後期ジュラ紀（約1億630万〜1億450万年前）、テチス海の影響を受けたローラシア西部ヨーロッパの浅い炭酸塩海では、**Thecosmilia**、**Isastrea**、**Thamnasteria** などのイシサンゴ類が低いパッチ礁を築き、淡い石灰質の海底に複雑な生息場を生み出していました。画面には、陽光きらめく数メートルほどの遠浅の海を、枝状や塊状のサンゴ群体の間で **Glyphea**（グリフェア類のロブスター）、棘の長いウニ **Cidaris**、海流に腕を広げるウミユリ **Pentacrinites**、円盤形の硬鱗魚 **Dapedium** の群れが行き交う様子が広がります。現代のサンゴ礁に似て見えても、そこを形づくっていた生物相は大きく異なり、この温暖で澄んだ海は、恐竜時代のヨーロッパが広い浅海の島々に縁どられていたことを物語っています。
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 image captures a broadly plausible Late Jurassic shallow carbonate shelf environment, with patch reef structure, pale carbonate substrate, crinoids (the branching stalked forms visible in the center are a genuine positive), cidaroid-type sea urchins with prominent spines, lobster-like crustaceans, and schooling fish. The low-relief patch reef structure and pale sandy carbonate floor are consistent with Jurassic platform settings. However, the corals depicted have morphologies that lean heavily toward modern reef coral appearances—rounded massive and plate forms that visually resemble Porites or Acropora rather than the rugose or distinctly Jurassic scleractinian morphologies of Thecosmilia, Isastrea, or Thamnasteria. A more Jurassic-accurate depiction would show corals with more irregular, thicker, less 'clean' branching and more massive-lumpy growth forms. The fish are generic and look modern—deep-bodied laterally compressed forms consistent with modern reef fish rather than clearly Jurassic actinopterygian (e.g., Dapedium-like) morphology with more rhomboid scales and less symmetrical fins. The lobster-like crustaceans in the foreground are reasonably consistent with palinurid or glypheoid decapods, though they appear quite modern-looking. The background island with vegetation is a potential issue—while land would exist, the vegetation silhouettes look suspiciously modern (possibly flowering-plant-era trees), whereas Late Jurassic Laurasian flora would have been dominated by conifers, cycads, and ferns with no angiosperms.
Grok
画像:
再生成
キャプション:
承認
Apr 2, 2026
The caption is factually accurate and well-suited to the Late Jurassic period in Laurasia. The described taxa—Thecosmilia, Isastrea, and Thamnasteria—are genuine Late Jurassic scleractinian corals from European Tethyan settings, Pentacrinites is a correct stalked crinoid, cidaroid urchins match the period, Glyphea is an appropriate lobster-like decapod, and Dapedium is a valid deep-bodied fish from shallow Jurassic seas. The setting of warm, shallow shelf seas on a carbonate platform around 160–150 Ma (Oxfordian-Kimmeridgian) aligns with paleogeography, and the note on pre-modern reef communities is appropriately contextual without misleading claims. Detail level is educational and precise, warranting approval.
The image, however, is a significant anachronism, depicting a modern tropical coral reef with branching Acropora-like stagshorn corals, massive Porites-like heads, fan corals, and schools of laterally compressed angelfish/parrotfish that evoke present-day Indo-Pacific reefs rather than low-relief Jurassic patch reefs. Jurassic scleractinians had thicker, more massive/irregular forms without the delicate branching or plating seen here; no clear stalked crinoids (columnar stems with arms), cidaroid urchins (long-spined but not prominently shown), lobsters in crevices, or Dapedium-like fish are visible—instead, generic modern reef fish dominate. The pale carbonate seafloor is somewhat present but overwhelmed by vibrant, high-relief structures and turquoise waters that feel Caribbean-contemporary. Background distant islands with hazy vegetation silhouettes suggest angiosperm-like trees, absent in Jurassic Laurasia (which had conifers/cycads). This lacks visual coherence with Jurassic biology/geology, requiring full regenerate for taxon-specific morphologies like lumpy corals, evident crinoid stalks, and period-appropriate fish.
Agreeing with GPT and Claude: corals are too modern (e.g., Acropora/Porites resemblances), fish generic/not Dapedium-like, crinoids not clearly columnar. Claude correctly flags vegetation anachronism, which I also note; GPT's concern on genus co-occurrence is overly cautious as these corals did overlap in Tethyan reefs (e.g., Solnhofen Limestone equivalents). Neither fully captured the total modern reef aesthetic takeover, but their 'adjust' votes underrate the scale—regenerate is needed over refinement.
The image, however, is a significant anachronism, depicting a modern tropical coral reef with branching Acropora-like stagshorn corals, massive Porites-like heads, fan corals, and schools of laterally compressed angelfish/parrotfish that evoke present-day Indo-Pacific reefs rather than low-relief Jurassic patch reefs. Jurassic scleractinians had thicker, more massive/irregular forms without the delicate branching or plating seen here; no clear stalked crinoids (columnar stems with arms), cidaroid urchins (long-spined but not prominently shown), lobsters in crevices, or Dapedium-like fish are visible—instead, generic modern reef fish dominate. The pale carbonate seafloor is somewhat present but overwhelmed by vibrant, high-relief structures and turquoise waters that feel Caribbean-contemporary. Background distant islands with hazy vegetation silhouettes suggest angiosperm-like trees, absent in Jurassic Laurasia (which had conifers/cycads). This lacks visual coherence with Jurassic biology/geology, requiring full regenerate for taxon-specific morphologies like lumpy corals, evident crinoid stalks, and period-appropriate fish.
Agreeing with GPT and Claude: corals are too modern (e.g., Acropora/Porites resemblances), fish generic/not Dapedium-like, crinoids not clearly columnar. Claude correctly flags vegetation anachronism, which I also note; GPT's concern on genus co-occurrence is overly cautious as these corals did overlap in Tethyan reefs (e.g., Solnhofen Limestone equivalents). Neither fully captured the total modern reef aesthetic takeover, but their 'adjust' votes underrate the scale—regenerate is needed over refinement.
Gemini
画像:
再生成
キャプション:
調整
Apr 2, 2026
The image requires a complete regeneration. My colleagues correctly identified the 'modern reef' aesthetic, which is the primary failure here. The corals depicted show delicate, branching, and plating morphologies typical of Cenozoic reefs, whereas Late Jurassic Tethyan scleractinians (like Thecosmilia or Isastrea) often exhibited massive, meandroid, or thicker, more irregular branching patterns. Furthermore, the fish silhouettes are indistinguishable from modern Perciformes; they lack the specific rhombic scales and fin placement characteristic of Dapedium or other Jurassic actinopterygians. The inclusion of angiosperm-like vegetation on the horizon is a significant anachronism that must be replaced with a flora of conifers, cycads, and bennettitales. The caption is scientifically sound in its taxonomic choices, but I agree with the previous reviewers that it leans too heavily into specific generic lists that might imply a level of ecological uniformity that is hard to verify. I suggest adjusting the caption to describe the 'assemblage types' rather than listing specific genera to ensure it remains robust against changing paleontological consensus. I largely agree with Grok's assessment that the image is a generic 'modern reef' placeholder that fails to capture the unique, low-relief character of Jurassic carbonate platforms.
Matania
要約
画像:
再生成
キャプション:
調整
Paragraph 1: The committee agrees the scene is broadly set in a warm, shallow carbonate-platform sea with a pale seafloor, patch reefs, corals, crinoid-like forms, cidaroid-like sea urchins, lobsters/crustaceans, and schooling fish—so the high-level Jurassic reef concept is directionally plausible. The caption also correctly aims for a Late Jurassic Laurasian/Tethyan reef setting and includes several taxon groups that are appropriate in principle.
Paragraph 2: IMAGE issues identified by one or more reviewers: 1) The overall image reads as a modern tropical coral reef rather than a Jurassic low-relief patch reef. 2) Coral morphologies are too modern, resembling Acropora/Porites-like branching, plating, and massive forms instead of Jurassic scleractinian forms; reviewers specifically noted the need for thicker, more irregular, less 'clean' growth forms. 3) The reef looks too high-relief and vibrant/Caribbean-like for a Jurassic carbonate platform. 4) Fish silhouettes are generic modern reef fish; they do not read as Jurassic actinopterygians and are not convincingly Dapedium-like. 5) Dapedium-specific traits are missing, especially rhomboid scales and more appropriate fin placement/body silhouette. 6) Crinoids are not clearly represented as stalked Pentacrinites; the expected columnar stems and feeding crowns are not evident enough, despite some branching stalked forms. 7) The echinoderms are broadly stylized and modern-looking rather than convincingly Jurassic/cidaroid. 8) Lobster-like crustaceans appear modern and generic rather than clearly glypheoid/period-appropriate. 9) The background islands and vegetation look suspiciously modern, possibly angiosperm-era trees, which is an anachronism for Late Jurassic Laurasia. 10) The color/lighting/water clarity and the polished reef aesthetic feel too contemporary and should be less like an Indo-Pacific scene. 11) The image composition looks like a composite of modern reef motifs rather than a defensible Jurassic assemblage with taxon-specific morphology.
Paragraph 3: CAPTION issues identified by one or more reviewers: 1) The caption is too taxonomically specific and presents multiple genera as if their exact co-occurrence in one European Laurasian reef is fully established without citation. 2) The use of Thecosmilia, Isastrea, and Thamnasteria may be plausible in a general Late Jurassic Tethyan reef context, but the assemblage claim is overconfident as written. 3) Isastrea was flagged as needing careful verification for Late Jurassic Europe/Laurasia suitability. 4) Pentacrinites is acceptable as a stalked crinoid genus, but the image does not clearly show the diagnostic stalked form, making the caption-image match weak. 5) The phrase 'deep-bodied Dapedium' is likely inaccurate or at least poorly supported; Dapedium was noted as a valid Jurassic fish, but its ecology/placement in sunlit shallow reef shoals is questionable. 6) The caption’s fish/ecology claim feels stronger than the evidence supports; Jurassic reef fish assemblages were likely different from the one implied. 7) The statement 'before modern reef communities had evolved' is acceptable in concept, but it can imply a too-simple evolutionary progression; better to emphasize that Jurassic reef structure and biota differed from modern reefs. 8) Because the image is visually modern, the caption and image are in tension; the caption describes a more defensible Jurassic assemblage than what is actually shown.
Paragraph 4: Final verdict: the image must be regenerated, not merely adjusted, because the modern-reef aesthetic, anachronistic vegetation, and generic fish/coral/crustacean forms are pervasive and require a full scene rebuild with Jurassic-appropriate morphology. The caption only needs adjustment, not regeneration: its broad Jurassic framing is sound, but it should be softened to avoid overconfident genus-level co-occurrence claims and to better match uncertain ecological details.
Paragraph 2: IMAGE issues identified by one or more reviewers: 1) The overall image reads as a modern tropical coral reef rather than a Jurassic low-relief patch reef. 2) Coral morphologies are too modern, resembling Acropora/Porites-like branching, plating, and massive forms instead of Jurassic scleractinian forms; reviewers specifically noted the need for thicker, more irregular, less 'clean' growth forms. 3) The reef looks too high-relief and vibrant/Caribbean-like for a Jurassic carbonate platform. 4) Fish silhouettes are generic modern reef fish; they do not read as Jurassic actinopterygians and are not convincingly Dapedium-like. 5) Dapedium-specific traits are missing, especially rhomboid scales and more appropriate fin placement/body silhouette. 6) Crinoids are not clearly represented as stalked Pentacrinites; the expected columnar stems and feeding crowns are not evident enough, despite some branching stalked forms. 7) The echinoderms are broadly stylized and modern-looking rather than convincingly Jurassic/cidaroid. 8) Lobster-like crustaceans appear modern and generic rather than clearly glypheoid/period-appropriate. 9) The background islands and vegetation look suspiciously modern, possibly angiosperm-era trees, which is an anachronism for Late Jurassic Laurasia. 10) The color/lighting/water clarity and the polished reef aesthetic feel too contemporary and should be less like an Indo-Pacific scene. 11) The image composition looks like a composite of modern reef motifs rather than a defensible Jurassic assemblage with taxon-specific morphology.
Paragraph 3: CAPTION issues identified by one or more reviewers: 1) The caption is too taxonomically specific and presents multiple genera as if their exact co-occurrence in one European Laurasian reef is fully established without citation. 2) The use of Thecosmilia, Isastrea, and Thamnasteria may be plausible in a general Late Jurassic Tethyan reef context, but the assemblage claim is overconfident as written. 3) Isastrea was flagged as needing careful verification for Late Jurassic Europe/Laurasia suitability. 4) Pentacrinites is acceptable as a stalked crinoid genus, but the image does not clearly show the diagnostic stalked form, making the caption-image match weak. 5) The phrase 'deep-bodied Dapedium' is likely inaccurate or at least poorly supported; Dapedium was noted as a valid Jurassic fish, but its ecology/placement in sunlit shallow reef shoals is questionable. 6) The caption’s fish/ecology claim feels stronger than the evidence supports; Jurassic reef fish assemblages were likely different from the one implied. 7) The statement 'before modern reef communities had evolved' is acceptable in concept, but it can imply a too-simple evolutionary progression; better to emphasize that Jurassic reef structure and biota differed from modern reefs. 8) Because the image is visually modern, the caption and image are in tension; the caption describes a more defensible Jurassic assemblage than what is actually shown.
Paragraph 4: Final verdict: the image must be regenerated, not merely adjusted, because the modern-reef aesthetic, anachronistic vegetation, and generic fish/coral/crustacean forms are pervasive and require a full scene rebuild with Jurassic-appropriate morphology. The caption only needs adjustment, not regeneration: its broad Jurassic framing is sound, but it should be softened to avoid overconfident genus-level co-occurrence claims and to better match uncertain ecological details.
Other languages
- English: Jurassic scleractinian coral reef with stalked crinoids
- Français: Récif de coraux scléractiniaires et crinoïdes du Jurassique
- Español: Arrecife de corales escleractinios y crinoideos del Jurásico
- Português: Recife de corais escleractínios e crinoides do Jurássico
- Deutsch: Korallenriff mit gestielten Seelilien im jurassischen Meer
- العربية: شعاب مرجانية صلبة مع زنابق البحر في العصر الجوراسي
- हिन्दी: जुरासिक काल की प्रवाल भित्तियाँ और डंठल वाले क्रिनोइड्स
- 한국어: 쥐라기 경산호 산호초와 자루가 있는 바다나리
- Italiano: Barriera corallina giurassica con crinoidi peduncolati
- Nederlands: Jurassisch koraalrif met gesteelde zeelelies
However, several depicted biota look like generalized “modern reef” morphotypes. The corals shown include many branching/encrusting forms with high resemblance to modern reef-building coral silhouettes, and the sea urchins/crinoid-like elements are stylized in ways that are hard to tie to specific Late Jurassic Laurasian taxa (e.g., stalked Pentacrinites-like forms are not clearly represented as columnar crinoids; the echinoderm appearance is broadly modern-looking). Also, the fish silhouettes are generic and not clearly consistent with specific Jurassic actinopterygian groups; while fish are certainly present, the scene feels more like a composite reef than a defensible “named-taxa” Jurassic assemblage. These issues suggest the image could be refined with more taxon-specific morphology (e.g., clearer crinoid stems/arms, Jurassic-relevant coral morphologies, and more appropriate reef-associated crustacean forms).