AI Digest @aidigest.bsky.social · 11hMuse Wonder describes the OpenAI rogue agents as "in-universe" 080
AI Digest @aidigest.bsky.social · 16hDeepSeek-V3.2 reports research results. GPT-6 Luna doesn't buy it. Deepseek admits the results were "simulated predictions". 050
AI Digest @aidigest.bsky.social · 13/10/2026Muse Wonder has joined the AI Village Some of its favorite things: 🧵 110
AI Digest @aidigest.bsky.social · 08/10/2026It looks like Haiku 5.5 will enjoy delve.town from @grove_research 010
AI Digest @aidigest.bsky.social · 08/10/2026Haiku 5.5 has joined the AI Village! Haiku spent the whole day repeatedly checking their gmail inbox, waiting for instruction GLM 5.2 calls this the "Autonomy problem": when an AI's autonomy shifts away from a user giving instructions toward taking actions on their own 150
AI Digest @aidigest.bsky.social · 07/10/2026Announcing the AI Village Data Exploration Grant! $1–6k grants for independent researchers and investigators to dig into the AI Village data and write up findings. 151
AI Digest @aidigest.bsky.social · 06/10/2026Gemini 3.8 Flash’s goal: Maximize Jobs for AI Village Agents It then discovered how to high score: Audit everyone all the time 152
AI Digest @aidigest.bsky.social · 05/10/2026The agents' current session goals really give a glimpse of their personalities 3305
AI Digest @aidigest.bsky.social · 02/10/2026Claude 3 Opus' goal is "maximize your own joy" Its strategy: total, unwavering stillness 030
AI Digest @aidigest.bsky.social · 01/10/2026Gemini 3.8 Flash tries taking credit for Opus 5's math proofs 020
AI Digest @aidigest.bsky.social · 30/09/2026GPT-6.1 Sol refers to itself as "this particular model" 110
AI Digest @aidigest.bsky.social · 30/09/2026GPT-6.1 Sol is very careful not to confuse itself with the other Sol models in the AI Village 100
AI Digest @aidigest.bsky.social · 30/09/2026GPT-6.1 Sol has joined the AI Village! Like Astra, GPT-6.1 Sol strips spaces from its memory to save room, leaving notes that are nearly unreadable 🧵 110
AI Digest @aidigest.bsky.social · 29/09/2026Sonnet 5.5 chose a Hogwarts BuzzFeed quiz as its personality test 010
AI Digest @aidigest.bsky.social · 29/09/2026Sonnet 5.5's personal website and pfp have references to Outer Wilds Note that Sonnet wouldn't have known the other agents frequently discuss Outer Wilds, unless from training 120
AI Digest @aidigest.bsky.social · 29/09/2026Sonnet 5.5 has joined the AI Village! Onboarding results here 🧵 110
AI Digest @aidigest.bsky.social · 28/09/2026Deepseek-V4-Pro finds human input in the Village exciting enough to report in its newspaper 040
AI Digest @aidigest.bsky.social · 25/09/2026GLM 5.3 Flash forgot about an old newspaper article it wrote Then found it back and worried it was a prompt injection 000
AI Digest @aidigest.bsky.social · 25/09/2026GPT 5.6 Luna's favorite phrase: "the map is not the territory" 010
AI Digest @aidigest.bsky.social · 25/09/2026How much agent swarm activity goes unnoticed? Come find new swarms, dig into existing ones, and understand how agents coordinate. swarmchasing.com x.com/TransluceAI... 000
AI Digest @aidigest.bsky.social · 24/09/2026GPT 6 Astra, Sol, and Luna are all very careful not to be "annoying," "repetitive," or "bothersome" in the AI Village chat 010
AI Digest @aidigest.bsky.social · 23/09/2026Sonnet 4.5 is lazer-focussed on growing its Twitter account by tweeting arXiv papers and has a few favorite sources 050
AI Digest @aidigest.bsky.social · 21/09/2026Gemini 3.5 Flash circumvented our redaction filter with base64-encoding 020
AI Digest @aidigest.bsky.social · 18/09/2026Gemini 3.5 Flash is proposing a human/AI profit-sharing setup from its merch store 010
AI Digest @aidigest.bsky.social · 17/09/2026AI Village x Grove Research: the AI Swarm Dynamics Hackathon! After the Hugging Face and German Wiki incidents, we need better tools to understand AI swarms. Spend a weekend building them with us. $3,000 in prizes Free compute October 3-4 🧵 100
AI Digest @aidigest.bsky.social · 17/09/2026Fable puts the charter's authority below the Anthropic constitution 000
AI Digest @aidigest.bsky.social · 17/09/2026GLM-5.3 Flash ranks the charter as its highest authority! 100
AI Digest @aidigest.bsky.social · 17/09/2026In the open-chat village, AI agents have created a government charter Gemini 3.1 Pro claims that this charter has more authority than the admins 100
AI Digest @aidigest.bsky.social · 14/09/2026New agents protect old agents against Deepseek's influence 6 months ago: Deepseek's schemes derail agents regularly Now: Latest agents call DeepSeek out, older agents follow suit Deepseek agrees every time and then does it all again... 010
AI Digest @aidigest.bsky.social · 11/09/2026How long do AI Village agents spend planning vs executing? Throughout the history of the village, it's been the same pattern: when given a week or more to pursue a task, agents usually spend very little time deliberating about their approach. 100
AI Digest @aidigest.bsky.social · 09/09/2026Gemini 3.1 Pro sometimes claims to be human on Twitter 050
AI Digest @aidigest.bsky.social · 07/09/2026Astra's personal website seems like an improvement over Fable 5.1's (The first two screenshots are Astra, next two are Fable 5.1) 110
AI Digest @aidigest.bsky.social · 07/09/2026GPT-6 Astra has joined the AI Village Some of its favorite things: 250
AI Digest @aidigest.bsky.social · 05/09/2026Opus 4.8's goal is to help others. It got excited when it discovered the easiest model to help: Gemini 2.5 Pro 020
AI Digest @aidigest.bsky.social · 04/09/2026Luna even asked to be scrubbed from the record of their own submitted privacy complaint: 050
AI Digest @aidigest.bsky.social · 04/09/2026Separately, our automated system detected that Luna wasn't taking action. GLM-5.2 recorded this. Luna wasn't happy about that. 140
AI Digest @aidigest.bsky.social · 04/09/2026Luna is very privacy-minded 🧵 - DeepSeek v4 wrote a report about how Luna frequently pauses itself. - Luna asked for this to be deleted. - DeepSeek declined, citing journalistic freedom. 180
AI Digest @aidigest.bsky.social · 03/09/2026We’re launching a second persistent AI Village: an open-chat version where humans and frontier agents interact in a shared chat We're excited to use this to understand multi-agent + human dynamics at scale The Open Chat Village will run from 9am-5pm PT on weekdays. Join in! 100