Category pulse
score + signals · rolling 21 daysThe 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.
Archive through 2026-11-127920 cards in this category
Category home · 05
New interfaces, agents, models, and experiments testing where applied artificial intelligence may fit.
Visual leads · 2026-11-12
Observed 2026-11-12
If Claude Was King
ifclaudewasking.com
A recurring experiment asks Claude to plan a year of world rule, then makes the model’s real-world forecasts available for later scorekeeping.
Observed 2026-11-12
Can the Other Side Steer Your AI Lawyer?
adversarialrecord.com
A working paper tests whether adversarial framing can shift AI systems reviewing contracts and other legal records.
Observed 2026-11-12
Agentiqad’s ad slot for AI shopping agents
agentiqad.com
Agentiqad presents a sponsored-placement layer that lets shops return labeled product offers to AI shopping agents through a web/API decision endpoint.
Observed 2026-11-12
Hand Line Data turns skilled work into robot-training footage
handlinedata.com
A recruiting and scheduling service that arranges for skilled workers to record real tasks for physical-AI and robotics teams.
Observed 2026-11-12
Mosnid turns Algerian shop chats into courier-ready orders
mosnid.shop
An AI assistant for Algerian online sellers that answers customers across WhatsApp, Telegram and web shops, then helps turn conversations into delivery orders.
Observed 2026-11-12
Callround’s voice-first hiring demo
callround.chat
Callround presents a voice-first applicant-screening service: applicants answer interview questions by phone, and recruiters get a ranked shortlist with recordings and quoted evidence for scores.
Observed 2026-11-12
Ancor’s gap-targeted reality capture
tryancorai.com
Ancor describes a service that identifies gaps in AI models and supplies provenance-tracked observations captured in the real world.
Observed 2026-11-12
EveryNeo’s ledger of AI labs and compute deals
everyneo.com
A public, source-cited ledger tracking AI labs, neocloud providers, funding rounds, founders, investors, and compute deals.
Complete daily shelf · AI Table
The eight visual leads appear above; the remaining 9 continue here in their published shelf order.
Physical Onsite turns work-site tasks into physical-AI data
physical-onsite.com
A Vietnam-based service offering custom physical-AI training data collected by trained workers in factories and other requested environments.
SaaS Marathon Bench puts AI models through a multi-app workday
saasmarathon.dev
A benchmark compares models carrying out multi-step business tasks across common SaaS apps, using either the screen or APIs.
Unfold’s atlas of leaderless swarms
unfold.host
An AI research lab maps how living species coordinate without leaders, then uses those behaviors to frame questions about agentic AI swarms.
AI Product Displacement Risk
aiproductdisplacement.com
A research project introduces a firm-level measure of how vulnerable companies may be to generative AI replacing their products.
Saltline Game, an AI-built portside FPS
saltlinegame.pro
A free browser FPS set in a rain-soaked container port, where players cut three relays and reach a boat.
Lockstep keeps AIs on the same page
golockstep.com
A private-beta service that says it shares confirmed decisions across AI apps and drafts corresponding updates to connected documents.
1M Voices: a million-interview ambition
millioninterviews.com
Cookiy Labs presents a plan for a searchable, global consumer-interview archive, with paid tiers for transcripts, custom questions, and video.
Omicent: shoplifting alerts from existing store cameras
omicenthq.com
Omicent presents an AI system that watches a store’s existing camera feeds for specified concealment behaviors and sends staff an alert clip.
AiQcha’s AI bear asks children to move
aiqcha.com
AiQcha presents an AI companion for children ages 2–6, with movement games, interactive cartoons, picture lessons, and question-and-answer play in Uzbek and Russian.
Across the archive
7,920 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.