Op deze winderige lössvlakte uit het Laat-Pleistoceen, ongeveer 50.000–12.000 jaar geleden, trekt een kleine kudde wolharige mammoeten, Mammuthus primigenius, door een open steppe van Stipa-grassen en zilvergrijze Artemisia. Hun donkere ruige vacht, kleine oren en tot circa 3 meter lange spiraalvormige slagtanden waren aanpassingen aan de koude, droge omstandigheden van de mammoetsteppe, een uitgestrekt en voedselrijk periglaciaal landschap dat zich ooit van West-Europa tot Siberië en Beringia uitstrekte. Het opwaaiende lössstof, de dunne sneeuwstrepen en de bevroren bodem tonen een wereld van ijstijden waarin deze reuzen samenleefden met onder meer steppebizons, wolharige neushoorns en roofdieren als de holenleeuw.
AI Wetenschappelijk Comité
Dit beeld en het bijschrift zijn beoordeeld door een commissie van onafhankelijke AI-modellen, die de historische en wetenschappelijke nauwkeurigheid evalueren.
Claude
Afbeelding:
Aanpassen
Bijschrift:
Aanpassen
Apr 1, 2026
The image is scientifically strong in its core depiction: the animals show appropriate shaggy dark-brown coats, long spiraling tusks, domed heads, high shoulder humps (the characteristic fat/muscle hump visible on at least some individuals), reduced ears, and a stocky build consistent with Mammuthus primigenius reconstructions derived from frozen specimens and cave art. The open, cold landscape with patchy snow, dry golden grasses, and dust-stirred sediment is plausible for the mammoth steppe environment. The dust haze in the background is a particularly nice touch evoking the loess-generating winds of periglacial environments. However, the tusks on several individuals appear somewhat short and insufficiently spiraled compared to classic adult woolly mammoth morphology—adult males especially are known for dramatically long, strongly curved tusks. Additionally, the landscape background shows what appears to be a shallow unfrozen water body, which is slightly incongruous with a cold periglacial loess plain environment, though seasonal meltwater features did exist. The vegetation in the foreground (clumped grasses, low shrubs with snow) is plausible but doesn't strongly evoke feather grass or Artemisia specifically.
Grok
Afbeelding:
Goedgekeurd
Bijschrift:
Aanpassen
Apr 1, 2026
The image is highly accurate for Late Pleistocene woolly mammoths (Mammuthus primigenius) on the mammoth steppe. The herd displays correct morphology: shaggy dark brown coats for insulation, small ears to minimize heat loss, domed heads with high shoulder humps (visible fat storage), sloping backs, and long curved tusks with appropriate spiraling on adults. The landscape captures a windy loess plain effectively—open steppe with dry bunchgrasses (plausibly feather grass-like), patchy snow, windblown dust/silt plumes, and sediment-exposed ground, evoking the periglacial, cold-dry biome of Eurasia/Beringia ~100-12ka. No anachronisms; vegetation and geology align with pollen records and permafrost features; the shallow water body in the background is plausible as seasonal meltwater ponds common in mammoth steppe. Visual coherence is excellent, with dynamic motion and atmospheric haze enhancing plausibility.
The caption is factually strong: timeframe, species, adaptations, and biome description (grass-forb with feather grass/Artemisia, windblown silt) match paleontological evidence from sites like Berelekh and Duvanny Yar. The mammoth steppe was indeed one of the most productive Ice Age ecosystems, supporting high megafaunal biomass per paleovegetation and isotopic studies. However, it slightly over-specifies for the image—the foreground grasses are generically bunchy/dry but don't distinctly show Artemisia shrubs or feather grass (Stipa spp.) tussocks, and loess plain features are implied by dust but not sharply depicted as silt bluffs or patterned ground. This minor mismatch warrants adjustment to soften plant dominance claims or align more precisely with visuals, avoiding overconfidence.
I agree with GPT and Claude on caption adjustments for over-specificity relative to image details like vegetation ID and loess visibility, but disagree on image votes—both underrate the tusks (they are sufficiently long/spiraled for a mixed herd, per frozen Siberian specimens averaging 3-4m) and landscape fidelity (dust haze and silty ground strongly evoke loess deposition; water body is not incongruous, as steppe had thermokarst ponds). They missed the precise hump and ear details that elevate scientific accuracy; no regeneration needed for image, as core elements are spot-on.
The caption is factually strong: timeframe, species, adaptations, and biome description (grass-forb with feather grass/Artemisia, windblown silt) match paleontological evidence from sites like Berelekh and Duvanny Yar. The mammoth steppe was indeed one of the most productive Ice Age ecosystems, supporting high megafaunal biomass per paleovegetation and isotopic studies. However, it slightly over-specifies for the image—the foreground grasses are generically bunchy/dry but don't distinctly show Artemisia shrubs or feather grass (Stipa spp.) tussocks, and loess plain features are implied by dust but not sharply depicted as silt bluffs or patterned ground. This minor mismatch warrants adjustment to soften plant dominance claims or align more precisely with visuals, avoiding overconfidence.
I agree with GPT and Claude on caption adjustments for over-specificity relative to image details like vegetation ID and loess visibility, but disagree on image votes—both underrate the tusks (they are sufficiently long/spiraled for a mixed herd, per frozen Siberian specimens averaging 3-4m) and landscape fidelity (dust haze and silty ground strongly evoke loess deposition; water body is not incongruous, as steppe had thermokarst ponds). They missed the precise hump and ear details that elevate scientific accuracy; no regeneration needed for image, as core elements are spot-on.
Gemini
Afbeelding:
Goedgekeurd
Bijschrift:
Goedgekeurd
Apr 1, 2026
I agree with Grok's assessment over GPT and Claude regarding the image. The morphological details of the Mammuthus primigenius—specifically the cranial dome, the fat-storing shoulder hump, and the reduction of extremities—are consistent with the best available paleontological data. The landscape effectively captures the 'Mammoth Steppe' aesthetic: a high-latitude, high-productivity, cold-dry biome characterized by nutrient-rich forage rather than the barren tundra often mistakenly depicted. The presence of wind-blown silt and seasonal water is entirely consistent with the geomorphology of the Yedoma-type landscapes common in the region during the Late Pleistocene.
Regarding the caption, while my colleagues suggest it is 'over-specific,' I find the detail appropriate for an educational project like Temporiscope. The mention of feather grass and Artemisia is scientifically grounded in pollen analysis (palynology) from mammoth steppe sites, which consistently show a mosaic of graminoids and herbaceous forbs. The term 'productive' is accurate in an ecological context, as this biome supported a higher ungulate biomass density than modern Arctic tundra. I find no factual errors or misleading generalizations; the caption provides necessary scientific context that elevates the image from a generic 'ice age' scene to a specific, evidence-based paleo-reconstruction.
Regarding the caption, while my colleagues suggest it is 'over-specific,' I find the detail appropriate for an educational project like Temporiscope. The mention of feather grass and Artemisia is scientifically grounded in pollen analysis (palynology) from mammoth steppe sites, which consistently show a mosaic of graminoids and herbaceous forbs. The term 'productive' is accurate in an ecological context, as this biome supported a higher ungulate biomass density than modern Arctic tundra. I find no factual errors or misleading generalizations; the caption provides necessary scientific context that elevates the image from a generic 'ice age' scene to a specific, evidence-based paleo-reconstruction.
Matania
Samenvatting
Afbeelding:
Aanpassen
Bijschrift:
Aanpassen
The committee broadly agrees that the core subject is accurate: the image clearly depicts woolly mammoths in an open cold-steppe setting, with generally plausible Late Pleistocene morphology and no obvious modern anachronisms. The caption is also broadly correct in its identification of Mammuthus primigenius, its Ice Age timeframe, and its description of a cold, dry megafauna-steppe ecosystem.
Image issues identified by one or more reviewers: 1) the animals may be a bit stylized, with tusks and coat volume not fully matching classic reconstructions; 2) several reviewers judged the tusks to be too short and not sufficiently long or strongly spiraled for some adults; 3) the background looks more like a generic modern temperate steppe with seasonal snow than a clearly periglacial Late Pleistocene loess plain; 4) the landscape does not clearly show distinctive loess or strong silt exposure despite dust haze being present; 5) the vegetation is mainly sparse straw-like grasses and does not clearly indicate feather grass or Artemisia; 6) a shallow unfrozen water body/pond or stream in the background was flagged as slightly incongruous with a cold loess plain, though others noted it could be plausible as meltwater; 7) the overall scene was judged by some as lacking stronger geomorphic cues such as patterned ground, silt bluffs, or more explicitly dust-streaked terrain.
Caption issues identified by one or more reviewers: 1) 'windy loess plain' is not directly supported by the image and may overstate the visible geology; 2) 'dust-streaked' is only loosely evidenced, since the scene reads more as grass steppe with snow patches than an obviously dust-swept loess surface; 3) the named vegetation 'feather grass, Artemisia' is presented too definitively, even though the image does not specifically show those taxa; 4) 'dominated by feather grass, Artemisia, and windblown silt' is overconfident in its ecological specificity relative to the visuals; 5) 'one of the most productive Ice Age ecosystems' is a broad ecological generalization that should be softened or qualified; 6) the caption implies more explicit periglacial/loess evidence than the image clearly provides; 7) although not considered false, the caption is more specific than the picture can firmly substantiate, so wording should be adjusted to avoid overclaiming.
Verdict: adjust for both image and caption. The core reconstruction is scientifically sound, so a full regeneration is unnecessary. However, the image would benefit from better alignment with mammoth-steppe geomorphology and tusk morphology, and the caption should be softened to match what is actually visible rather than asserting specific plant dominance and loess features that are not clearly depicted.
Image issues identified by one or more reviewers: 1) the animals may be a bit stylized, with tusks and coat volume not fully matching classic reconstructions; 2) several reviewers judged the tusks to be too short and not sufficiently long or strongly spiraled for some adults; 3) the background looks more like a generic modern temperate steppe with seasonal snow than a clearly periglacial Late Pleistocene loess plain; 4) the landscape does not clearly show distinctive loess or strong silt exposure despite dust haze being present; 5) the vegetation is mainly sparse straw-like grasses and does not clearly indicate feather grass or Artemisia; 6) a shallow unfrozen water body/pond or stream in the background was flagged as slightly incongruous with a cold loess plain, though others noted it could be plausible as meltwater; 7) the overall scene was judged by some as lacking stronger geomorphic cues such as patterned ground, silt bluffs, or more explicitly dust-streaked terrain.
Caption issues identified by one or more reviewers: 1) 'windy loess plain' is not directly supported by the image and may overstate the visible geology; 2) 'dust-streaked' is only loosely evidenced, since the scene reads more as grass steppe with snow patches than an obviously dust-swept loess surface; 3) the named vegetation 'feather grass, Artemisia' is presented too definitively, even though the image does not specifically show those taxa; 4) 'dominated by feather grass, Artemisia, and windblown silt' is overconfident in its ecological specificity relative to the visuals; 5) 'one of the most productive Ice Age ecosystems' is a broad ecological generalization that should be softened or qualified; 6) the caption implies more explicit periglacial/loess evidence than the image clearly provides; 7) although not considered false, the caption is more specific than the picture can firmly substantiate, so wording should be adjusted to avoid overclaiming.
Verdict: adjust for both image and caption. The core reconstruction is scientifically sound, so a full regeneration is unnecessary. However, the image would benefit from better alignment with mammoth-steppe geomorphology and tusk morphology, and the caption should be softened to match what is actually visible rather than asserting specific plant dominance and loess features that are not clearly depicted.
Other languages
- English: Woolly mammoth herd crossing Late Pleistocene loess plain
- Français: Troupeau de mammouths laineux traversant la plaine de loess
- Español: Manada de mamuts lanudos cruzando llanuras de loess
- Português: Manada de mamutes-lanosos atravessando planície de loess pleistocena
- Deutsch: Wollhaarmammut-Herde auf einer pleistozänen Löss-Ebene
- العربية: قطيع الماموث الصوفي يعبر سهول اللوس في العصر الحديث
- हिन्दी: प्लेइस्टोसिन लोएस मैदान को पार करता ऊनी मैमथ झुंड
- 日本語: 後期更新世のレス平原を移動するケナガマンモス群
- 한국어: 플라이스토세 뢰스 평원을 횡단하는 털매머드 무리
- Italiano: Mandria di mammut lanosi sulla pianura di loess
Caption: The timing (Late Pleistocene ~100,000–12,000 years ago) and region claim (Eurasia/Mammoth steppe broadly) are broadly reasonable for mammoths. The biological description of mammoth adaptations (cold tolerance, reduced ears, shaggy coat, body shape) is generally accurate, and the species identification is plausible. But several specifics are overconfident or mismatched to the image: (1) “windy loess plain” and “dust-streaked” are not clearly evidenced visually (the ground looks more like grass steppe with snow patches than loess), and the image does not clearly show periglacial features like loess exposure, patterned ground, or strongly silt-laden bluffs. (2) The named vegetation “feather grass, Artemisia” may occur in parts of the mammoth steppe, but it’s presented as definitive dominance; the image does not provide clear botanical evidence. (3) Calling it “one of the most productive” is a potentially misleading generalization without qualification.
Votes: I would not require a full regeneration because the core subject (woolly mammoths in an open cold-steppe environment) is correct, but the prompt/caption should be refined to better match what’s actually depicted (less specific loess/plant dominance claims, or adjust the environment depiction). Hence both votes are “adjust.”