쥐라기 후기, 약 1억 6,000만~1억 4,500만 년 전 곤드와나 동부의 마다가스카르 외해 대륙붕에서는 거대한 눈을 지닌 안구어룡류(ophthalmosaurid) 어룡이 푸른빛 바다를 가르며 사냥에 나섰습니다. 관람객은 길이 약 4 m의 유선형 어룡이 베렘노프시스(Belemnopsis) 무리를 추격하는 사이, 규칙적인 늑골 장식을 지닌 페리스핑크테스(Perisphinctes)와 더 매끈하고 둥근 필로세라스(Phylloceras) 암모나이트가 물기둥 속에 떠다니는 장면을 보게 됩니다. 이 바다는 리프트의 영향을 받던 곤드와나 대륙 가장자리의 얕은 바다 바깥쪽 선반 환경으로, 두족류가 특히 풍부한 생태계였습니다. 햇빛이 스며드는 물속에서 번쩍이는 어룡의 커다란 눈은, 공룡 시대 바다 또한 치열한 추격과 생존의 무대였음을 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 captures a convincing Late Jurassic marine shelf scene with several well-executed elements. The central ichthyosaur has a good tuna-shaped body plan, visible large eye (consistent with ophthalmosaurid anatomy), flippers, and a heterocercal-style tail — all appropriate for the group. The coiled ammonites are varied in form, with some showing ribbing (plausibly Perisphinctes-type) and some smoother (plausibly Phylloceras-type), which aligns with the caption's claims. Small fish in the background are reasonable for a Jurassic shelf setting. However, the image lacks any identifiable belemnites (elongated dart-shaped guards or squid-like forms), which the caption specifically names as 'Belemnopsis.' The cephalopods with visible tentacles and nautilus-like shells appear to be nautiloids rather than belemnoids, making that part of the scene scientifically inconsistent with the caption.
Grok
이미지:
조정
캡션:
조정
Apr 2, 2026
The image depicts a plausible Late Jurassic marine shelf scene with a central ophthalmosaurid ichthyosaur showing tuna-like proportions, large eyes, elongated snout, flippers, and a tail appropriate for fast pursuit predators—consistent with genera like Brachypterygius or Sveltonectes from Late Jurassic deposits. Coiled ammonite shells with varied ribbing (ribbed ones evoking Perisphinctes, smoother ones like Phylloceras) and small schooling fish add ecological coherence for a Gondwanan shelf sea. The blue-green water, sandy seafloor, and open-water composition fit a shallow outer shelf environment near Madagascar's margin. No anachronisms like modern boats or fish are present, and visual plausibility is high. However, putative belemnites (elongated, dart-like forms or squid-like bodies labeled as scattering) are ambiguous—many resemble nautiloids or soft-bodied coleoids rather than distinct Belemnopsis guards with typical bullet-shaped rostra, making them hard to identify specifically. Minor adjustments could refine shell shapes and behaviors for better taxonomic match and realism (e.g., less static floating of large ammonites).
The caption is factually strong on timeframe (Late Jurassic, 163–145 Ma aligns with Oxfordian-Kimmeridgian), Gondwana-Madagascar context (accurate, as rifting intensified later), and ecosystem (cephalopod-rich shelves with ichthyosaurs). Perisphinctes and Phylloceras are cosmopolitan Late Jurassic ammonites with Gondwanan records; ophthalmosaurids were indeed fast, open-water predators; Belemnopsis is attested in Jurassic Gondwana (e.g., Australia, possibly Madagascar margins). The phrasing 'before the rise of modern marine vertebrates' is acceptable as interpretive, noting dominance shifts post-Jurassic. Issues are minor: Belemnopsis presence on exact Madagascar shelves is plausible but not richly documented (stronger in other Gondwanan sites), and image-caption mismatch on belemnite visuals risks misleading. Adjust for precise sourcing or generalization.
Agreeing with GPT and Claude on core strengths (ichthyosaur anatomy, general scene coherence) and main flaw (lack of clear Belemnopsis identification—image shows ambiguous coleoid/nautiloid forms, not diagnostic guards). GPT correctly flags vague 'modern vertebrates' phrasing (teleosts existed but weren't dominant); Claude rightly notes nautiloid-like elements over belemnites. Neither missed much, but both underplay strong Gondwanan fit of named taxa (e.g., Perisphinctes in Indian plates); image's fish are apt lepisosteid-like forms for Jurassic shelves, enhancing accuracy.
The caption is factually strong on timeframe (Late Jurassic, 163–145 Ma aligns with Oxfordian-Kimmeridgian), Gondwana-Madagascar context (accurate, as rifting intensified later), and ecosystem (cephalopod-rich shelves with ichthyosaurs). Perisphinctes and Phylloceras are cosmopolitan Late Jurassic ammonites with Gondwanan records; ophthalmosaurids were indeed fast, open-water predators; Belemnopsis is attested in Jurassic Gondwana (e.g., Australia, possibly Madagascar margins). The phrasing 'before the rise of modern marine vertebrates' is acceptable as interpretive, noting dominance shifts post-Jurassic. Issues are minor: Belemnopsis presence on exact Madagascar shelves is plausible but not richly documented (stronger in other Gondwanan sites), and image-caption mismatch on belemnite visuals risks misleading. Adjust for precise sourcing or generalization.
Agreeing with GPT and Claude on core strengths (ichthyosaur anatomy, general scene coherence) and main flaw (lack of clear Belemnopsis identification—image shows ambiguous coleoid/nautiloid forms, not diagnostic guards). GPT correctly flags vague 'modern vertebrates' phrasing (teleosts existed but weren't dominant); Claude rightly notes nautiloid-like elements over belemnites. Neither missed much, but both underplay strong Gondwanan fit of named taxa (e.g., Perisphinctes in Indian plates); image's fish are apt lepisosteid-like forms for Jurassic shelves, enhancing accuracy.
Gemini
이미지:
조정
캡션:
조정
Apr 2, 2026
The image succeeds in establishing a plausible Late Jurassic marine atmosphere, with the ophthalmosaurid ichthyosaur demonstrating appropriate, high-speed, pelagic morphology. The varied ammonite morphologies are a strong touch for biodiversity, but as my colleagues correctly noted, the 'Belemnopsis' are visually absent or indistinguishable from nautiloid-like soft-bodied forms. Belemnites should exhibit a distinct, elongated, bullet-shaped rostrum (guard) and a specific tentacle arrangement, which is not present here. The scene is visually coherent, but the lack of diagnostic belemnite features creates a disconnect between the visual evidence and the scientific claims in the caption.
Regarding the caption, I agree with the previous reviewers that the claim regarding the 'rise of modern marine vertebrates' is imprecise and potentially misleading; modern teleost lineages were already established, even if their ecological dominance was limited. While the inclusion of Perisphinctes and Phylloceras is scientifically defensible for the region and period, the caption should be tempered to reflect that these are representative of the broader Gondwanan margin rather than implying an exclusive, documented inventory for the specific Madagascar margin. I agree with the committee's consensus: the ichthyosaur is well-rendered, but the cephalopod diversity needs to be corrected to visually include distinct belemnite anatomy, and the caption requires a more nuanced tone regarding vertebrate evolution.
Regarding the caption, I agree with the previous reviewers that the claim regarding the 'rise of modern marine vertebrates' is imprecise and potentially misleading; modern teleost lineages were already established, even if their ecological dominance was limited. While the inclusion of Perisphinctes and Phylloceras is scientifically defensible for the region and period, the caption should be tempered to reflect that these are representative of the broader Gondwanan margin rather than implying an exclusive, documented inventory for the specific Madagascar margin. I agree with the committee's consensus: the ichthyosaur is well-rendered, but the cephalopod diversity needs to be corrected to visually include distinct belemnite anatomy, and the caption requires a more nuanced tone regarding vertebrate evolution.
Matania
요약
이미지:
조정
캡션:
조정
The committee agrees that the overall scene is broadly accurate for a Late Jurassic Gondwanan marine shelf: the water column, open-marine composition, and the central ichthyosaur’s tuna-like, fast-pursuit body plan are convincing. The coiled ammonite-like shells also generally fit a Jurassic cephalopod-rich ecosystem, and there are no obvious modern intrusions or other blatant anachronisms in the image.
Image issues identified by the committee: 1) the putative belemnites are not visually diagnostic of Belemnopsis or belemnites at all; they read as nautiloid-like or generic soft-bodied cephalopods rather than showing distinct elongated bullet-shaped rostra/guards; 2) the scene contains multiple large ammonite shells that look overly static, floating, and somewhat collection-like rather than behaving like a believable assemblage in water; 3) the ammonites are somewhat idealized/interchangeable rather than clearly tied to specific Late Jurassic Gondwanan taxa; 4) the ichthyosaur, while generally correct, is stylized in a broad modern-fish way and is not anatomically specific enough to confidently verify an ophthalmosaurid; 5) the image-caption match is weak because the image does not clearly support the caption’s specific Belemnopsis claim; 6) the reviewers noted minor realism issues in the cephalopod forms and their buoyancy/behavior, though not enough to make the scene broadly implausible.
Caption issues identified by the committee: 1) the specific taxonomic identification of Belemnopsis is unsupported by the image, because no diagnostic belemnite rostra/guards are visible; 2) the caption overreaches by naming Perisphinctes, Phylloceras, and Belemnopsis as though they are directly evidenced by the visual, when the image only shows generalized ammonite- and cephalopod-like forms; 3) the phrase “before the rise of modern marine vertebrates” is misleading/imprecise, because modern teleost lineages were already present in the Jurassic even if marine ecosystems were organized differently; 4) the wording implies a level of locality-specific certainty for the Madagascar margin that is stronger than what the image can justify; 5) the caption’s taxonomic and regional specificity should be tempered to reflect broader Late Jurassic Gondwanan shelf faunas unless a sourced reconstruction is intended.
Final verdict: adjust for both image and caption. The scene is scientifically plausible in broad strokes, but the image does not clearly depict the belemnites named in the caption, and the caption uses taxonomic and evolutionary language that is too specific or too strong relative to what is visibly supported. The committee therefore recommends tightening the visual cephalopod anatomy and softening the caption’s taxonomic claims and evolutionary phrasing rather than regenerating the whole piece.
Image issues identified by the committee: 1) the putative belemnites are not visually diagnostic of Belemnopsis or belemnites at all; they read as nautiloid-like or generic soft-bodied cephalopods rather than showing distinct elongated bullet-shaped rostra/guards; 2) the scene contains multiple large ammonite shells that look overly static, floating, and somewhat collection-like rather than behaving like a believable assemblage in water; 3) the ammonites are somewhat idealized/interchangeable rather than clearly tied to specific Late Jurassic Gondwanan taxa; 4) the ichthyosaur, while generally correct, is stylized in a broad modern-fish way and is not anatomically specific enough to confidently verify an ophthalmosaurid; 5) the image-caption match is weak because the image does not clearly support the caption’s specific Belemnopsis claim; 6) the reviewers noted minor realism issues in the cephalopod forms and their buoyancy/behavior, though not enough to make the scene broadly implausible.
Caption issues identified by the committee: 1) the specific taxonomic identification of Belemnopsis is unsupported by the image, because no diagnostic belemnite rostra/guards are visible; 2) the caption overreaches by naming Perisphinctes, Phylloceras, and Belemnopsis as though they are directly evidenced by the visual, when the image only shows generalized ammonite- and cephalopod-like forms; 3) the phrase “before the rise of modern marine vertebrates” is misleading/imprecise, because modern teleost lineages were already present in the Jurassic even if marine ecosystems were organized differently; 4) the wording implies a level of locality-specific certainty for the Madagascar margin that is stronger than what the image can justify; 5) the caption’s taxonomic and regional specificity should be tempered to reflect broader Late Jurassic Gondwanan shelf faunas unless a sourced reconstruction is intended.
Final verdict: adjust for both image and caption. The scene is scientifically plausible in broad strokes, but the image does not clearly depict the belemnites named in the caption, and the caption uses taxonomic and evolutionary language that is too specific or too strong relative to what is visibly supported. The committee therefore recommends tightening the visual cephalopod anatomy and softening the caption’s taxonomic claims and evolutionary phrasing rather than regenerating the whole piece.
Other languages
- English: Ichthyosaur and Perisphinctes ammonites off Madagascar coast
- Français: Ichthyosaure et ammonites Perisphinctes au large de Madagascar
- Español: Ictiosaurio y amonites Perisphinctes frente a Madagascar
- Português: Ictiossauro e amonites Perisphinctes na costa de Madagascar
- Deutsch: Ichthyosaurier und Perisphinctes-Ammoniten vor Madagaskar
- العربية: إكتيوصور وأمونيت بيريسفينكتس قبالة سواحل مدغشقر
- हिन्दी: मेडागास्कर तट के पास इक्थियोसौर और पेरिस्फ़िंक्टेस अमोनाइट्स
- 日本語: マダガスカル沖のイクチオサウルスとアンモナイト
- Italiano: Ittiosauro e ammoniti Perisphinctes al largo del Madagascar
- Nederlands: Ichthyosaurus en Perisphinctes ammonieten voor de kust van Madagaskar
However, the depicted cephalopods and ichthyosaur anatomy look somewhat idealized and interchangeable rather than closely tied to specific, Late Jurassic Gondwanan taxa. In particular, several ammonite-like shells appear very large and some are shown floating near the surface without much realism in buoyancy/behavior (while not impossible, it reads more like a “collection” than a specific predation event). The “Belemnopsis” in the caption cannot be confirmed from the visuals because the shell shapes in the image are not clearly elongated belemnite guards versus coiled ammonite shells. The ichthyosaurs are also drawn with broad, modern fish-camouflage styling; ophthalmosaurids are generally believable, but the exact look (eye size/position, jaw proportions) isn’t specific enough to verify an ophthalmosaurid, and the overall targeting focus is unclear.
For the caption, the broad timeframe (Late Jurassic) and Gondwana shelf context are reasonable. Madagascar being part of Gondwana in the Late Jurassic is broadly correct. The species-level claims are the weak point: naming “Perisphinctes” and “Phylloceras,” and explicitly “Belemnopsis,” requires confidence that these exact genera match the Madagascar margin setting for 163–145 Ma; the caption doesn’t justify or show clear evidence, and several are not unambiguously supported by what’s visibly present. Also, “before the rise of modern marine vertebrates” is a vague and slightly misleading phrasing—modern-style lineages (e.g., teleosts) were already present by the Jurassic, even if marine “megafauna” ecosystems differ from today. Overall, the scene concept is good, but the caption overreaches in taxonomic specificity and uses a generalization about “modern” vertebrates.