Science

Current Picks

Behavioural Economics Took the Field Experiment From Development Studies

By Karim Osman / Sep 18, 2026

Randomised field trials moved from development economics into behavioural policy. Here is how the method travelled, what it changed, and how to read trial claims.

Three Photon Lab Codes Forced One Optogenetics Lab to Switch Laser Wavelengths

By Renu Shah / Jul 17, 2026

A neuroscience lab abandoned two-photon imaging for a 1320 nm three-photon laser to reach hippocampal CA1 neurons. The switch required red-shifted opsins like ChRmine, revealing trade-offs between depth and temporal precision.

Software Heritage Outlasts University Repositories as Code Citations Diverge

By Renu Shah / Sep 18, 2026

University repositories archive papers, not code. Software Heritage issues persistent identifiers for source code, and citation practice is slowly splitting from paper DOIs.

Dictator Game Splits Shrink Sixfold When the Recipient Can Retaliate

By Karim Osman / Sep 18, 2026

A new experiment finds dictator game offers fall from about 25% to near 4% when recipients can retaliate, challenging purely altruistic explanations.

Barcode of 2,000 Stars Per Sky Frame Trained One Spiral Model and Rejected Another

By Renu Shah / Jul 20, 2026

Gaia's astrometry of 2,000 stars per square degree trained a density wave model of Milky Way spiral arms and rejected tidal interaction at 12 sigma confidence.

A 1986 Primate Lab's Per Diem Rate Outlived Its Neuroscience Findings

By Alice Chen / Sep 18, 2026

A 1986 primate lab per diem still shapes neuroscience grant budgets, even though the findings it supported never replicated. Here's how to read the mismatch.

Reproducibility Checks Stall Where Code Repositories Stop Accepting Deposits

By Karim Osman / Sep 18, 2026

Reproducibility checks stall when code repositories stop accepting deposits. A methodology breakdown of the procedural choices that produce unverifiable results.

The Bell Test's Coincidence Monitor Was a Nuclear Physicist's Spare Parts Build

By Karim Osman / Sep 18, 2026

A nuclear physicist built the coincidence monitor central to early Bell tests from surplus spectroscopy gear, and the undocumented build choices still complicate replication.

When a Fortran Solver’s Unit Tests Outlive Its Physics

By Jonas Eriksen / Sep 18, 2026

Legacy Fortran solvers often pass every unit test while their physics has quietly drifted. How test oracles fossilize, and how to audit them.

Giving Up the Last Octave: How Submillimeter Survey Bands Trade Sensitivity for Redshift Reach

By Jonas Eriksen / Sep 18, 2026

Submillimeter surveys trade sensitivity for redshift reach by choosing longer bands. How observatories, detectors, and survey strategy collide in practice.

Single-Molecule Force Clamps Reveal How Chaperonins Fold One Protein at a Time

By Jonas Eriksen / Sep 18, 2026

Force clamps pull one protein at a time through a chaperonin barrel, exposing folding trajectories that bulk assays average away. A methodology breakdown.

Beamtime Fees Now Outweigh Synthesis Costs in Battery Catalyst Papers

By Renu Shah / Sep 18, 2026

Synchrotron beamtime fees now rival or exceed synthesis costs in battery catalyst papers, reshaping funding, authorship, and what evidence gets published.

A Survey Pipeline Built for Galaxies Ends Up Pricing Exoplanet Candidates

By Renu Shah / Sep 18, 2026

How probabilistic tools built for galaxy surveys crossed into exoplanet candidate vetting, reshaping follow-up economics and raising new reproducibility questions.

Cryo-EM Time Now Costs More Than the Protein Prep It Images

By Karim Osman / Sep 18, 2026

Cryo-EM beam time now costs more than the protein prep it images. A look at how infrastructure economics shape which structures get solved and published.

Mapping How Place Cells Fire Taught Neuroscientists to Ignore the Room

By Alice Chen / Sep 18, 2026

How single-neuron recordings in rodents turned space into the hippocampus's main variable, and why the room itself became background noise in the process.

A Two-Million-Dollar Synchrotron Beamline Sits Idle While One Grant Cycle Reallocates Its Operating Budget

By Karim Osman / Jul 31, 2026

A two-million-dollar synchrotron beamline sits idle as grant cycles reallocate its operating budget. This article explores the funding misalignment, the chemistry left undone, and potential fixes.

A 1973 Rat Colony Kept One Memory Paper Alive Through Four Funding Droughts

By Alice Chen / Sep 18, 2026

How a 1973 rat colony and one technician's breeding logs kept a memory-consolidation paper alive through four funding droughts, and what that means for reading survivor papers.

Deep-Sea Mooring Fees Outlast the Paleoclimate Record They Were Built to Recover

By Alice Chen / Sep 18, 2026

A paleoclimate finding took fifteen years to settle, not because the science failed but because mooring fees, ship-time costs and grant cycles outlasted the data itself.

A Coronagraph's Deformable Mirror Now Resolves Exozodiacal Dust at Four Wavelengths

By Jonas Eriksen / Sep 18, 2026

A deformable-mirror coronagraph now senses exozodiacal dust in four wavelength bands at once, sharpening the contrast that direct imaging of Earth-like planets demands.

Optogenetics Papers Depend on a Virus Batch Nobody Standardised

By Karim Osman / Sep 18, 2026

Optogenetics depends on viral vectors that no one standardises. Batch variation in titre and purity can flip circuit results, and methods sections rarely report it.