Key points
Late August 2026 saw three independent results land almost simultaneously, spanning strong-field relativity, cosmic-scale structure, and dark matter physics.
- Relativity — star S301: A team led by Stefan Gillessen (Max Planck Institute for Extraterrestrial Physics), using ESO's Very Large Telescope (VLT) with continuous tracking since 2023, determined that star S301 orbits Sagittarius A* (Sgr A*), the Milky Way's central supermassive black hole (~4 million solar masses), with:
- Peak speed above 8% of light speed (~24,000 km/s) — the fastest star yet measured.
- Pericenter distance of only ~1.8 billion km from Sgr A* (about 12 times the Earth–Sun distance), ten times closer than the previous record holder.
- Gillessen's analogy: from a planet around S301 at closest approach, the black hole would appear as large as the full Moon in Earth's sky.
- Scientific value of S301 — measuring frame dragging: At such close range, S301 traverses one of the most extreme gravitational environments known. If Sgr A* rotates, frame dragging should measurably alter the star's orbit. Gillessen likens the method to "dropping a leaf into the wind" — the star is the leaf probing spacetime. He estimates a full spin measurement from S301 alone would take about a decade; a population of similar stars would accelerate this. UCLA astronomer Tuan Do notes measuring black hole spin is far harder than Earth's analogous (far weaker) effect because black holes have no surface to observe. UCL astronomer Ziri Younsi agrees one star already gives the tightest spin constraints, but a population would improve matters further.
- Cosmic web — first direct filament observation: Using the VLT's MUSE instrument, scientists directly observed a ~3-million-light-year-long cosmic filament connecting two galaxies — the first direct imaging of a filament. This confirms the standard picture that galaxies sit along dark-matter-and-gas filaments and are fed gas (fuel for star formation) along them, moving the cosmic web from simulation-based theory toward direct observational evidence.
- Dark matter — dark stars as black hole seeds: On August 29, Sohan Ghodla and Cosmin Ilie (Colgate University) published a Letter in *Physical Review D* proposing that dark stars — hypothetical early-universe objects powered partly by dark matter rather than fusion, potentially growing to millions-to-hundreds-of-millions of solar masses before collapsing — could be the seeds of the first supermassive black holes. Analyzing the nanohertz gravitational-wave background detected by pulsar timing arrays (PTAs), they find dark-star-remnant seed densities of ~10⁻³ per cubic megaparsec, versus ~10⁻⁶ /Mpc³ for direct-collapse seeds — meaning dark star remnants could dominate the PTA signal, while direct-collapse seeds contribute negligibly.
- Ghodla: "Produce too many of these massive seeds and you end up over-producing the PTA-detected signal. Produce too few, and you need other sources to efficiently assemble these supermassive black holes later in the life of the universe to match PTA observations."
- Ilie: PTAs are usually seen as probes of relatively nearby supermassive black hole binaries, "what our work shows is that the signal may also contain information about how the ancestors of those black holes formed at cosmic dawn."
- Toutiao / "AI Scientists" (8-29): S301 star details and expert commentary
- Tencent News / "Popular Science New Vision" (8-29): MUSE filament; JWST early black holes
- Tencent News / "Cosmos Probing" (8-29): dark star Physical Review D Letter, Colgate University
- Time News (8-29): dark stars, PTA nanohertz gravitational waves
- Additional 8-29/8-30 reports on the Roman telescope launch and related cosmology topics
Caveats
1. One star vs. a population: S301 alone needs ~10 years to constrain Sgr A*'s spin; the milestone is not yet a measurement. A population of pericenter stars is needed. 2. One filament vs. the full web: A single 3-million-light-year filament is an important data point, but validating the cosmic web requires filaments across multiple scales and directions (filaments + nodes + voids). 3. Hypothesis vs. evidence for dark stars: The dark-star result is an indirect inference from PTA data; dark stars have never been directly observed. Dark-star and direct-collapse seed models remain mutually exclusive alternatives that current data cannot discriminate — direct signals may require JWST, Roman, or ngVLA.
Open questions for the next 6–12 months
1. Can follow-up orbital data on S301 (plus any newly confirmed pericenter stars) yield preliminary Sgr A* spin constraints by H2 2027, potentially compressing the timeline from 10 years to 3–5? 2. Can MUSE/JWST/Roman directly observe more filaments at varied scales by H1 2027? 3. Will improved PTA precision falsify one of the two early black-hole-seed models? 4. Can JWST/Roman/ngVLA detect first-generation dark stars directly in 2027–2028 — a "holy grail" for early-universe cosmology?