This Week’s Awesome Tech Stories From Around the Web (Through November 14) Every week, we scour the web for im
A drill reaches down into hot rock, while far above it, data centers hum through the computational demands of artificial intelligence. In a roundup of technology stories through Nov. 14, one of the most consequential projects is a geothermal plant that Fervo Energy is developing for Google, designed to deliver 396 megawatts by the end of 2028. The company has connected its first 100-megawatt plant to the grid, according to the roundup. [singularityhub]
The stories in this week’s collection are less a parade of gadgets than a tour of technology pressing against practical limits: electricity, materials, human communication and trust. Together, they show why innovation is not simply a matter of making something work. It is also about whether it can be built affordably, scaled responsibly and used to solve the problem it promises to address.
Power beneath our feet
AI’s appetite for electricity is helping make the search for new power sources urgent. Geothermal energy draws on heat beneath Earth’s surface, offering a potential source of electricity that does not depend on sunshine or wind. Fervo’s project is intended to supply a large amount of power to Google, whose data centers are part of the broader expansion driving demand. [singularityhub]
The milestone is real, but the promise still has to be demonstrated at scale. A 100-megawatt connection is not the same as completing a 396-megawatt plant, and the project’s stated timeline reaches to the end of 2028. The wider question is whether geothermal systems can be built reliably and quickly enough to meet the rising needs of data centers, without treating a target as an accomplished fact.
Small tests, big materials
A different kind of infrastructure story begins with cement particles under a microscope. Engineers at the University of Illinois Urbana-Champaign have developed a test that uses a microscope-based framework costing about $500 to measure cement particle characteristics. The aim is to help the construction industry save money. [techxplore]
Cement is an everyday material with a complicated job: its properties affect how concrete behaves, and testing those properties can influence cost and construction decisions. A comparatively inexpensive measurement tool could make useful analysis more accessible. But a low equipment price alone does not establish how widely the method will be adopted or how much it will save; those outcomes depend on performance in real-world use.
The roundup also points to work on silicon carbide, a material used in power electronics. Silicon carbide is harder and more expensive to process than ordinary silicon, making the way it is manufactured consequential. Infineon demonstrated a technique called Cold Split, which separates wafers from a silicon-carbide crystal with less material loss than conventional methods, according to a technology digest. [guruchahal]
That is a manufacturing improvement, not a magic shortcut around the material’s challenges. Reducing waste could matter when a material is costly to process, but the account does not establish how much the technique will change production costs or how broadly it will be used. The important idea is more modest and more tangible: sometimes efficiency comes not from inventing a new material, but from losing less of the one already in hand.
Communication and control
For people living with ALS, Parkinson’s disease or brain injuries, technology may offer something more personal than cheaper construction or cleaner power: a route back to communication. Science News describes technology that could help people with these conditions “talking” again. The brief item does not provide enough detail to assess the system or how far it has progressed, but the goal points to a major frontier in assistive technology: translating signals from the brain into a form other people can understand. [sciencenews]
That possibility comes with a distinction worth keeping clear. A technology described as capable of helping people communicate is not necessarily a widely available treatment, and the short summary does not say how many people have used it or how well it works. For a reader, the promise is vivid; for science, the next questions are about reliability, access and whether the system can serve people with very different needs.
Other stories in the roundup examine technology’s influence on judgment and privacy. One digest reports that Meta is rolling out AI age-estimation technology on Instagram, using signals such as posts, photos and account activity to identify suspected teenagers and place them in protected Teen Accounts, even if they entered an adult birthday. [benjaminbennettalexander]
The effort illustrates a persistent tension: platforms may use automated systems to protect young users, but those systems make inferences about people from their online behavior. The account describes the rollout, not its accuracy or the consequences of mistaken classifications. Those unanswered questions matter because an age estimate is not merely a technical output; it can change what a user is allowed to see or do.
The answers AI gives
Technology also raises questions about what people value in knowledge itself. Science News reports on a DeepMind study prompted by AI-generated mathematical proofs, and asks what may be lost when attention focuses only on answers. [sciencenews]
A proof is not just a result; it is a chain of reasoning that can reveal why a claim is true. If AI systems help generate proofs, they could expand what mathematicians can explore. But the source’s framing leaves open a deeper issue: whether a proof that is difficult for people to follow can still deepen human understanding, and how researchers should balance speed with explanation.
The week’s most striking image may be the simplest: heat rising from deep underground to meet the electricity demands of machines above it. Between those depths lie the questions that make technology worth watching—not only what it can do, but what it will cost, whom it will serve and what we still need to understand.
Sources
- This Week’s Awesome Tech Stories From Around the Web (Through November 14)
- Top news of the week - Science News, Physics News, Technology News
- Weekly Technology Digest — November 14, 2026
- GeekWire – Breaking News in Technology & Business
- Scientific American
- 10 Big Tech Stories This Week (Nov 3–10, 2026)
- New Atlas: Technology-Innovation-Outdoor News
- TechLatest AI & Tech Weekly #42 - Techlatest.net
- This Week’s Awesome Tech Stories From Around the Web (Through November 14)
- Tech | Science News
- AI in Bio Weekly Review: Schrödinger's $2.45B Novartis Pact, BioNTech's $950M Biotheus Buyout, AlphaFold 3 Code Release, and Boltz-1 Open-Source Breakthrough — November 7–13, 2026 | Data Analysis | OpenBio News
- Scientists Develop Injectable Glucose Sensor That Could Change Diabetes Care