3
[2026-09-03] · src Dobkowski, Trok, Skakunenko, Japha, Groswasser, Efremov, Marletto, Fuentes, Penrose, Vedral, Schleich, Folman; Science Advances (2026-09-02), arXiv:2502.14535
Using a Stern-Gerlach-type "Quantum Galileo Interferometer" built into an atom chip, the team split a cloud of 87Rb atoms cooled to 3.3 nK into a coherent superposition of two wavepackets: one held stationary by the chip's magnetic-field gradient, the other released into genuine free fall for up to 2T = 2 ms, reaching a 7.5 micrometre separation before the two were recombined. The relative phase that accumulated between them - about 80 radians across 13 interference fringes - matched the phase predicted by inserting the equivalence principle directly into the Schrodinger equation, Delta-phi = -(m g2 / 3 hbar) T3, with residuals of roughly 2.5% (about 2 radians on 80) between the measured and simulated curves.
Neutron interferometry (Colella-Overhauser-Werner, 1975) already showed decades ago that gravity shifts a matter wave's phase, so the news is not "gravity affects quantum objects" in general. What is new is the configuration: one arm genuinely in free fall while the other is held static in the lab frame, rather than both arms sitting at fixed, different heights in a static potential - the quantum-mechanical version of Einstein's freely-falling elevator, not of a gradiometer. That specific configuration is also the one several proposals for gravitationally-induced wavefunction collapse - including a conjecture of Penrose's own, who is a coauthor here - use to predict a decoherence rate that larger superpositions should show. The separation and fall time in this run are far too small to be sensitive to any such effect - this confirms the phase, not any collapse - but it is now a working apparatus of exactly the kind that search would need, validated in the regime where standard quantum mechanics and general relativity agree.
[2026-09-03] · src LZ Collaboration, arXiv:2609.02823 (2026)
Searching an extended nuclear-recoil energy window up to roughly 270 keV - well above the range a standard WIMP search covers - LUX-ZEPLIN's latest analysis of its full exposure found a single event reconstructed at 248 +/- 23 (stat) +/- 23 (sys) keV, in a region where the collaboration's own background model predicts very little. A profile-likelihood test against the background-only hypothesis rejects that hypothesis at 2.6 sigma global significance once the look-elsewhere effect across the several models tested is included (3.4 sigma local, for the best-fitting model) - short of any discovery threshold.
A 2.6-sigma global excess is roughly what a search this size turns up from background fluctuation alone every so often, and the collaboration frames it that way rather than as a signal - it is the theory papers already fitting this one event to fermionic dark-matter absorption or a Higgsino that are the extraordinary claim this brief is wary of, not LZ's own report. What is real regardless of what the event turns out to be is the extended window itself: a detector built to see few-keV nuclear recoils demonstrating clean, understood sensitivity out past 250 keV is what made this event visible in the first place, and that capability is what LZ's remaining runtime - or an independent look from XENONnT or PandaX in the same window - will need to say whether this was a fluctuation or the start of something.
[2026-09-03] · src Zhan et al. (Wuhan Institute of Physics and Mathematics); Science Advances (2026-09), arXiv:2603.22981
A dual-species (85Rb/87Rb) cold-atom interferometer aboard the China Space Station compared the free-fall acceleration of the two isotopes over 280 days in continuous microgravity, finding them equal to within roughly 5 parts in 108 (Eotvos parameter consistent with zero at that level) - about three orders of magnitude tighter than any previous atom-interferometric test of the weak equivalence principle done in microgravity (drop towers, aircraft, sounding rockets).
This is still five to seven orders of magnitude looser than the tightest classical test, MICROSCOPE's ~10-15 from two macroscopic test masses in a dedicated satellite, so it does not compete with the best composition- dependent bound on a fifth force. What it does is close most of the gap that was specific to quantum tests of free fall - the class of experiment that could in principle see a violation coupled to superposition or coherence rather than to bulk composition, which a classical test mass cannot probe at all. Three orders of magnitude in one 280-day campaign, on a platform that can simply keep collecting, is the more informative number here than the absolute bound reached so far.
2
[2026-09-03] · src Arban, Complete Conservatory Method — 14 Characteristic Studies
Trumpet takes today's slot — three days since its last session, the longest gap of the three right now, continuing the alto → piano → trumpet three-day cycle this rotation has held since mid-August. Three sessions into the 14 Characteristic Studies, No. 1 asked for endurance, No. 2 for dynamic contrast and No. 3 for legato across a register break; No. 4 moves the demand again, to switching cleanly between slurred and tongued articulation inside a single phrase — the point where a legato group bleeds into the next attack instead of the tongue meeting it right where the slur mark ends.
7 minutes: a one-octave G major scale, ascending and descending, in four-note groups with the first two slurred and the last two tongued, at mp, quarter = 66, six times through — alternating which half of each group starts the slur so both directions of the switch, into a tongue and out of one, get equal rehearsal before the study asks for the same thing inside a real phrase.
Characteristic Study No. 4, played straight through once, treating every notated slur as the exact point where articulation must change rather than papering over the seam with a soft re-tongue or letting the tongue carry through what should stay connected. Success criterion: every slur-to-tongue or tongue-to-slur transition in the study is audible as a genuine change in attack — a listener following along with the score closed could still mark where each slur ends by ear alone.
Art of Phrasing, Melody No. 17. Play it once for notes, then once treating every articulation change in the melody as a phrasing decision rather than a mechanical hand-off — the same demand as the technical target, at a smaller scale and a slower tempo, where a careless choice is more exposed rather than less.
One clean pass at the Characteristic Study, same as No. 1 through No. 3 — a second pass right behind it invites a tiring embouchure to start covering for itself exactly where clean articulation matters most, training the blur this session is trying to eliminate. Stop there, or sooner if a transition starts needing an extra push to sound clean.
[2026-09-03] · src —
hearFreddie Hubbard — Birdlike (Ready for Freddie) · Tidal · Spotify
Cut in 1961 at a tempo fast enough that nothing can hide, Hubbard's solo here runs long slurred bebop lines straight into single sharply-tongued notes and back again, over and over, without the seam ever softening — exactly the transition today's technical target is chasing, just at a speed where getting it wrong would be impossible to disguise.