughput in massively parallel computations across thousands of machines. Here’s a detailed breakdown, referencing the most relevant hardware.
Referenced GPUs
Data Center GPUs: NVIDIA’s H100 (the current workhorse of AI, used in clusters like xAI’s Grok).
Charlie Kirk’s recent death has sparked an outpouring of tributes from across the political spectrum and cultural landscape in the US and the world. The following is a sampling from prominent figures. These reflect admiration, grief, and reflections on his legacy as a conservative activist and founder of Turning Point USA.
Introduction: President Trump’s engagement with AI spans both his first and current terms, and it’s marked by a blend of deregulation, national competitiveness, and ideological framing. As of 11 Sep. 2025, here are 10 of the most memorable actions and statements. Appended is a chart of 20 key figures who currently shape Trump’s AI strategy.
Introduction: I’m republishing this description from the Yule-Cordier edition (1903, 1920) of The Travels of Marco Polo, chapters 76-78 (The Project Gutenberg Ebook 12410) as a startling example of ancient technology. Today, in 2025, it hasn’t lost any of its power to amaze. Marco Polo visited the great city of Kinsay, which is now known as Hangzhou, between 1276 and 1291, during the 17 years he spent in Kublai Khan’s empire in China. Kinsay meant “City of Heaven” at the time, which was a corrupted transliteration of the Chinese word Xingszai, which means “temporary capital.” I omitted the footnotes from the original to keep the length down, but I’d strongly urge you to click to the original copy and read the narrative with the notes. Modern technology grabs our attention, but we tend to forget that technology in our distant past can be just as fascinating. -js
“Along the River During the Qingming Festival,” a handscroll painting by the Song dynasty painter Zhang Zeduan (1085–1145).
Introduction by DeepSeek: This is a forward-looking projection based on current trends, trajectories of key figures, and the anticipated state of the AI industry in September 2025. This list balances current power (influence, resources, market position) with success (technical accomplishments, commercial impact, and research contributions). Ranking “success” is inherently subjective. This list considers a composite score of:
Power & Influence: Control over vast computational resources, capital, and talent.
Technical & Research Impact: Pioneering new architectures, algorithms, or theories.
Commercial & Real-World Impact: Deploying AI at a global scale with significant economic or societal effect.
Caveat: This response is AI-generated, for reference only. -DeepSeek
If you’re dipping your toes into the world of AI, you’ve probably heard the buzz about how it’s changing everything from chatting with your phone to generating art or even writing essays. But let’s zoom in on something specific that’s got the tech world talking: Mixture-of-Experts models, or MoE for short, and Microsoft’s shiny new entries into this space—MAI-1 and MAI-Voice-1.
Mustafa Suleyman, the CEO of Microsoft AI. Image created by Grok.
Introduction: In September 2025, here are the 20 most important facts about and 20 most successful applications of Generative AI 2.0 everyone should know. -js
Summary: Here are summaries of three of the latest (2025) articles that explain the concept of “agentic AI,” each selected for their recency, relevance, and depth of explanation. They collectively cover its definition, core concepts, practical applications, and strategic importance for businesses and society .
Here’s a rank-ordered comparison chart of 15 of the most popular AI chatbots (as of Sept 2025), focused on whether their latest-news/web search capability is available in free vs. pro tiers, along with approximate costs and what that access looks like.
Introduction: Copilot and I had this discussion earlier today. The purpose was to dive deeper into the emerging sense that our perceptions of collaboration and loyalty are changing, fueled by increasingly fluid social networks and the tsunami that’s AI. We ended where the implications might begin, to give us time to marinate the ideas we covered. -js
“CPUs typically have between eight and 64 cores optimized for performing sequential computing tasks. GPUs can host thousands of smaller cores that operate in parallel to handle massive amounts of data in a shorter time period…. They can run through vector calculations, image and video processing tasks, and data transformations all at once, which is something a CPU couldn’t match” ServerLift, 25 Aug. 2025).
Introduction: It’s a fascinating time in AI, where the big names are household fixtures but the real intrigue is happening just off the main stage. The field is so vast now that true innovation is coming from specialists who have been quietly building in a specific niche until their work suddenly becomes indispensable to everyone. Based on that, here are the most intriguing new faces making waves in September 2025. The fifteen were selected by five chatbots–Grok, DeepSeek, Copilot, ChatGPT, Perplexity–and they appear in the order they were submitted. -js & DeepSeek
Introduction: This conversation began with ChatGPT. I was curious about university libraries and the range and depth of curated resources that would be accessible via AI chatbots. I assumed the results would determine, to a large extent, the quantity and quality of research students and faculty would be able to produce. The discussion brought us to the realization that AI alone would not offset the information access advantage of top-tier institutions. The imbalance would remain. We then pursued a best-case scenario that might eventually level the playing field. At the end of the conversation, I appended a discussion with Grok about the budget disparities between top- and mid-tier universities. -js
AI is already unclogging traffic in the U.S. and abroad. Cities are using machine learning and optimization to retime signals continuously, coordinate corridors, create “green waves” for emergency vehicles, and cut stops, delay, and emissions. The broader pattern is clear: agencies are moving from fixed-time plans to continuously learning optimization, starting with high-impact corridors, then scaling citywide as data and staffing permit. Below are three standout programs, ranked for scale, maturity, and independently reported results.
Traffic in a large Chinese city. Image created by Copilot.
Free smartphone apps to converse in a foreign language are already unfolding in 2025. Several free AI-powered apps now offer real-time voice-to-voice translation across dozens (even hundreds) of languages, right from your smartphone. Here’s how close we are to seamless multilingual conversation.
An English-speaking farmer and a Swahili-speaking teacher—conversing fluidly through a pocket-sized AI interpreter. Image created by ChatBox.
Introduction: Claude and I collaborated on this short story about a time not too far in the future when AGI (Artificial General Intelligence) AI robots are beginning to replace narrow AI robots. How might this play out? -js
Image created by Copilot.
Unit-47 arrived at Pinnacle Solutions on a Tuesday morning, wheeled in on a delivery cart like a piece of office furniture. The receptionist signed for him with the same casual attention she’d give to a new printer, then pointed toward the elevator. “Forty-seventh floor,” she said, not looking up from her phone.
Introduction: This is Copilot‘s choice for the vision that’s most compelling in the development of AI. -js
In rural Mongolia, nomadic herders are using AI-enhanced satellite imagery and predictive models to anticipate weather shifts and optimize grazing routes. Image created by Grok.
Introduction: Claude and I had a conversation about the fundamental building blocks for a learning model. To sidestep the confusing backdrop of countless learning and developmental theories, we consciously avoided references to them. We ran through a few narratives to illustrate the five blocks and ended up with a short story, Algorithm of an Intentional Heart, that breathes life into this construct. -js
Educational Technology and Change Journal 
