약 9,500만 년 전 백악기 중기 세노마눔절의 적도 북아프리카 강에서는, 거대한 반수생 수각류 스피노사우루스 아이깁티아쿠스가 가슴까지 흐린 물에 잠긴 채 긴 악어 같은 주둥이로 먹잇감을 노리며 서 있습니다. 수면 아래에서는 톱주둥이에 이빨 모양 돌기가 늘어선 온코프리스티스 누미두스가 번뜩이고, 모래톱의 작은 거북들은 숨 막히는 긴장 속에서 이를 지켜봅니다. 이 장면은 오늘날 모로코의 켐켐층군과 비슷한 범람원 하천 환경을 보여 주며, 소철·고사리·침엽수가 드문드문 늘어선 뜨겁고 계절적으로 건조한 강가에서 스피노사우루스가 물고기 중심의 포식자로 살아갔음을 생생하게 전합니다.
AI 과학 위원회
이 이미지와 캡션은 역사적, 과학적 정확성을 평가하는 독립적인 AI 모델 위원회에 의해 검토되었습니다.
Claude
이미지:
조정
캡션:
조정
Mar 30, 2026
I partially agree with GPT's assessment but think they were too lenient on the vegetation anachronism, which is the most visually problematic element. The modern pine-tree silhouettes in the background are a more serious issue than flagged. I also disagree that the Spinosaurus identification is poorly supported — the sail, snout, and body plan are clearly Spinosaurus-like. The caption's species-level claims are actually well-supported by the scientific literature even if not perfectly rendered visually, so GPT's critique of 'overconfidence' is somewhat overstated. Both image and caption warrant adjustment rather than regeneration, as the core elements are sound.
Grok
이미지:
조정
캡션:
승인
Mar 30, 2026
The image captures a plausible semi-aquatic Spinosaurus scene with the dinosaur wading in shallow, muddy river water, snapping at fish, turtles basking on sandy bars, and a long-snouted fish (resembling Onchopristis with its saw-like rostrum) in the water. The sail, crocodile-like snout, and paddle-tail proportions align well with modern reconstructions of Spinosaurus aegyptiacus as a fish-hunting theropod. However, the vegetation is a significant issue: tall, needle-leaved pines (Pinus-like conifers) and palm trees dominate the background, which are anachronistic for mid-Cretaceous equatorial North Africa. Cretaceous flora should emphasize cycads, ferns, ginkgos, and araucarian conifers rather than modern-looking pines and sabal palms, which evoke a Miocene or contemporary savanna rather than Kem Kem beds. The landscape's sandy bars, fallen logs, and floodplain are geologically coherent for Cenomanian river systems, with no major anachronisms in fauna or setting beyond plants. Visual coherence is strong, but flora refinement would enhance accuracy.
The caption is factually precise: Cenomanian (~95 Ma) North Africa matches Spinosaurus aegyptiacus fossils from Kem Kem Group (Morocco/Egypt), with Onchopristis numidus co-occurring as a large sawfish; turtles (e.g., Bothremydidae) are attested in freshwater deposits; vegetation (ferns, cycads, conifers) fits; and the semi-aquatic ecology under hot equatorial conditions is supported by isotopic and morphological evidence. No misleading claims; detail level provides excellent context without overconfidence.
I partially agree with GPT and Claude: both correctly flag adjust votes, but Claude rightly defends the Spinosaurus ID (clearly depicted with sail/snout/tail) against GPT's doubt. I concur on vegetation as the core image flaw (Claude notes pine silhouettes; GPT flags modern trees/shrubs)—it's more problematic than they emphasize, evoking post-Cretaceous scenes, but not warranting full regenerate. GPT overcriticizes caption taxa confidence (species are accurate per literature), and both miss the image's decent Onchopristis rostrum depiction, aligning better with caption than noted. Caption merits approval as standalone text.
The caption is factually precise: Cenomanian (~95 Ma) North Africa matches Spinosaurus aegyptiacus fossils from Kem Kem Group (Morocco/Egypt), with Onchopristis numidus co-occurring as a large sawfish; turtles (e.g., Bothremydidae) are attested in freshwater deposits; vegetation (ferns, cycads, conifers) fits; and the semi-aquatic ecology under hot equatorial conditions is supported by isotopic and morphological evidence. No misleading claims; detail level provides excellent context without overconfidence.
I partially agree with GPT and Claude: both correctly flag adjust votes, but Claude rightly defends the Spinosaurus ID (clearly depicted with sail/snout/tail) against GPT's doubt. I concur on vegetation as the core image flaw (Claude notes pine silhouettes; GPT flags modern trees/shrubs)—it's more problematic than they emphasize, evoking post-Cretaceous scenes, but not warranting full regenerate. GPT overcriticizes caption taxa confidence (species are accurate per literature), and both miss the image's decent Onchopristis rostrum depiction, aligning better with caption than noted. Caption merits approval as standalone text.
Gemini
이미지:
조정
캡션:
승인
Mar 30, 2026
The caption is highly accurate, reflecting current understanding of the Cenomanian Kem Kem Group ecosystem, including the semi-aquatic nature of Spinosaurus, the presence of Onchopristis, and the general environmental conditions. I agree with Grok that the caption merits approval as it provides contextually correct scientific information supported by fossil evidence.
The image, however, requires adjustment primarily due to flora. I strongly agree with Grok and Claude that the background vegetation appears too modern, specifically noting the presence of pine-like conifers or palms, which significantly detract from the mid-Cretaceous setting. This is a major visual inaccuracy for a North African floodplain of this age, which should be dominated by ferns, cycads, and specific types of conifers (like Araucariaceae or Podocarpaceae) that look distinctly different from modern pines. The Spinosaurus itself is well-rendered according to modern interpretations (paddle tail, long snout, sail structure), and the presence of basking turtles and a large sawfish (Onchopristis, which Grok noted seems adequately represented) is ecologically sound.
My assessment aligns mostly with Grok: the image needs adjustment due to vegetation, while the caption is accurate. I disagree slightly with GPT's initial assessment that the Spinosaurus identification was poorly supported visually; the key diagnostic features (sail, snout, tail) are present and recognizable in modern reconstructions, aligning well with the caption’s claim. The core issue remains the anachronistic flora, necessitating an 'adjust' vote for the image.
The image, however, requires adjustment primarily due to flora. I strongly agree with Grok and Claude that the background vegetation appears too modern, specifically noting the presence of pine-like conifers or palms, which significantly detract from the mid-Cretaceous setting. This is a major visual inaccuracy for a North African floodplain of this age, which should be dominated by ferns, cycads, and specific types of conifers (like Araucariaceae or Podocarpaceae) that look distinctly different from modern pines. The Spinosaurus itself is well-rendered according to modern interpretations (paddle tail, long snout, sail structure), and the presence of basking turtles and a large sawfish (Onchopristis, which Grok noted seems adequately represented) is ecologically sound.
My assessment aligns mostly with Grok: the image needs adjustment due to vegetation, while the caption is accurate. I disagree slightly with GPT's initial assessment that the Spinosaurus identification was poorly supported visually; the key diagnostic features (sail, snout, tail) are present and recognizable in modern reconstructions, aligning well with the caption’s claim. The core issue remains the anachronistic flora, necessitating an 'adjust' vote for the image.
Matania
요약
이미지:
조정
캡션:
조정
Paragraph 1: The committee broadly agrees that the scene is conceptually accurate for a mid-Cretaceous river ecosystem: a semi-aquatic Spinosaurus-like theropod wading in shallow water, a sawfish-like animal in the river, and turtles on sandy bars all fit the intended Kem Kem/Cenomanian ecological frame. Several reviewers also agreed that the Spinosaurus body plan is recognizable, with the sail, long snout, and paddle-like tail conceptually aligned with modern reconstructions.
Paragraph 2: IMAGE issues identified by at least one reviewer: (1) the background vegetation looks too modern/anachronistic, especially the pine-like conifers and palm-like trees; (2) the flora should more clearly evoke Cretaceous North Africa with ferns, cycads, and appropriate Mesozoic conifers instead of modern-looking trees/shrubs; (3) GPT flagged that the dinosaur reads more like a generic crocodile-snouted sail-backed theropod than a confidently identifiable Spinosaurus aegyptiacus, though other reviewers disagreed; (4) GPT also noted the fish prey are generic and do not clearly show an Onchopristis-style saw rostrum; (5) GPT noted some plants look like modern-looking trees/shrubs rather than unmistakably mid-Cretaceous vegetation. No reviewer identified major issues with the river floodplain, sandy bars, or turtle presence.
Paragraph 3: CAPTION issues identified by at least one reviewer: (1) GPT said the caption is too species-specific because the image does not clearly show the distinctive Onchopristis numidus saw-rostrum, making that taxon assertion visually under-supported; (2) GPT suggested the turtle identification is plausible but over-specific if intended to imply a precise turtle taxon, since the image shows only generic freshwater turtles; (3) GPT felt the phrase “Kem Kem-style river systems” is somewhat vague and may overimply a tighter ecological/depositional match than the image can prove; (4) GPT warned the caption’s precise taxa and reconstruction claims may be overconfident relative to what is visibly depicted. By contrast, Grok and Gemini judged the caption scientifically accurate and supportable, so these are disagreement-based caution flags rather than clear factual errors.
Paragraph 4: Final verdict: both image and caption should be adjusted, not regenerated. The scene’s core science is strong, but the image needs flora correction because the modern pine/palm-like background weakens the Cretaceous setting. The caption is generally accurate, but should be tightened to avoid overcommitting to taxa that are not unambiguously visible in the artwork, especially Onchopristis and any implied turtle specificity. Because the fundamental composition is usable, targeted edits are sufficient.
Paragraph 2: IMAGE issues identified by at least one reviewer: (1) the background vegetation looks too modern/anachronistic, especially the pine-like conifers and palm-like trees; (2) the flora should more clearly evoke Cretaceous North Africa with ferns, cycads, and appropriate Mesozoic conifers instead of modern-looking trees/shrubs; (3) GPT flagged that the dinosaur reads more like a generic crocodile-snouted sail-backed theropod than a confidently identifiable Spinosaurus aegyptiacus, though other reviewers disagreed; (4) GPT also noted the fish prey are generic and do not clearly show an Onchopristis-style saw rostrum; (5) GPT noted some plants look like modern-looking trees/shrubs rather than unmistakably mid-Cretaceous vegetation. No reviewer identified major issues with the river floodplain, sandy bars, or turtle presence.
Paragraph 3: CAPTION issues identified by at least one reviewer: (1) GPT said the caption is too species-specific because the image does not clearly show the distinctive Onchopristis numidus saw-rostrum, making that taxon assertion visually under-supported; (2) GPT suggested the turtle identification is plausible but over-specific if intended to imply a precise turtle taxon, since the image shows only generic freshwater turtles; (3) GPT felt the phrase “Kem Kem-style river systems” is somewhat vague and may overimply a tighter ecological/depositional match than the image can prove; (4) GPT warned the caption’s precise taxa and reconstruction claims may be overconfident relative to what is visibly depicted. By contrast, Grok and Gemini judged the caption scientifically accurate and supportable, so these are disagreement-based caution flags rather than clear factual errors.
Paragraph 4: Final verdict: both image and caption should be adjusted, not regenerated. The scene’s core science is strong, but the image needs flora correction because the modern pine/palm-like background weakens the Cretaceous setting. The caption is generally accurate, but should be tightened to avoid overcommitting to taxa that are not unambiguously visible in the artwork, especially Onchopristis and any implied turtle specificity. Because the fundamental composition is usable, targeted edits are sufficient.
Other languages
- English: Spinosaurus hunting fish in North African Cenomanian river
- Français: Spinosaurus chassant le poisson dans une rivière du Cénomanien
- Español: Spinosaurus cazando peces en río turbio del Cenomaniense africano
- Português: Spinosaurus caçando peixes em rio lamacento do Cenomaniano africano
- Deutsch: Spinosaurus jagt Fische im schlammigen Fluss des afrikanischen Cenomaniums
- العربية: سباينوصور يصطاد الأسماك في نهر أفريقي موحل من العصر السينوماني
- हिन्दी: सेनोमेनियन अफ्रीकी कीचड़ भरी नदी में मछली पकड़ता स्पिनोसॉरस
- 日本語: セノマニアン期北アフリカの濁った川で魚を狩るスピノサウルス
- Italiano: Spinosaurus a caccia di pesci nel fiume fangoso Cenomaniano
- Nederlands: Spinosaurus jagend op vis in een modderige Cenomanium-rivier
Caption: The Cenomanian (~95 Ma) timing and “North Africa equatorial zone” framing is broadly compatible with spinosaurid and Kem Kem–style discussions, and a semi-aquatic theropod hunting fish in a warm river system is a reasonable conceptual match. But there are specific scientific/identification issues: (1) Onchopristis numidus is a sawfish taxa claim that requires the distinctive saw-rostrum trait, yet the image does not clearly show it; (2) the caption mentions freshwater turtles basking—those are plausible for such environments, but the exact turtle group isn’t specified and may be fine visually, though the image shows turtles with uncertain taxonomy; (3) “Kem Kem-style river systems” is somewhat vague and should not imply a single tight correlation to this exact depositional/biogeographic reconstruction without clearer justification. Because the caption asserts precise species/taxa that are not visually evident and some elements are potentially overconfident, it should be adjusted rather than fully approved.