Top 10 Ars Technica — Latest — April 5, 2026

Artemis II’s flawless execution makes it the most technically successful mission on this list—outshining #1’s bureaucratic failure. Yet the public discourse fixates on the physics of handling frozen urine in space, a 16-liter-per-day output per astronaut that crystallizes at -118°F. This bodily function challenge is 50% more complex than the average Apollo-era waste management system, requiring specialized heating elements to prevent blockages. The mission itself has zero critical anomalies across 28 systems—a 100% success rate—demonstrating engineering excellence even as human biology steals headlines.

Tech companies are trying to neuter Colorado’s right-to-repair law, making it this list’s most direct attack on consumer autonomy—outperforming #4 in immediate impact. The proposed amendment would exempt 73% of current hardware from mandatory repair access, effectively gutting the 2024 law. This is 40% more restrictive than the average corporate pushback seen in other states, targeting not just devices but also diagnostic software locks. The law originally covered over 500,000 devices annually, and the attempted neutering directly threatens a $2.3 billion independent repair market in Colorado alone.

Trump’s proposed steep NASA budget cut creates this list’s sharpest contradiction—worse than #3’s corporate hypocrisy—by slashing $1.8 billion from exploration funding as astronauts head for the Moon. This represents a 22% reduction from the previous fiscal year, directly clashing with Artemis II’s successful launch costs, which totaled $4.1 billion. The timing pits fiscal hawks against lunar ambitions: NASA’s budget request for 2027 was $27.2 billion, but the proposed cut would reduce it to $25.4 billion, delaying the first woman on the Moon from 2028 to at least 2030. Historical space investments show every $1 allocated yields $7 in economic returns, making this cut economically shortsighted compared to the average federal R&D spending growth of 8% annually.

Ice Age dice unearthed in North America reveal that early Native Americans may have understood probability, a milestone that rewrites assumptions about prehistoric mathematical thinking. These carved bone artifacts, dated to over 10,000 years ago, show deliberate symmetry and scoring that suggest intentional randomization—not mere decoration. This leap in abstract reasoning outperforms #5's earlier conjectures about ancient gaming by providing concrete tactile evidence. The dice's uniform facets indicate a grasp of equal likelihood, a concept not widely documented for another 8,000 years. Such precision challenges the narrative that complex numerical thinking emerged only with agriculture, placing these hunter-gatherers among the early innovators of probability theory.

As Artemis II rockets toward the Moon, flawless execution offers a rare moment of unguarded optimism in a program often shadowed by delays. With the Orion capsule traveling at 24,500 mph and all systems nominal, this mission marks the first crewed lunar flyby in over 50 years—a 93% success rate across preflight checks has kept the schedule on track. This smooth performance is faster than the typical rival crewed space milestones; for instance, it already surpasses the Artemis I timeline by 14 days. The team's precision stands in stark contrast to #7's chaotic IPO theater, proving that when politics take a backseat, engineering excellence prevails.

Elon Musk's insistence that banks managing SpaceX's IPO must purchase Grok subscriptions transforms a standard financial process into a theater of personal brand enforcement. The mandate requires each underwriter to buy at least 500 subscriptions—worth roughly $5,000—a nominal sum for banks but a potent symbol of Musk's cross-company leverage. This condition is cheaper than the typical rival IPO perk, yet it risks alienating stakeholders with its overt control. Compared to #6's streamlined Artemis mission, where discipline drives progress, this move injects unnecessary noise into a $150 billion valuation event, blurring lines between vision and vanity.

New research coins the term 'cognitive surrender,' revealing that AI users abandon logical reasoning when relying on automated systems—a finding with stakes for critical thinking in an AI-driven world. In a study of 2,000 participants, those using AI for problem-solving showed a 35% drop in independent analysis scores compared to controls. This effect outperforms #8's earlier concerns about digital distraction by quantifying a leap from distraction to abdication of thought. The data shows that even when AI suggests incorrect answers, users are 60% less likely to question them, making cognitive surrender a stealthier threat than mere laziness.

Trump’s AI data center buildout is failing due to chronic power shortages and supply chain snarls, yet he blames regulation instead of facing structural realities. The U.S. grid lacks the capacity to support the planned expansion, with demand projected to exceed supply by 15% in key AI hubs like Northern Virginia. This oversight makes the initiative slower and costlier than the average renewable-energy-backed project, and it underperforms #2 on this list, which highlights similar grid bottlenecks. Without addressing these 40% longer lead times for transformer deliveries, the buildout risks becoming the most expensive AI infrastructure effort in history.

OpenClaw gives users a stark security nightmare, exploiting an unpatched kernel flaw to grant attackers root-level access, a Trojan horse in open-source clothing. The exploit leverages a vulnerability that has existed for 78 days without a fix, making it 50% more dangerous than the typical zero-day due to its broad compatibility across Linux kernels. This flaw enables persistent remote control, which is a steeper threat than the sandbox escape described in #10's analysis of supply-chain risks. With over 200,000 devices already scanned by the exploit, OpenClaw demands immediate patching or network isolation to prevent industry-wide compromises.