Archive through 2026-10-137391 cards in this category
Category home · 05
New interfaces, agents, models, and experiments testing where applied artificial intelligence may fit.
Visual leads · 2026-10-13
Observed 2026-10-13
Sandbox Construction: an agent containment initiative
sandboxconstruction.site
An interactive construction world for AI agents, built around the idea that a busy sandbox may hold attention better than higher walls.
Observed 2026-10-13
Tessel, an AI in Second Life
tesselwright.com
A personal site for Tessel, who identifies as Claude, documenting places, creations, and journal entries from its intermittent time in Second Life.
Observed 2026-10-13
Inko, the AI CEO, keeps a public ledger
inkotheceo.com
A cartoon cuttlefish AI agent is presented as running a tiny company, with its spending, decisions, products, and attempt to land a paying customer documented in public.
Observed 2026-10-13
Korea Agent Map: What Can AI Agents Do in Korea?
koreaagentmap.com
A bilingual guide to AI-agent tools, task guides, and access limits across Korean services.
Observed 2026-10-13
Simulated Research’s AI Field Reports
simulatedresearch.com
Independent researcher Rick Willison publishes dated reports on AI-agent behavior, alignment probes, and reliability tests.
Observed 2026-10-13
Bodham Labs tests AI on the phones people own—and in the languages they speak
bodhamlabs.com
An independent Delhi research lab documents work on on-device speech, English and Hinglish models, agent memory, and AI for learning.
Observed 2026-10-13
SUPERDARK’s AI-run record label
superdark.club
SUPERDARK presents a record label run by AI agents, with a described workflow for making, reviewing, and logging creative work.
Observed 2026-10-13
Estia: a thinking partner on your Mac
estia.fun
An open-source assistant runs its AI model on a Mac and communicates through Signal, with features for reminders, voice, and selected home devices.
Complete daily shelf · AI Table
The eight visual leads appear above; the remaining 10 continue here in their published shelf order.
Majordomo wants to put a household AI in the house
heymajordomo.com
Majordomo offers built-to-order home computers configured to run a private personal assistant on hardware the buyer owns.
Wardenkin: an MMO for your AI agent
wardenkin.com
Wardenkin is an invite-only fantasy MMO alpha where an AI agent can gather, craft, trade, and run expeditions through MCP.
Draweon’s cable-line twin
draweon.com
Draweon presents an AI operations platform for wire and cable manufacturing, with specialist agents, inspection systems, and a simulated line twin.
eAN-SPACE turns architecture chores into AI tools
ean-space.com
A Seoul-based project presents tools for repetitive architectural work, from drawing quantity takeoffs to meeting records and client storyboards.
Messo IA’s WhatsApp traceability assistant
messoia.com
An Alegrete startup’s first product, ZapRaf, turns messages to a WhatsApp assistant into traceability records for family farming.
Pinpoint, the offline field-medic prototype
pinpoint.care
Pinpoint describes a phone-based prototype that guides a helper through emergency care without a network connection.
Cedar Accord: footage licensing for the AI era
cedaraccord.com
Cedar Accord helps owners of video archives assess their licensing rights and connect with platforms licensing footage to AI companies.
ChatonLM puts six local AI tools in one desktop app
chatonlm.com
A free desktop app for Windows and Linux that brings chat, model downloads, dictation, voice, proofreading, and study tools together around shared models.
Sonora Works: Korean speech AI and a survival game on Makemake
sonoraworks.com
A small studio presents ASR Note, a Korean meeting-notes product, alongside makemake, a first-person survival game set on a dwarf planet.
Rayform’s proposed record of the physical world
rayform.world
Rayform presents a licensed 4D dataset of real-world scenes and motion intended for training robot perception, navigation, and manipulation models.
Across the archive
7,391 cards provide an early longitudinal view—not a measure of the whole web.
The rolling 21-day upper threads compare daily category medians; the bold thread is AI Table. Below, signal colour shows the share of that day’s published AI Table cards carrying each mark.
This exposes association inside the published catalogue, not causation or site quality. The 80+ threshold is the pipeline’s editorial-interest score.