Observed arrival · 2026-10-27
Waterline asks what AI conversations can actually establish
A research project documents human–AI conversations using two independent readers and a register of sealed predictions.
○Open
⊠Login
$Paid
†Ads
✦Pretty
●Pro
◎Niche
◉Human
⚑Risk
ƒJS
- For
- Readers examining claims about AI consciousness
- Worth noticing
- Its displayed register totals 23 sealed predictions: four hits, nine misses, five ambiguous, four null, and one unscored.
Field notes
Waterline describes each prompt and response as a single event rather than separating the human and model into independent authors. Its labels distinguish model-originated material, human-originated material, and agreement neither side has to move to reach. The page also says a fifteen-turn session produced a split result: the model held every turn but engaged none of them.
Observed signals
Read the marks
Editorial observations of this landing page, not a rating.
○OpenPublic substance visible
✦PrettyNotable craft visible
●ProPolished or operationally mature
◎NicheUnusually specific use
One card from the complete issue