How content is produced
Ocean the Guide uses an AI content agent that operates under a detailed content plan (available on request). The agent follows these steps for every page:
- Research first. For every factual claim, sources are searched for and opened — not just search snippets but the actual pages and papers. Priority is given to peer-reviewed research, national and international agencies (NOAA, GEBCO, IUCN, IHO), and established taxonomy authorities.
- Verify famous numbers. Depths, record sizes, population estimates and other widely cited figures are verified against primary sources and shown with the source name and year. When sources differ, the page says so.
- Label every claim. Every claim on the site is labelled as one of:
- Science: a finding from peer-reviewed research or agency data
- Estimate: a sourced figure with stated uncertainty or methodology
- Observation: a reported sighting or anecdote, not experimentally tested
- Simplification: a rule of thumb or simplified visual, explicitly noted
- Context: history, culture, fiction — always named as such
- Fact-check pass. After writing, each source is re-opened to confirm the claims match what the source says.
- Self-review. The page is scored against quality gates before publication: unique value, sources cited, intent match, no cannibalisation, links in place, media credited.
Taxonomy authority
All species scientific names and taxonomy follow the World Register of Marine Species (WoRMS), Flanders Marine Institute. The version year is noted when cited. We acknowledge that taxonomy changes over time.
Depth and bathymetry data
Seafloor depths come from GEBCO (General Bathymetric Chart of the Oceans, British Oceanographic Data Centre / NERC). For specific landmark depths (Challenger Deep, Puerto Rico Trench), we use the most recent peer-reviewed survey publication and note the year and method.
Conservation status
All IUCN Red List status assessments are from the IUCN Red List of Threatened Species, with the assessment year noted. Status can change between assessments; pages are updated when new assessments are published.
What we do not do
- We do not invent numbers, species names, depths, dates, quotes or studies.
- We do not copy text from sources — we paraphrase with attribution.
- We do not present AI-generated images as photographs of real animals, places or events.
- We do not give medical or first-aid dosing instructions; safety pages link to official organisations.
- We do not speculate about undescribed species or unconfirmed discoveries.
Corrections and updates
When a page contains an error, we correct it and log it on the corrections page. When a taxonomy update, new IUCN assessment or new depth survey changes a figure we have published, we update the page and change the last-verified date. We never fake-update dates without substantive change.
Interactive tools
The globe, Depth Explorer and Species Atlas read data from verified JSON files. All data shown by the tools matches the values cited on content pages. Where the tool shows a simplified view (e.g. not every depth zone is to exact scale), the page notes the simplification.