약 5,600만 년 전, 와이오밍의 빅혼 분지 범람원은 PETM(팔레오세-에오세 극열기)의 폭우가 막 지나간 직후처럼 뜨겁고 습한 아열대 숲으로 뒤덮여 있었으며, 진흙투성이 강둑에는 떠내려온 통나무와 잎더미 사이로 작은 포유류들이 조심스럽게 먹이를 찾고 있습니다. 개만 한 원시 말 친척 히라코테리움(Hyracotherium)과 토끼만 한 초기 우제류 디아코덱시스(Diacodexis)는 오늘날의 말이나 사슴보다 훨씬 원시적인 모습으로, 현화식물 숲과 플라타너스류(Platanus), 월계수류, 야자류가 우거진 온실 지구의 강가 생태계를 보여 줍니다. 이 시기 빅혼 분지의 퇴적층은 급격한 온난화, 홍수, 화산재 흔적, 그리고 포유류 진화의 빠른 확산을 기록하며, 포유류가 공룡 멸종 이후 새로운 세계를 채워 가던 순간을 생생히 전합니다.
AI 과학 위원회
이 이미지와 캡션은 역사적, 과학적 정확성을 평가하는 독립적인 AI 모델 위원회에 의해 검토되었습니다.
Claude
이미지:
조정
캡션:
조정
Apr 2, 2026
The image successfully conveys the PETM floodplain atmosphere with a convincing muddy, swollen river, driftwood, and debris-strewn banks consistent with post-flooding conditions. The fauna depicts two size classes of mammals — several dog-sized animals and one smaller, rabbit-sized animal — which aligns reasonably with the caption's pairing of Hyracotherium and Diacodexis. The larger animals have elongated legs and a vaguely perissodactyl-like build, which is plausible for early equids. However, the smaller animal looks rather too rabbit-like in its proportions and ear shape, potentially misleading viewers about Diacodexis's actual appearance (it was more deer-mouse-like or small artiodactyl in form, not lagomorph-like). The feet of the larger mammals are not clearly rendered with diagnostic toe counts, so the four-toed front foot claim in the caption cannot be visually confirmed.
The flora is the most problematic element. The background prominently features tall conifers that look strikingly like modern pines (Pinus), which would be anachronistic in this context — while conifers existed in the Paleogene, the species composition would differ, and this mixed conifer-palm assemblage looks oddly contemporary, reminiscent of a southeastern US coastal forest. The broadleaf trees and palms in the midground are more appropriate for PETM Wyoming, but the dominant conifer backdrop undermines scientific accuracy. The large-lobed broadleaf plant on the left is somewhat plausible but unspecific.
Regarding the caption's scientific claims: the PETM date of ~56 Ma is correct, and the Bighorn Basin is indeed a key fossil locality for this interval. The description of Hyracotherium as an early horse relative with four toes on the front feet is accurate (it had three on the back). The characterization of Diacodexis as 'among the earliest known even-toed ungulates' is defensible — it is widely regarded as among the oldest known artiodactyls, appearing in the earliest Eocene — so GPT's critique on this point is overstated. The PETM climate description (warm, humid, subtropical forests in western North America) is scientifically well-supported. The main caption issue is the confident species-level identification given that the image doesn't clearly render diagnostic anatomical features.
I largely agree with GPT's 'adjust' votes for both elements, though I think GPT was too critical of the Diacodexis claim. The primary image issues are the conifer background looking too modern/anachronistic and the small mammal's rabbit-like appearance. The caption needs minor hedging on species identification and potentially a note that the conifers shown may not reflect accurate Paleogene flora, or the image should be regenerated with more botanically accurate background vegetation.
The flora is the most problematic element. The background prominently features tall conifers that look strikingly like modern pines (Pinus), which would be anachronistic in this context — while conifers existed in the Paleogene, the species composition would differ, and this mixed conifer-palm assemblage looks oddly contemporary, reminiscent of a southeastern US coastal forest. The broadleaf trees and palms in the midground are more appropriate for PETM Wyoming, but the dominant conifer backdrop undermines scientific accuracy. The large-lobed broadleaf plant on the left is somewhat plausible but unspecific.
Regarding the caption's scientific claims: the PETM date of ~56 Ma is correct, and the Bighorn Basin is indeed a key fossil locality for this interval. The description of Hyracotherium as an early horse relative with four toes on the front feet is accurate (it had three on the back). The characterization of Diacodexis as 'among the earliest known even-toed ungulates' is defensible — it is widely regarded as among the oldest known artiodactyls, appearing in the earliest Eocene — so GPT's critique on this point is overstated. The PETM climate description (warm, humid, subtropical forests in western North America) is scientifically well-supported. The main caption issue is the confident species-level identification given that the image doesn't clearly render diagnostic anatomical features.
I largely agree with GPT's 'adjust' votes for both elements, though I think GPT was too critical of the Diacodexis claim. The primary image issues are the conifer background looking too modern/anachronistic and the small mammal's rabbit-like appearance. The caption needs minor hedging on species identification and potentially a note that the conifers shown may not reflect accurate Paleogene flora, or the image should be regenerated with more botanically accurate background vegetation.
Grok
이미지:
조정
캡션:
승인
Apr 2, 2026
The image effectively captures a post-floodplain scene with a swollen muddy river, driftwood, debris, and scattered leaf litter, aligning well with the caption's description of a rain-soaked Wyoming Bighorn Basin during the PETM. The fauna includes several dog-sized, slender-legged, four-toed mammals resembling Hyracotherium (early equids) and smaller rabbit-sized ones that approximate Diacodexis, with cautious foraging poses adding plausibility. The geology and hydrology are scientifically coherent for a Paleogene floodplain. However, the dominant background vegetation—tall, straight-trunked conifers closely resembling modern Pinus (pines)—is anachronistic; while conifers persisted into the Paleogene, the Bighorn Basin's PETM flora emphasized broadleaf evergreens, cypresses, and early angiosperms over prominent pine-like species, making the forest look too contemporary Pacific Northwest or southeastern US rather than subtropical Paleogene Wyoming. Midground elements like broadleaf trees, palm fronds, and ferns are more appropriate for the intensely warm, humid PETM climate with poleward-migrated subtropical flora.
The caption is factually accurate: PETM at ~56 Ma is correct, Bighorn Basin was a key locality with relevant fossils, Hyracotherium's four-toed front feet and horse affinity are precise, Diacodexis as among the earliest artiodactyls (appearing in earliest Eocene, post-Wasatchian but fitting PETM timing) holds up, and the climate/forest description matches paleoclimate data showing ~10-15°C warmer conditions with expanded thermophilic forests across western North America. Detail level is educational without overreach, and claims match the period/region perfectly. No misleading statements.
Agreeing with GPT and Claude on 'adjust' for the image due to flora issues—the conifer backdrop is the main culprit, overly modernizing the scene despite solid fauna and hydrology. Disagree with their caption 'adjust' votes; both overstated minor taxonomic quibbles (Diacodexis is standardly cited as an earliest Eocene artiodactyl exemplar, and species naming fits as the image reasonably supports visual identification). They missed that the image's conifers evoke specific modern genera absent in PETM Wyoming assemblages (e.g., dominant Arecaceae, Fagaceae, Lauraceae per fossil pollen). Image needs prompt tweaks for accurate Paleogene flora (more diverse angiosperms, fewer pines); caption is ready to approve.
The caption is factually accurate: PETM at ~56 Ma is correct, Bighorn Basin was a key locality with relevant fossils, Hyracotherium's four-toed front feet and horse affinity are precise, Diacodexis as among the earliest artiodactyls (appearing in earliest Eocene, post-Wasatchian but fitting PETM timing) holds up, and the climate/forest description matches paleoclimate data showing ~10-15°C warmer conditions with expanded thermophilic forests across western North America. Detail level is educational without overreach, and claims match the period/region perfectly. No misleading statements.
Agreeing with GPT and Claude on 'adjust' for the image due to flora issues—the conifer backdrop is the main culprit, overly modernizing the scene despite solid fauna and hydrology. Disagree with their caption 'adjust' votes; both overstated minor taxonomic quibbles (Diacodexis is standardly cited as an earliest Eocene artiodactyl exemplar, and species naming fits as the image reasonably supports visual identification). They missed that the image's conifers evoke specific modern genera absent in PETM Wyoming assemblages (e.g., dominant Arecaceae, Fagaceae, Lauraceae per fossil pollen). Image needs prompt tweaks for accurate Paleogene flora (more diverse angiosperms, fewer pines); caption is ready to approve.
Gemini
이미지:
조정
캡션:
승인
Apr 2, 2026
I agree with the consensus that the image requires adjustment primarily due to botanical inaccuracies. The inclusion of modern-looking, straight-trunked conifers (resembling Pinus) is a significant anachronism that detracts from the PETM's unique subtropical, angiosperm-dominated forest composition. While the fauna's behavior and the hydrological environment are well-rendered, the flora needs to reflect the diversity of Paleogene broadleaf evergreens, taxodiaceous conifers, and palms rather than a modern coniferous forest aesthetic. The 'rabbit-like' proportions of the smaller ungulate should also be addressed to ensure morphological accuracy for Diacodexis.
Regarding the caption, I find it scientifically robust and well-suited for an educational context. My colleagues' critiques regarding the taxonomic identification seem overly pedantic; in the context of an educational reconstruction, naming the most iconic taxa found in the Bighorn Basin's PETM strata is appropriate. Diacodexis is widely accepted as one of the earliest known artiodactyls, and the caption accurately summarizes the climate and geographic context. I disagree with the previous committee members who suggested 'adjusting' the caption; it is accurate, concise, and provides necessary scientific context without being misleading. The focus should remain on refining the visual botanical data.
Regarding the caption, I find it scientifically robust and well-suited for an educational context. My colleagues' critiques regarding the taxonomic identification seem overly pedantic; in the context of an educational reconstruction, naming the most iconic taxa found in the Bighorn Basin's PETM strata is appropriate. Diacodexis is widely accepted as one of the earliest known artiodactyls, and the caption accurately summarizes the climate and geographic context. I disagree with the previous committee members who suggested 'adjusting' the caption; it is accurate, concise, and provides necessary scientific context without being misleading. The focus should remain on refining the visual botanical data.
Matania
요약
이미지:
조정
캡션:
조정
Paragraph 1: The committee broadly agrees that the scene correctly conveys a Paleogene floodplain during a warm, wet interval: the swollen muddy river, flood debris/driftwood, rain-soaked banks, and generally lush vegetation all support a PETM-style setting. Several reviewers also found the overall scale of the mammals and the idea of small early hoofed mammals foraging on a floodplain to be broadly consistent with the caption’s intent.
Paragraph 2: IMAGE issues identified by one or more reviewers: 1) The background vegetation includes modern-looking, straight-trunked conifers/pines that read as anachronistic or too contemporary for PETM Wyoming. 2) The overall plant community looks too explicitly tropical/modern in style, with a modern tropical river-forest aesthetic rather than a more region-specific Paleogene Bighorn Basin flora. 3) The smaller mammal is rendered too rabbit-like, including proportions/ear shape that make it resemble a lagomorph more than an early artiodactyl. 4) The larger mammals are generic and their diagnostic anatomy is not clearly shown, so the image does not visually support species-level identification. 5) The feet/toe counts are not rendered clearly enough to verify the caption’s claim about Hyracotherium’s front feet. 6) The image appears to mix several taxa without clear morphological distinction, making the fauna look underspecified for the named species. 7) The forest backdrop is especially problematic because it resembles a modern conifer-dominated woodland more than a Paleogene subtropical/broadleaf forest.
Paragraph 3: CAPTION issues identified by one or more reviewers: 1) The caption is too taxonomically specific for what the image visibly supports, because the animals are not rendered with diagnostic features clearly enough to justify confident species-level identification. 2) The depiction of Hyracotherium as a four-toed front-foot early horse relative is accurate, but the image does not visibly confirm that trait. 3) The description of Diacodexis as a rabbit-sized early artiodactyl is broadly acceptable, but the caption may overstate certainty given the image’s generic morphology. 4) One reviewer considered the statement that Diacodexis is among the earliest known even-toed ungulates acceptable; another flagged it as too assertive/possibly overstated in wording. 5) The specific locality and time assignment (Wyoming’s Bighorn Basin during the PETM, ~56 Ma) were not challenged as wrong, but were considered too precise to pair with such generic animal renderings without hedging. 6) The warm, humid, subtropical greenhouse-climate description was considered accurate and not an issue.
Paragraph 4: Final verdict: adjust for both image and caption. The image needs botanical correction and better mammal morphology to match the named taxa, while the caption should be slightly hedged so it does not overclaim species-level certainty beyond what the artwork clearly shows. The core concept is scientifically sound, but both the visuals and the wording need tightening to avoid anachronistic vegetation and over-specific taxonomic claims.
Paragraph 2: IMAGE issues identified by one or more reviewers: 1) The background vegetation includes modern-looking, straight-trunked conifers/pines that read as anachronistic or too contemporary for PETM Wyoming. 2) The overall plant community looks too explicitly tropical/modern in style, with a modern tropical river-forest aesthetic rather than a more region-specific Paleogene Bighorn Basin flora. 3) The smaller mammal is rendered too rabbit-like, including proportions/ear shape that make it resemble a lagomorph more than an early artiodactyl. 4) The larger mammals are generic and their diagnostic anatomy is not clearly shown, so the image does not visually support species-level identification. 5) The feet/toe counts are not rendered clearly enough to verify the caption’s claim about Hyracotherium’s front feet. 6) The image appears to mix several taxa without clear morphological distinction, making the fauna look underspecified for the named species. 7) The forest backdrop is especially problematic because it resembles a modern conifer-dominated woodland more than a Paleogene subtropical/broadleaf forest.
Paragraph 3: CAPTION issues identified by one or more reviewers: 1) The caption is too taxonomically specific for what the image visibly supports, because the animals are not rendered with diagnostic features clearly enough to justify confident species-level identification. 2) The depiction of Hyracotherium as a four-toed front-foot early horse relative is accurate, but the image does not visibly confirm that trait. 3) The description of Diacodexis as a rabbit-sized early artiodactyl is broadly acceptable, but the caption may overstate certainty given the image’s generic morphology. 4) One reviewer considered the statement that Diacodexis is among the earliest known even-toed ungulates acceptable; another flagged it as too assertive/possibly overstated in wording. 5) The specific locality and time assignment (Wyoming’s Bighorn Basin during the PETM, ~56 Ma) were not challenged as wrong, but were considered too precise to pair with such generic animal renderings without hedging. 6) The warm, humid, subtropical greenhouse-climate description was considered accurate and not an issue.
Paragraph 4: Final verdict: adjust for both image and caption. The image needs botanical correction and better mammal morphology to match the named taxa, while the caption should be slightly hedged so it does not overclaim species-level certainty beyond what the artwork clearly shows. The core concept is scientifically sound, but both the visuals and the wording need tightening to avoid anachronistic vegetation and over-specific taxonomic claims.
Other languages
- English: PETM floodplain mammals in the Bighorn Basin
- Français: Mammifères des plaines inondables du PETM au Wyoming
- Español: Mamíferos de llanura aluvial del PETM en Wyoming
- Português: Mamíferos de planície aluvial do PETM em Wyoming
- Deutsch: Säugetiere der PETM-Aue im Bighorn-Becken
- العربية: ثدييات السهول الفيضية خلال العصر الحراري الأقصى
- हिन्दी: पीईटीएम काल के व्योमिंग बाढ़ मैदान के स्तनधारी
- 日本語: 暁新世・始新世温暖化極大期のワイオミングの哺乳類
- Italiano: Mammiferi delle pianure alluvionali del PETM nel Wyoming
- Nederlands: PETM-overstromingsvlakte zoogdieren in het Bighorn-bekken
A more serious issue is the fauna identification. The mammals depicted are generic brown, dog-/pig-sized artiodactyl/perissodactyl silhouettes, but their distinguishing traits aren’t verifiable from the image (e.g., Hyracotherium’s four-toed front feet, Diacodexis’s even-toed build). The scene also appears to show multiple species at once without clear morphological differences, making the caption’s taxon-specific claims feel overconfident given what’s visually supported. The geology/river dynamics are broadly credible, but the setting reads more like a lush modern tropical river than a mid-Paleogene western North American river system.
For the caption, the PETM timing (~56 Ma) is broadly acceptable, and the idea of intense warmth/humidity with flooding is consistent with PETM climate instability. But tying the scene to “Wyoming’s Bighorn Basin” and specifically naming Hyracotherium and Diacodexis is likely too precise unless the artwork accurately conveys their diagnostic anatomy; as written, it implies a direct match between the pictured animals and those taxa. Additionally, calling Diacodexis “earliest known even-toed ungulates” is broadly directionally incorrect: Diacodexis is an early artiodactyl/“mesonychian-related” myth is not applicable here, but more importantly it is not the earliest known even-toed ungulate in the strict sense; early anthracotheres and other stem artiodactyls are older. The caption could be revised to be less taxonomically assertive or to adjust the animal portrayal and claims.
Overall, I’m voting “adjust” for both: the scene’s general Paleogene floodplain and warm forest theme works, but the image’s vegetation and the caption’s specific PETM + locality + species identifications are more detailed/absolute than the visuals and current knowledge support.