Google fired proton beams at its AI chips to see whether they could survive five years of cosmic radiation in orbiting data centers

An analysis of satellite data from 2004 to 2024 showed that land surface temperatures in the areas surrounding AI data centers typically increase by an average of 2 degrees Celsius (3.6 degrees Fahrenheit) after they begin running. Across the 6,000 global data centers analyzed by the study, temperatures increased by as much as 9.1 degrees Celsius (16.4 degrees Fahrenheit).

Big Tech has rebranded nuclear power as a green solution to address the strain on the grid from millions of people using power-hungry AI tools. As of mid-2025, generating one image uses as much electricity as charging the average smartphone, or leaving a household light bulb on for 87 consecutive days.

Google used its 53-qubit processor on an extremely difficult problem for traditional computers to simulate to showcase quantum supremacy. IBM later challenged Google’s claim by arguing that its Summit supercomputer could solve the same problem in just 2.5 days using a more efficient method.

The project seeks to reduce errors within its quantum computer and integrate 1 million physical qubits (or quantum bits) into the room-sized computer. This powerful device will attempt to solve complex problems in medicine, computing, and more.

For years, Google was the cutting-edge of the AI world, including the development of the Transformer—the basis of contemporary AI chatbots. But since OpenAI debuted ChatGPT in 2022, the company was had to play catch-up. This is a four-hour deep dive into Google's AI history.

Several safety mechanisms failed during Anatoli Bugorski's inspection of malfunctioning equipment at the U-70 synchrotron, the Soviet Union's largest particle accelerator. Despite receiving as much as 600 times the lethal dose of radiation to the brain, he has shown no intellectual damage and worked until the age of 77.

Analysts believe that generative AI tools—not antitrust rulings—are Google's biggest threat because they draw users away from traditional search and potential ad revenue. AI partnerships, like the one between Apple and OpenAI, may further reduce Google's market share.

Its semiconductor chips were used in the automotive industry, for data centers, and by crypto miners. And starting in 2006, Nvidia took steps that would ultimately transform it into the AI company it's known as today. This podcast is the second part of a three-part series diving deep into Nvidia's company evolution.

Property boundaries, such as anti-climb fencing, overlapping cameras, vehicle crash barriers, and circulating patrols, are coupled with lobby iris scans and ID and badge checks. The security operations center monitors individuals throughout their visit, including the less than 1% of Google users who ever receive permission to visit the data center floor.

Workers, particularly in marketing, are experiencing mental fatigue in the workplace due to expectations—often unspoken—that they should use AI to multitask beyond their comfort levels. Many feel they are underperforming if they don't embrace this technology. This situation differs from typical burnout. This podcast discussion suggests that this "brain fry" tends to intensify when employees are required to use more than three AI tools.

Most AI systems are so-called "black box" models, where we only see inputs and outputs while the inner workings remain obscure. This piece, written prior to the launch of ChatGPT, cautions developers to know why and how a system produces its results, and not to sacrifice explainability for accuracy.

DeepSeek unveiled an AI model in 2025 that it said was made in two months for roughly $6M. For context, OpenAI has suggested it took over $100M to make its AI model GPT-4. The news rattled the stocks of American tech companies, including leading chipmaker Nvidia.

Surface hardening and embedding specific elements or coatings can improve the biocompatibility and bone tissue integration of these implants. Particle accelerators are significant sources of the radioisotopes needed in medical diagnostics, and microwave linear accelerators are used in the majority of cancer irradiation treatments.

Unlike regular computers, where each bit is either 0 or 1, quantum computers use qubits, which can simultaneously be both—allowing for massive parallel processing power. Google’s quantum processor once solved a problem in seconds that would have taken a supercomputer 47 years.

An MIT study finds lower brain engagement in students who used ChatGPT, resulting in consistent underperformance in neural, linguistic and behavioral levels. AI users struggled to recall their own essays, while the brain-only group showed the highest neural connectivity and satisfaction.

Unveiled in 2023, the GH200 Grace Hopper Superchip, designed for generative AI applications in data center-scale systems, combines a Grace CPU with a Hopper GPU to create a coherent memory model. This allows the Hopper to access Grace data directly, reducing the need to repeatedly copy data back and forth and vastly improving performance.

Launched in 1998, Google revolutionized search engines with its clean design and PageRank algorithm. As of 2025, Google has about 90% of the search market share but faces new challenges from AI-powered competitors and regulatory scrutiny.

Although ChatGPT, Google's AI overviews, and other similar software can save time, a study of more than 10,000 adults showed that reliance on them yielded work products that were more generic and included fewer facts than those produced solely by Google search. Weaker brain connectivity has also been observed when users write using AI.

Moore’s law, named after Intel’s cofounder, predicts that the number of transistors on a chip will double every two years. But Nvidia could be outpacing that prediction.

Though the particle—which is its own antiparticle—remains theoretical, scientists from Microsoft to Google continue developing systems to mimic this property towards building error-resistant quantum computers. Scientists at UCSB have been attempting to get the particle using just pencil lead and Scotch tape.

An interview with Alphabet CEO Sundar Pichai warns that the trillion-dollar surge in AI valuations may contain clear signs of irrationality, echoing past tech bubbles. He says no firm—including Google—would be immune if the boom cools or bursts.

Through anti-Stokes cooling, incoming energy is combined with vibrations in a material's crystal lattice to produce a net reduction in its temperature. Using this effect, lasers can be directed at microchip hot spots to help keep data centers cool.

According to a report from the International Energy Agency, a typical AI data center in 2025 consumes as much electricity as 100,000 households, but the largest under construction are expected to consume 20 times as much. Global data center electricity consumption is expected to surpass Japan's total consumption by 2030.

Traditional quantum computers are incredibly fragile, with encoded information easily disrupted and requiring significant error correction systems. Microsoft’s approach involves storing data on a surface, which is more stable, but involves an elusive particle first theorized in 1938.

Scientists exposed meteorite samples to proton irradiation at CERN to mimic the effects of a nuclear detonation. The material exhibited greater resistance to breaking after the exposure, suggesting nuclear weapons may better serve to deflect asteroids, not break them.

Punch card computers from the late 1940s were used to store and process a company's data locally before the advent of desktop computers. As the amount of data and company sizes increased alongside the miniaturization of technology, mainframes were replaced by local server rooms and external facilities with interconnected servers running cloud-based and AI software.

Australian start-up Cortical Labs unveiled CL1, a hybrid chip designed to help scientists study brain function in real-time. The 800,000 lab-grown neurons on the chip are kept viable for six months via a built-in life support system.

The development of transformer architecture and the release of GPTs across AI labs led to rapid changes in LLM scale and power. Companies like OpenAI, Google, Meta, Microsoft, and Anthropic have invested billions of dollars to develop reliable, versatile models.

With the New START treaty expired, researchers at the Federation of American Scientists have proposed using satellites and AI to remotely verify nuclear arsenals—replacing the onsite inspections that once built trust between the US and Russia during the Cold War.

When a can of ground beef and pork spoiled despite attempts to sterilize it with gamma radiation, researcher Arthur Anderson isolated the bacteria in the food and discovered Deinococcus radiodurans. The polyextremophile has since been found on the walls of nuclear reactors and overcomes radiation damage through highly optimized DNA repair mechanisms.

Date from 2023 suggests that data centers accounted for around 4.4% of US electricity consumption, with new facilities supporting the AI boom consuming more power than the state of New Hampshire annually. However, the energy impact of individual use varies widely depending on the prompt and AI model.

In IBM’s quantum lab, quantum processors are kept at a temperature of 15 millikelvin using dilution refrigerators. These refrigerators use nested thermal layers, cryogens like liquid nitrogen and helium, and powerful compressors to maintain the ultra-cold temperatures required for qubit stability and performance.

Outside the protective atmosphere of Earth, astronauts are constantly bombarded by radiation, which, among other dangers, increases the risk cancer. While those of us on Earth can mitigate the effects of similar radiation by limiting exposure,...

Every second, lasers strike 50,000 droplets moving at 250 kilometers (155 miles) per hour, vaporizing the tin in supernova-like miniature explosions that produce ultraviolet light. Knowing this light will spread via diffraction as it passes through a filter during manufacturing, designers reverse-engineer the filter design based on the required final microchip pattern.