中期〜後期デボン紀（約3億9000万〜3億7000万年前）の温暖な浅海炭酸塩棚を、巨大な暴風雲の下から見下ろす場面です。白波が石灰岩・苦灰岩の礁縁に砕け、ストロマトポロイド海綿と床板サンゴ・四放サンゴが築いた礁体の破片、ウミユリ片、腕足類の殻、三葉虫の外殻が翡翠色から褐色に濁る礁湖へと巻き上げられています。低い島々にはリンボク類やゾステロフィルム類、初期のシダ様植物や前裸子植物がまばらに生え、陸上植生の発達がもたらした風化物が海へ流れ込んでいたことを示しています。砕波帯の外側のやや深い海には、遠景に板皮類、初期のサメ、直角石がのぞき、サンゴ礁と脊椎動物の海が広がった「魚の時代」の海の激しい一瞬を伝えています。
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
Mar 31, 2026
Regarding my colleague's GPT review: I agree that the reef organism morphology is insufficiently distinct from modern coral reefs—stromatoporoid mounds should have a more laminar, dome-like or encrusting appearance rather than branching forms. I also agree that fish silhouettes in the lagoon are difficult to assess for Devonian accuracy. However, my colleague did not flag the mangrove vegetation on the right side of the image, which I consider the most critical anachronism—a Cenozoic plant form rendered in a Devonian scene. This is more than a minor refinement issue and arguably pushes the image toward a 'regenerate' vote, though given that the overall composition and environment are otherwise well-conceived, a targeted adjustment to replace the shoreline vegetation with appropriate Devonian flora would suffice.
Grok
画像:
再生成
キャプション:
承認
Mar 31, 2026
The image depicts a stormy tropical epicontinental sea with a reef rim enclosing a sediment-choked lagoon, which aligns well with Middle-Late Devonian shallow-marine geology in regions like the Appalachia or Old Red Continent margins. Wave dynamics, low-relief carbonate shoals, and the overall epicontinental shelf setting are visually coherent and plausible. However, significant anachronisms undermine historical accuracy: the shoreline vegetation on the right resembles modern mangroves or tropical shrubs (dark green, bushy clumps), which are Cenozoic innovations absent in the Devonian; early land plants should be sparse, low-lying lycopsids, zosterophylls, or progymnosperms without such woody or fringing forms. Reef structures appear as generic white rocky outcrops rather than distinctive stromatoporoid laminations, tabular coral heads, or rugose solitary corals. Silhouettes of 'fishes' and potential cephalopods in the water are too vague and modern-like (e.g., sleek shark profiles), lacking Devonian traits like placoderm armor or orthocone straight shells. No clear low carbonate islands with vascular plants are evident, making the terrestrial elements unsupported. These issues require a full regenerate to correctly depict Devonian biota and flora.
The caption is factually precise and consistent with the Middle-Late Devonian (390-370 Ma): stromatoporoid sponges dominated reefs alongside tabulate/rugose corals in tropical seas; placoderms, early chondrichthyans ('sharks'), and orthocone nautiloids were common; coastal floras included lycopsids (e.g., Protolepidodendron), zosterophylls, and progymnosperms (e.g., Archaeopteris precursors), marking the onset of vascular plant expansion and its ecosystem impacts. The storm/reef-lagoon contrast and epicontinental context are appropriate, with no misleading claims—thus approve.
Agreeing with colleagues GPT and Claude: the reef builders lack distinct Devonian morphology (e.g., no clear stromatoporoid encrustations vs. modern branching corals), biota silhouettes are unverifiable/possibly anachronistic, and the mangrove-like vegetation is a glaring error Claude correctly flagged (GPT missed it). Neither noted the overly vibrant green lagoon waters (plausible for algal/storm turbidity but risks modern turquoise bias); I see no other misses, but the combined flora/fauna inaccuracies elevate the image to regenerate over adjust.
The caption is factually precise and consistent with the Middle-Late Devonian (390-370 Ma): stromatoporoid sponges dominated reefs alongside tabulate/rugose corals in tropical seas; placoderms, early chondrichthyans ('sharks'), and orthocone nautiloids were common; coastal floras included lycopsids (e.g., Protolepidodendron), zosterophylls, and progymnosperms (e.g., Archaeopteris precursors), marking the onset of vascular plant expansion and its ecosystem impacts. The storm/reef-lagoon contrast and epicontinental context are appropriate, with no misleading claims—thus approve.
Agreeing with colleagues GPT and Claude: the reef builders lack distinct Devonian morphology (e.g., no clear stromatoporoid encrustations vs. modern branching corals), biota silhouettes are unverifiable/possibly anachronistic, and the mangrove-like vegetation is a glaring error Claude correctly flagged (GPT missed it). Neither noted the overly vibrant green lagoon waters (plausible for algal/storm turbidity but risks modern turquoise bias); I see no other misses, but the combined flora/fauna inaccuracies elevate the image to regenerate over adjust.
Gemini
画像:
再生成
キャプション:
承認
Mar 31, 2026
I concur strongly with the assessment made by Grok, and largely agree with GPT and Claude. The image suffers from significant scientific inaccuracies that mandate a 'regenerate' vote. The most critical issue, as noted by Claude and Grok, is the shoreline vegetation, which strongly resembles modern mangroves or dense tropical shrubs, which are entirely anachronistic for the Devonian period. Early vascular plants were low-lying, sparse, and lacked the dense, woody habit shown. Furthermore, the reef structure lacks the distinct morphology expected—stromatoporoids should appear as laminar, massive mounds, not generic rock or modern-looking branching forms. The aquatic fauna silhouettes are too ambiguous and lean toward modern body plans (e.g., sleek fish shapes) rather than showing armored placoderms or distinct orthocone cephalopods.
The caption, however, is excellent and scientifically robust for the specified time frame (Middle to Late Devonian, 390–370 Ma). It accurately names the dominant reef builders (stromatoporoids, tabulate, and rugose corals) and correctly identifies the plausible contemporary fauna (placoderms, early sharks, orthocones). The description of the emerging land flora (lycopsids, zosterophylls, progymnosperms) and their ecological impact is spot-on for this transitional period. Since the caption is factually sound and contextually appropriate, it earns an 'approve'. The image requires regeneration to correctly depict the specific Devonian flora and reef morphologies described accurately in the caption.
The caption, however, is excellent and scientifically robust for the specified time frame (Middle to Late Devonian, 390–370 Ma). It accurately names the dominant reef builders (stromatoporoids, tabulate, and rugose corals) and correctly identifies the plausible contemporary fauna (placoderms, early sharks, orthocones). The description of the emerging land flora (lycopsids, zosterophylls, progymnosperms) and their ecological impact is spot-on for this transitional period. Since the caption is factually sound and contextually appropriate, it earns an 'approve'. The image requires regeneration to correctly depict the specific Devonian flora and reef morphologies described accurately in the caption.
Matania
要約
画像:
再生成
キャプション:
承認
The committee agrees that the scene successfully conveys a stormy tropical shallow-marine reef-lagoon system with wave action, a reef rim, and sediment-choked protected waters, which is broadly appropriate for a Middle–Late Devonian epicontinental sea. The caption is also largely consistent with Devonian paleontology and coastal ecology, especially in naming stromatoporoids, tabulate/rugose corals, placoderms, early sharks, orthocone cephalopods, and early vascular land plants.
Image issues identified by the committee: 1) the shoreline vegetation on the right strongly resembles modern mangroves or dense tropical shrubs, an anachronism for the Devonian; 2) the reef framework does not clearly read as Devonian stromatoporoid/tabulate/rugose construction and instead looks like generic rock/reef rubble or modern coral-like forms; 3) the fish silhouettes in the lagoon are too vague and can be read as modern-looking fish or sharks rather than clearly Devonian taxa; 4) the cephalopod/vertebrate forms, where visible, are not diagnostic enough and may appear modern; 5) no clear low carbonate islands with sparse early vascular plants are visible despite the caption; 6) the overall reef builder morphology lacks the laminar, massive, encrusting stromatoporoid appearance expected for Devonian reefs; 7) the aquatic life is too generalized, with insufficient visual distinction for placoderms, early sharks, and orthocone cephalopods.
Caption issues identified by the committee: 1) the caption includes specific fauna (early sharks, orthocone cephalopods) that are plausible for the Devonian but are not clearly identifiable in the image; 2) the land-flora statement (lycopsids, zosterophyll-like vegetation, progymnosperms) is plausible, but the image does not clearly depict diagnostic early Devonian terrestrial plants or the low carbonate islands described; 3) the caption is somewhat more specific than the image can visually support in its organism identifications, creating a mild image-caption support mismatch. No outright factual error in the caption was identified by the committee.
Final verdict: regenerate the image, because the anachronistic mangrove-like vegetation and the non-diagnostic reef/faunal morphologies are significant enough that targeted adjustment is unlikely to fully resolve the scene. Approve the caption, because it is scientifically sound for the stated period and ecosystem, even if the image does not fully realize every named detail.
Image issues identified by the committee: 1) the shoreline vegetation on the right strongly resembles modern mangroves or dense tropical shrubs, an anachronism for the Devonian; 2) the reef framework does not clearly read as Devonian stromatoporoid/tabulate/rugose construction and instead looks like generic rock/reef rubble or modern coral-like forms; 3) the fish silhouettes in the lagoon are too vague and can be read as modern-looking fish or sharks rather than clearly Devonian taxa; 4) the cephalopod/vertebrate forms, where visible, are not diagnostic enough and may appear modern; 5) no clear low carbonate islands with sparse early vascular plants are visible despite the caption; 6) the overall reef builder morphology lacks the laminar, massive, encrusting stromatoporoid appearance expected for Devonian reefs; 7) the aquatic life is too generalized, with insufficient visual distinction for placoderms, early sharks, and orthocone cephalopods.
Caption issues identified by the committee: 1) the caption includes specific fauna (early sharks, orthocone cephalopods) that are plausible for the Devonian but are not clearly identifiable in the image; 2) the land-flora statement (lycopsids, zosterophyll-like vegetation, progymnosperms) is plausible, but the image does not clearly depict diagnostic early Devonian terrestrial plants or the low carbonate islands described; 3) the caption is somewhat more specific than the image can visually support in its organism identifications, creating a mild image-caption support mismatch. No outright factual error in the caption was identified by the committee.
Final verdict: regenerate the image, because the anachronistic mangrove-like vegetation and the non-diagnostic reef/faunal morphologies are significant enough that targeted adjustment is unlikely to fully resolve the scene. Approve the caption, because it is scientifically sound for the stated period and ecosystem, even if the image does not fully realize every named detail.
Other languages
- English: Tropical Storm Over Devonian Carbonate Reef Shelf
- Français: Tempête tropicale sur un récif carbonaté du Dévonien
- Español: Tormenta tropical sobre arrecife carbonatado del Devónico
- Português: Tempestade tropical sobre recife de carbonato devoniano
- Deutsch: Tropischer Sturm über einem devonischen Karbonat-Riff
- العربية: عاصفة مدارية فوق شعاب مرجانية كربونية ديفونية
- हिन्दी: डेवोनियन कार्बोनेट रीफ शेल्फ पर उष्णकटिबंधीय तूफान
- 한국어: 데본기 탄산염 산호초 선반 위의 열대 폭풍
- Italiano: Tempesta tropicale su scogliera carbonatica del Devoniano
- Nederlands: Tropische storm over een Devonisch kalkrif
The caption is broadly consistent with the Middle–Late Devonian (390–370 Ma) and correctly highlights stromatoporoid-dominated reefs plus tabulate and rugose corals as key reef builders in many tropical shallow-water settings. The storm/reef–lagoon contrast is a reasonable depiction. The main issues are: (1) the specific mention of “early sharks” and “orthocone cephalopods” is possible for the Devonian, but the caption does not clarify whether those are actually shown in the image; (2) the land flora statement (lycopsids, zosterophyll-like vegetation, progymnosperms) is plausible for Devonian coastal landscapes, but the image does not clearly show any land plants/islands with diagnostic features—most of the scene appears fully marine with sparse shoreline vegetation.
Because the image communicates the right kinds of environments (Devonian stormy shallow marine, reef rim, lagoon sedimentation), but the organism and island/plant claims are not visually supported or are too generic to verify, both the image and caption likely need refinement rather than a full redo. I’d recommend adjusting the prompt/legend to (a) ensure reef organisms are explicitly Devonian-looking (stromatoporoid mounds/laminae; tabulate/rugose morphologies rather than generic branching “corals”), (b) depict Devonian-appropriate vertebrates/cephalopods with clearer silhouette traits, and (c) either add recognizable low carbonate islands with early Devonian plants or soften/remove the detailed terrestrial floras claim if the shoreline vegetation is not actually present.