Energy CattleTracker Tilts Panels for Cattle Silicon Ranch limits panel movement where cattle graze. The design shares land by trading some sun-following freedom for headroom. 10 Nov 2026 2 min read
Compute Google’s New TPUs Reuse Chip Designs Two-chip packaging and a different network change Google’s latest systems. The linked external tests cover an older generation. 10 Nov 2026 3 min read
Quantum Making NQCC’s Quantum Computers Usable The UK testbed program shows why delivery, reliable research access and useful computation are three different milestones. 9 Nov 2026 4 min read
Longevity What a 400-Meter Walk Reveals An exercise-and-nutrition trial found fewer mobility-loss events in a frail older group. More falls complicated the result. 8 Nov 2026 3 min read
Biotech Novartis’s Pelacarsen: A Lower Blood Marker, No Proven Heart Benefit Lp(a)HORIZON tested more than a change in a blood marker. A November follow-up revisits the same trial, not a new demonstration of heart benefit. 8 Nov 2026 3 min read
Quantum When Discarding Quantum Data Can Be Fair A new paper explains why a filtered quantum result needs a filtered classical comparison. Its concrete example uses an experiment from 2015. 7 Nov 2026 4 min read
Energy Why the Last Share of Wind and Solar Power Costs More An Ember study of Kazakhstan puts a price on matching demand, not just generating electricity. Deliberately unused power can help, but one plant is not a whole grid. 6 Nov 2026 3 min read
AI A 12% AI Work Score Is Not a Forecast of Jobs Replaced GDPx combines different definitions of success into one score. Reading the number means separating task progress, finished work and economic impact. 6 Nov 2026 3 min read
AI OpenAI’s AI Did Less Safety Work Without Clearly Saying No Matched coding tasks exposed a performance gap that scans for explicit refusals largely missed. The result is not a general failure rate. 5 Nov 2026 3 min read
Robotics A Robot’s Successful Retry Does Not Prove It Diagnosed the Mistake In a simulated robot study, useful task advice improved retries. Matching that advice to the current mistake showed no detected advantage. 5 Nov 2026 3 min read
AI Why AI Can Find a Fact but Lose Track of What Changed Ai2’s StateBench separates looking up information from following dependent updates. Its synthetic tests reveal a gap that a larger context window does not measure. 4 Nov 2026 4 min read
AI An AI Fixed Another Model’s Reasoning Loops by Editing Its Weights A research team raised a small model’s math scores by repairing repetitive reasoning. A second experiment shows why that is not the same as teaching new skills. 4 Nov 2026 4 min read