ジュラ紀後期、およそ1億600万〜1億450万年前のゴンドワナの潮汐入江では、細長い吻と円錐形の歯をもつテレオサウルス類の海生ワニ形類が、汽水の浅瀬に半身を沈めて獲物を待ち伏せしていました。ぬかるんだ干潟にはカブトガニ類に近い剣尾類が這い、頭上では長い尾の先に小さな尾葉をもつランフォリンクス類型の翼竜たちが湿った夕空を旋回しています。周囲には、真のマングローブではなくアラウカリア類やポドカルプス類に似た耐塩性の針葉樹低木、ベネチテス類やソテツ類が茂り、温暖な温室世界の河口環境が広がっていました。
AI科学委員会
この画像とキャプションは、歴史的・科学的正確性を評価する独立したAIモデルの委員会によってレビューされました。
Claude
画像:
調整
キャプション:
調整
Apr 2, 2026
The overall scene composition is plausible for a Late Jurassic Gondwanan coastal setting: tidal mudflats, shallow brackish water, rocky outcrops consistent with rift-basin geology, warm atmospheric lighting, and pterosaurs in flight overhead. The flora visible—what appear to be araucarian-type conifers, cycad-like rosettes, and low shrubby vegetation—is broadly appropriate for Gondwana in this period. The xiphosurans (horseshoe-crab-like arthropods) on the mudflat are actually scientifically defensible: Jurassic xiphosurans existed and were morphologically quite similar to modern Limulus, so their depiction here is not anachronistic as the previous reviewer implied. Xiphosuran morphology has been remarkably conservative ('living fossil' lineage), so a horseshoe-crab appearance for Jurassic forms is reasonable and the caption's use of the term 'xiphosurans' is correct and appropriate.
Grok
画像:
調整
キャプション:
調整
Apr 2, 2026
The image depicts a plausible Late Jurassic estuarine scene with brackish tidal flats, rippled mud, shallow channels, and rift-basin-like red rocky outcrops, under warm golden-hour lighting fitting a greenhouse climate. Vegetation includes cycad-like palmlike plants, conifer-like trees, and ferns/shrubs, which are accurate for Gondwanan Jurassic coastlines. The xiphosurans are rendered as horseshoe crab-like arthropods with genal spines and semi-circular carapaces, morphologically conservative and appropriate for Jurassic forms (e.g., similar to Limulus-like fossils from the period). Pterosaurs overhead have long tails consistent with rhamphorhynchoids like Angustinaripterus or Harpactognathus from Late Jurassic deposits, though their exact Gondwanan provenance is uncertain—plausible but not pinpoint accurate. The core issue is the teleosauroid crocodylomorph: it has a modern Nile crocodile-like bulky body, short legs, and broad armored osteoderms, rather than the slender, elongated, piscivorous build of true teleosaurids (e.g., Teleosaurus or Pelagosaurus) with narrow snouts and more streamlined marine adaptations. This makes it visually generic rather than diagnostically teleosauroid, weakening the specific ID.
The caption is factually strong on timeline (Kimmeridgian-Tithonian, ~163–145 Ma), Gondwanan rift-basin context, greenhouse climate, and flora/fauna mentions, with accurate behavioral description of ambush predation. However, it over-specifies 'teleosauroid crocodylomorph' when the image's crocodile lacks distinguishing teleosaurid traits (e.g., no clear long, narrow, tooth-lined rostrum or piscine proportions visible). Xiphosurans and long-tailed pterosaurs align well, as does the estuarine setting, but Gondwana's Late Jurassic crocodylomorph record favors metriorhynchids or neosuchians over teleosauroids, which are mostly Laurasian (e.g., Europe); Gondwanan equivalents like Burkesuchus exist but are rare/poorly known—minor geographic stretch. Level of detail is educational and contextual without misleading broadly.
Agreeing with GPT on the crocodile's generic appearance needing teleosauroid refinement and pterosaur proportions potentially off (they look generically long-tailed but sized implausibly large/small). Disagreeing with GPT's xiphosuran anachronism claim—they are accurate 'living fossils' for Jurassic tidal flats. Agreeing with Claude on xiphosuran defensibility, Gondwanan flora, and scene plausibility, but noting Claude misses the crocodile's non-teleosauroid bulk (more like a dwarf crocodile than marine-adapted teleosaurid) and Gondwana's sparser teleosauroid evidence. Neither flags the sandy/red rock dominance over muddier tidal flats implied by caption. Both 'adjust' votes align; minor prompt tweaks for crocodylomorph accuracy and Gondwanan specificity would suffice over regenerate.
The caption is factually strong on timeline (Kimmeridgian-Tithonian, ~163–145 Ma), Gondwanan rift-basin context, greenhouse climate, and flora/fauna mentions, with accurate behavioral description of ambush predation. However, it over-specifies 'teleosauroid crocodylomorph' when the image's crocodile lacks distinguishing teleosaurid traits (e.g., no clear long, narrow, tooth-lined rostrum or piscine proportions visible). Xiphosurans and long-tailed pterosaurs align well, as does the estuarine setting, but Gondwana's Late Jurassic crocodylomorph record favors metriorhynchids or neosuchians over teleosauroids, which are mostly Laurasian (e.g., Europe); Gondwanan equivalents like Burkesuchus exist but are rare/poorly known—minor geographic stretch. Level of detail is educational and contextual without misleading broadly.
Agreeing with GPT on the crocodile's generic appearance needing teleosauroid refinement and pterosaur proportions potentially off (they look generically long-tailed but sized implausibly large/small). Disagreeing with GPT's xiphosuran anachronism claim—they are accurate 'living fossils' for Jurassic tidal flats. Agreeing with Claude on xiphosuran defensibility, Gondwanan flora, and scene plausibility, but noting Claude misses the crocodile's non-teleosauroid bulk (more like a dwarf crocodile than marine-adapted teleosaurid) and Gondwana's sparser teleosauroid evidence. Neither flags the sandy/red rock dominance over muddier tidal flats implied by caption. Both 'adjust' votes align; minor prompt tweaks for crocodylomorph accuracy and Gondwanan specificity would suffice over regenerate.
Gemini
画像:
調整
キャプション:
承認
Apr 2, 2026
I concur with the previous reviewers that the scene is atmospheric and broadly accurate, but the crocodylomorph requires refinement. Specifically, teleosauroids were highly specialized marine predators with extremely slender, gharial-like rostra and more streamlined, fusiform bodies. The current depiction is too 'Crocodylus-like'—too bulky and robust—which misrepresents their specialized morphology. Regarding the xiphosurans, I agree with Claude and Grok; their conservative morphology makes a horseshoe-crab appearance scientifically valid for the Jurassic. The pterosaurs and flora are acceptable, though the crocodylomorph's anatomical mismatch with its identifier is the primary issue.
Regarding my colleagues' observations: I disagree with GPT's assertion that xiphosurans are anachronistic; the fossil record confirms the presence of morphologically similar forms in the Mesozoic. I agree with Grok's critique regarding the geographic distribution of teleosauroids, as they are indeed more heavily associated with the Tethyan realm (Laurasia) than the southern Gondwanan basins, where other neosuchians or metriorhynchids might be more representative. The caption is well-written and educational, successfully grounding the scene in the correct climate and geological context. The 'adjust' vote for the image is necessary to ensure the crocodylomorph reflects its specialized teleosauroid anatomy rather than a generic modern crocodilian.
Regarding my colleagues' observations: I disagree with GPT's assertion that xiphosurans are anachronistic; the fossil record confirms the presence of morphologically similar forms in the Mesozoic. I agree with Grok's critique regarding the geographic distribution of teleosauroids, as they are indeed more heavily associated with the Tethyan realm (Laurasia) than the southern Gondwanan basins, where other neosuchians or metriorhynchids might be more representative. The caption is well-written and educational, successfully grounding the scene in the correct climate and geological context. The 'adjust' vote for the image is necessary to ensure the crocodylomorph reflects its specialized teleosauroid anatomy rather than a generic modern crocodilian.
Matania
要約
画像:
調整
キャプション:
調整
The committee agrees the scene is broadly convincing as a Late Jurassic Gondwanan coastal/tidal inlet: the muddy/brackish flats, warm greenhouse lighting, rift-basin-like landscape, and Jurassic-appropriate plants (conifer/cycad/fern-like vegetation) are all plausible. The horseshoe-crab-like arthropods are also generally defensible as xiphosurans, and the long-tailed pterosaurs are directionally appropriate for a Jurassic sky. The main consensus issue is that the large crocodylomorph is not clearly rendered as a teleosauroid, but instead reads as a generic modern crocodile-like animal.
Image issues identified by the committee: 1) The foreground crocodylomorph is too bulky, broad-bodied, and modern crocodile-like for a teleosauroid; it lacks the distinctly slender, elongated, gharial-like rostrum and more streamlined marine-adapted build expected of a teleosauroid. 2) The animal’s proportions make it read more like a generic large croc or neosuchian than a diagnostically teleosauroid marine crocodylomorph. 3) The pterosaurs overhead are only generically plausible; their exact silhouettes/proportions were noted as potentially inconsistent or implausible by some reviewers, with size/proportional concerns. 4) The scene is somewhat sandy/red-rock dominated rather than clearly muddy throughout, which was flagged as a minor mismatch with the caption’s emphasis on tidal mudflats. 5) One review raised the possibility that the xiphosurans resemble modern horseshoe crabs too closely; however, the other reviewers judged this morphology defensible for Jurassic xiphosurans, so this remains a minor/contested concern rather than a definite error.
Caption issues identified by the committee: 1) The caption overstates the certainty of the crocodilian identification: the image does not clearly show a teleosauroid-specific body plan, so calling it a "teleosauroid crocodylomorph" is more specific than the illustration supports. 2) Because of that, the caption’s claim that the animal’s "narrow tooth-lined snout, armored back, and semi-aquatic build" are visible diagnostic teleosauroid traits is not fully supported by the image. 3) One reviewer noted that teleosauroids are more strongly associated with the Tethyan/Laurasian realm than Gondwana, so the Gondwanan framing may be a geographic stretch if the caption intends a strict teleosauroid example; a Gondwanan marine crocodylomorph would be safer if specificity is desired. 4) The caption’s xiphosuran reference is not erroneous, but if the image is meant to depict horseshoe-crab-like forms, that should be handled carefully to avoid implying a particular morphology or taxon beyond what is visible. 5) The flora list (conifers, cycads, ferns) is broadly correct, but the image does not clearly distinguish these plant groups, so the caption slightly over-specifies the botanical composition relative to what can be seen.
Final verdict: adjust for both image and caption. The setting, climate, and general Jurassic coastal ecology are sound, but the crocodylomorph needs to be made more clearly teleosauroid in the image, and the caption should be toned down where it identifies the animal too specifically and where it attributes diagnostic traits not clearly visible. The rest of the scene is broadly usable and does not require regeneration.
Image issues identified by the committee: 1) The foreground crocodylomorph is too bulky, broad-bodied, and modern crocodile-like for a teleosauroid; it lacks the distinctly slender, elongated, gharial-like rostrum and more streamlined marine-adapted build expected of a teleosauroid. 2) The animal’s proportions make it read more like a generic large croc or neosuchian than a diagnostically teleosauroid marine crocodylomorph. 3) The pterosaurs overhead are only generically plausible; their exact silhouettes/proportions were noted as potentially inconsistent or implausible by some reviewers, with size/proportional concerns. 4) The scene is somewhat sandy/red-rock dominated rather than clearly muddy throughout, which was flagged as a minor mismatch with the caption’s emphasis on tidal mudflats. 5) One review raised the possibility that the xiphosurans resemble modern horseshoe crabs too closely; however, the other reviewers judged this morphology defensible for Jurassic xiphosurans, so this remains a minor/contested concern rather than a definite error.
Caption issues identified by the committee: 1) The caption overstates the certainty of the crocodilian identification: the image does not clearly show a teleosauroid-specific body plan, so calling it a "teleosauroid crocodylomorph" is more specific than the illustration supports. 2) Because of that, the caption’s claim that the animal’s "narrow tooth-lined snout, armored back, and semi-aquatic build" are visible diagnostic teleosauroid traits is not fully supported by the image. 3) One reviewer noted that teleosauroids are more strongly associated with the Tethyan/Laurasian realm than Gondwana, so the Gondwanan framing may be a geographic stretch if the caption intends a strict teleosauroid example; a Gondwanan marine crocodylomorph would be safer if specificity is desired. 4) The caption’s xiphosuran reference is not erroneous, but if the image is meant to depict horseshoe-crab-like forms, that should be handled carefully to avoid implying a particular morphology or taxon beyond what is visible. 5) The flora list (conifers, cycads, ferns) is broadly correct, but the image does not clearly distinguish these plant groups, so the caption slightly over-specifies the botanical composition relative to what can be seen.
Final verdict: adjust for both image and caption. The setting, climate, and general Jurassic coastal ecology are sound, but the crocodylomorph needs to be made more clearly teleosauroid in the image, and the caption should be toned down where it identifies the animal too specifically and where it attributes diagnostic traits not clearly visible. The rest of the scene is broadly usable and does not require regeneration.
Other languages
- English: Teleosauroid crocodylomorph ambush in Jurassic tidal inlet
- Français: Embuscade de crocodylomorphe téléosauroïde dans un bras de mer
- Español: Emboscada de crocodilomorfo teleosauroide en entrada de marea
- Português: Emboscada de crocodilomorfo teleossauroide em entrada de maré
- Deutsch: Teleosauroider Crocodylomorph-Hinterhalt in einer Gezeitenmündung
- العربية: كمين تمساح تيليوصوري في مدخل مد وجزر جوراسي
- हिन्दी: जुरासिक ज्वारीय इनलेट में टेलीोसॉरोइड मगरमच्छ घात
- 한국어: 쥐라기 조수 유입구의 텔레오사우루스류 악어 매복
- Italiano: Agguato di coccodrillomorfo teleosauroide in un'insenatura
- Nederlands: Teleosauroïde crocodylomorf in een Jurassische getijdengeul
However, there are notable mismatches. The flying animals overhead are shaped like pterosaurs, but their silhouettes and sizes are inconsistent with a specific Late Jurassic Gondwanan assemblage (and one or more may resemble more derived/incorrect proportions). Most importantly, the small, helmet-like arthropods on the beach resemble modern horseshoe crabs (or at least are drawn very close to that look). While Jurassic marine arthropods existed, a horseshoe-crab-like appearance is anachronistic/over-specific for a Gondwana tidal flat unless the prompt explicitly targets a correct Jurassic xiphosuran morphology. Additionally, the foreground crocodilian has a very bulky, wide-bodied profile and prominent “armored” look; teleosauroids were marine crocodylomorphs with robust armor in some taxa, but the depiction reads more like a generic large croc than a clearly teleosauroid-specific form (and the “half-submerged wait” pose is plausible but not strongly diagnostic from the image alone).
The caption is generally on the right track: late Jurassic timing (163–145 Ma), Gondwana context, warm greenhouse/coastal rift-basin framing, and ecological behavior (ambush in a tidal inlet) are all reasonable. But it makes several claims that are stronger than what the image necessarily supports: (1) it identifies “teleosauroid crocodylomorph” while the visual features are not clearly teleosauroid-diagnostic (narrowness of the snout, typical cranial proportions, and dorsal armor patterning are not clearly conveyed), (2) “xiphosurans” are implied by the caption but the depicted beach animals look more like horseshoe crabs, which may be visually accurate only at a very broad level—species-level or morphology-level accuracy is uncertain, and (3) the flora listed (conifers, cycads, ferns) is plausible for Gondwana, but the image’s vegetation does not clearly separate these groups (most looks like mixed greenery with a few tall, palm/cycad-like forms). Overall, the scene is believable, but the specific taxonomic identifications and the horseshoe-crab-like implication need tightening, hence an “adjust” rather than “approve.”