Botwin's Morning Wire

Nearly all Americans now view corruption as a major problem in government, with recent polls showing 88 to 92 percent expressing extreme or very high concern. A Gallup survey found 89 percent believe corruption is widespread, while a Brennan Center poll reported 92 percent of respondents across party lines identifying it as a serious issue. In a Midwest focus group that included voters who supported Donald Trump in 2024, participants linked high fuel costs to "corrupt politicians" and "billionaires," a reaction that surprised pollsters. The same research indicated that 79 percent of respondents feel corruption blocks solutions to everyday problems such as inflation and healthcare. Trust in Congress remains at historic lows, and public confidence in the government’s ability to address major challenges has hit an all-time low. While political rhetoric has long framed corruption as a voter grievance, the current data suggest a shift: concerns about graft are no longer peripheral but are increasingly tied to the issues that dominate voters’ daily lives. This growing linkage may influence how voters evaluate candidates, particularly incumbents, in the upcoming election.

A demonstration of error-corrected quantum bits shows error rates decline significantly as more hardware qubits are incorporated into logical units. Quantinuum reported that its Helios processor can scale logical qubit performance by factors exceeding ten with each incremental increase in size. The results indicate error correction scaling that surpasses previous demonstrations, though practical implementation remains constrained by hardware limits.

Logical qubits, formed by entangling multiple physical qubits, require error correction to maintain stability during computations. The effectiveness of such systems depends on how error rates respond to increases in code distance, a measure of a code's ability to detect and correct errors. When error rates decrease with each step up in code distance, a processor is considered to have passed the breakeven threshold. Quantinuum's latest findings suggest its hardware achieves this condition with measurable efficiency gains.

The company utilized a trapped ion architecture, where qubits are stored in charged atoms manipulated via laser-controlled electric fields. Helios employs a modular design that enables entanglement between any pair of ions by routing them through separate processing zones. This architecture supports complex error correction codes, including color codes with varying distances. In testing, logical qubits encoded with code distances of three, five, and seven were constructed using 13, 37, and 73 physical qubits respectively.

Error rates were assessed through repeated cycles of syndrome extraction and correction. Each cycle involved measuring ancillary qubits to detect anomalies without disturbing the encoded data. The logical error rate dropped from approximately 5 x 10^-4 at code distance three to 4 x 10^-5 at distance five and 5 x 10^-6 at distance seven. Each progression corresponded to roughly a twelvefold improvement in error suppression, marking some of the most consistent scaling observed to date.

A key advancement involved mitigating leakage errors, a source of instability where qubits occupy unintended energy states. Quantinuum introduced a laser-based correction method that restored leaked qubits to the computational manifold, reducing this error mode by a factor of 6.5. Additionally, a dynamic qubit recycling protocol allowed failed measurement qubits to be reset and reused within the same cycle, improving resource efficiency.

Syndrome interpretation, a computationally intensive step in error correction, was handled by three real-time processing systems: a conventional algorithm, a convolutional neural network, and a transformer-based model. These enabled rapid analysis of error patterns across multiple correction rounds. Twenty consecutive cycles were executed before final measurement to ensure reliable assessment of memory retention.

When applied to logic gate operations, the benefits of error correction became more nuanced. Direct hardware operations yielded error rates near 10^-3, while encoded operations using distance five codes reduced errors to about 7 x 10^-5. This suggests that error correction provides meaningful gains even during active computation, though performance gains were not uniform across all code distances.

The results present a mixed outlook. On one hand, the observed scaling indicates that modest improvements in hardware fidelity could yield substantial reductions in logical error rates. On the other, current limitations in qubit connectivity and code capacity restrict the scope of testable operations. The system currently supports only a single distance seven code at a time, and full algorithmic workloads remain out of reach.

Quantinuum's approach differs from some competitors by prioritizing scalable error correction architecture over aggressive reductions in physical error rates. If the company can expand qubit counts as projected, it may achieve functional error-corrected systems without requiring orders of magnitude improvements in base hardware quality. This strategy could shift the development trajectory toward larger, more modular error correction implementations.

The demonstration represents a technical milestone in quantum error management, showing that error correction can not only mitigate errors but also improve predictably with scale. While significant engineering challenges remain, the results offer a clear path toward more stable quantum computations, contingent on continued progress in qubit control and system integration.

NASA announced that commercial lunar terrain vehicles and robotic rovers will be essential for early development of the planned Moon Base, supporting tasks such as technology testing, astronaut transport, terrain mapping and resource prospecting. Data gathered from these early missions will shape future mobility systems and infrastructure needed for long‑term lunar habitation.

On Tuesday, the agency released a video update that highlighted recent advances in the commercial vehicles designed to provide surface mobility for the outpost.

"Lunar rovers are central to our surface mobility strategy," said Robert Pickle, director of the Surface Mobility Office within NASA’s Moon Base Program. "Their performance will lay the groundwork for the mobility systems that will define the next era of exploration and ultimately determine how effectively we can build, operate, and expand the infrastructure required for long‑duration exploration."

The agency also took the opportunity to promote its VIPER rover, which is scheduled to launch toward the lunar South Pole in 2027. The Volatiles Investigating Polar Exploration Rover will travel more than 18 miles (30 kilometers) across permanently shadowed craters to search for water and other resources that could support sustainable exploration. After being delivered by a Blue Moon Mark 1 lander, VIPER will operate from NASA’s Ames Research Center, where teams are simulating potential mission scenarios.

VIPER is currently housed at NASA’s Johnson Space Center in Houston, where it is undergoing final testing before being shipped to Kennedy Space Center for launch. The rover’s mission aims to gather data on temperature, lighting and terrain conditions in the South Pole region, helping scientists understand the origin and distribution of lunar water and how those resources might be used for the Moon Base.

In parallel, NASA awarded task orders to two companies — Astrolab and Lunar Outpost — to develop the first phase of crewed lunar terrain vehicles (LTVs). These vehicles, part of the Commercial Lunar Payload Services initiative, are slated for delivery to the lunar South Pole by 2028. The LTVs will be remotely operated from Earth, capable of traveling up to 3.7 miles (6 kilometers) per day and covering a total distance of six miles (10 kilometers) on a single battery charge. Each unit can support three moonwalks per Artemis mission, carry up to 176 pounds (80 kilograms) of samples and tools, and will be used by Artemis astronauts to explore geologically diverse areas of the South Pole.

Astrolab’s Crewed Lunar Vehicle 1 (CLV‑1) is based on the company’s FLEX architecture. A mockup of CLV‑1 was recently delivered to Axiom Space, where NASA astronauts evaluated the rover’s design, cabin layout and display interfaces while wearing Artemis spacesuits. The company plans to send a robotic precursor, FLIP (FLEX Lunar Innovation Platform), on a Voyager Technologies Griffin Mission One launch slated for late 2026. FLIP will study the South Pole environment and test subsystems that will be incorporated into CLV‑1, including tires, batteries, radios and cameras. Engineers from Astrolab will soon test the rover’s airless wheels in NASA’s Glenn Research Center’s Simulated Lunar Operations laboratory.

Lunar Outpost is developing the Pegasus crewed LTV, a lighter evolution of its Eagle LTV. Prior to the award, the company built several mockups that were delivered to Johnson Space Center for crewed testing. In 2025, its robotic rover MAPP (Mobile Autonomous Prospecting Platform) operated at the South Pole, serving as a testbed for thermal control, navigation cameras and mission‑control software that will be used to remotely operate Pegasus. A second MAPP rover is scheduled for launch in early 2027. Teams from Lunar Outpost will also test Pegasus’ spring‑mesh tires at Glenn’s facilities to assess soil mechanics and traction.

Both Astrolab and Lunar Outpost are on track to complete preliminary design reviews by the end of the year, with flight‑ready hardware expected to be delivered to NASA by late 2027. These developments are part of NASA’s broader Artemis program, which aims to send astronauts on increasingly complex missions near the lunar South Pole, establish a sustained human presence on the Moon and lay groundwork for future crewed missions to Mars.

For additional details about the Moon Base initiative, NASA’s website provides a dedicated resource page.