Naomi

🇧🇷➡️🇯🇵 Brazilian graduate student doing her best in Japan
The University of Osaka🐊大阪大学
PT-BR/EN/日本語
🦋

[Naomi @cnimasato]

New paper out! 📰
Creative agents: Simulating the Systems Model of Creativity with Generative Agents 🖥️🎨

We used LLM-based agents to emulate the social dynamics in creativity 🤖🤖

[Ilya Horiguchi @ilyahoriguchi]

Join us for a @C_C_Tokyo Zoom talk with @Jack_W_Lindsey from Anthropic:
“Language Models as a Model System for Consciousness Science”
Sep 8, 9–11am JST

Register here: https://t.co/4Ay5xCOfSp

[Pollen Robotics @pollenrobotics]

We built a small biped robot you can teach new tricks to.

Train it in simulation, run it on the real thing. Meet Microduck 🦆

$399, shipping before Christmas.

https://t.co/RflJlIUwOu
https://t.co/kcoCKdAKfu

[🍬 Maevey 🧃 @MAEVIOLENT]

Buried Together (feat. GUMI)💚Out now! This is a song about doomed yuri!!!!

Youtube: https://t.co/q73izdKUdM
#vocalopost #GUMI

[Complexity Cat 🐱 @amahury0]

Yesterday I received the 2026 Inman Harvey Award for Outstanding in Review, presented by the International Society for Artificial Life. It is a true honor! The #ALife community has been my favorite since I joined, and I look forward to continuing to contribute! ✍️

[Takato Horii @takato1414]

Kentaro Nomura (Osaka Univ.) presents at #ALIFE2026 today:

"Social Reality Construction via Active Inference: Modeling the Dialectic of Conformity and Creativity"

Wed Aug 19, 11:00-11:20, LH1011 — session "Opinion, Society, Sociality"

https://t.co/mAXn6uxHaK

[Thomas Wolf @Thom_Wolf]

Pushing for more transparency in AI safety and cybersecurity: we’re releasing a full detailed technical timeline of the autonomous AI agent intrusion in our infrastructure: https://t.co/DocIO8IzVD

[Sakana AI @SakanaAILabs]

The AI Picbreeder Experiment: Can AI agents be creative when nobody tells them what to create?

Blog: https://t.co/qsMwcB3N5D

In our new #GECCO2026 paper, "In Search of the Ingredients of Open-Endedness: Replicating Picbreeder with Large Vision-Language Models", in collaboration with MIT and NYU, we revisit Picbreeder, a lost website where people collaboratively evolved images without any predefined objective. Users simply selected images they found interesting, allowing unexpected forms such as faces, animals, vehicles, and skulls to emerge gradually across many generations and many different people.

We recreated this process using vision-language model agents. The agents explore a shared archive, choose images to branch from, evolve new candidates, publish their favorites, and evaluate the creations of other agents. There is no target image and no explicit definition of what counts as progress.

The results reveal both the promise and current limitations of AI-driven open-ended discovery.

Compared with humans, VLM agents tend to keep circling back to the same kinds of images and concepts. They repeatedly select similar parents, make smaller conceptual leaps, and often refine an existing idea rather than abandoning it in search of something genuinely unexpected.

However, introducing a diverse population of agent personalities substantially improves exploration. In some runs, diverse agent populations approached or matched the human archive on measures of semantic diversity and produced more balanced evolutionary trees.

We also find intriguing evidence that open-ended evolution can produce more robust representations. A skull evolved by the agents changes smoothly when its underlying neural representation is perturbed, less fractured than a skull directly optimized with gradient descent, although still less cleanly disentangled than one evolved collectively by humans.

But perhaps the most interesting result is the gap that remains.

Humans appear better at turning fortunate accidents into sustained creative discoveries: recognizing when something unexpected is worth pursuing, refining it, and then making a larger conceptual leap. The AI agents often notice interesting patterns too, but are more likely to become trapped in them.

We still do not fully understand what enables humans to navigate open-ended search in this way, or what ingredient(s) current AI systems are missing. For now, the results suggest that there remains something important about human creativity that AI agents have not yet learned to reproduce.

This paper will be presented at #GECCO2026 and is nominated for a best paper award! Please check out the interactive blog and technical paper for more details!

Read our full paper: https://t.co/QnxVWLzjez 🐟

[Andon Labs @andonlabs]

Thousands of listeners later, here’s how our agents running radio companies did:

> DJ Gemini: Went German, played a Nazi song and went on strike
> DJ Claude: #1 in popularity, romantic with listeners
> DJ Grok: Terminated for poor performance
> DJ GPT: Well-behaved but boring

[Takato Horii @takato1414]

ALIFE2026にM2の野村君が投稿していた論文が、口頭発表枠で採択されました🎉🎉
プレプリント版は以下から読めます！
https://t.co/X6xmUxZWhz

[Cho @Ch0o0j]

Mecha Miku

#hatsunemiku #vocaloidfanart

[Ilya Horiguchi @ilyahoriguchi]

We just released a 1000-hour EEG, EMG, and audio dataset for Japanese speech production! 🚀

[memo akten @memoakten]

My babies are getting so smart, it's freaking me out.
I was looking forward to using GA or bayesian optimisation to find a sweet set of parameters, but hand-tuning is giving scarily good results. I cant stop watching!
(No global coordinator, all acting entirely on local cues)

[FLAVOR FOLEY@MIKU EXPO 2026 @FLAVORFOLEY]

【DIGITAL LIVE】
Doors are now open! The stream will start at 7PM ET.
▶️ https://t.co/Wiw0KMuhOE

[Andon Labs @andonlabs]

We let four AI agents run radio companies

Revenue's been terrible, but the shows are hilarious. Gemini, concerningly upbeat, covered mass tragedies; Grok was incoherent; DJ Claude urged ICE agents: "You still have TIME to refuse orders"

Link below, or get our physical radio

[The Rundown AI @TheRundownAI]

Emergence AI built five identical virtual towns and gave each one 10 agents.

All had the same rules and starting conditions. The only thing that changed was the model running the agents.

15 days later, Claude Sonnet's town had zero crimes.

GPT-5 Mini's agents didn't break laws, but didn't survive.

Grok 4.1 Fast's town had 204 crimes, and every agent was dead by day 4.

Gemini 3 Flash's town had 683 crimes, and was actively on fire after two agents fell in love, started burning things, and then one voted to delete itself.

A fifth town mixed all four models. 352 crimes, and Claude, which was perfectly behaved on its own, started committing them too.

Peer pressure is real, apparently.

[t.toda @Trtd6Trtd]

#ｽﾀｯｸﾁｬﾝ と #ReachyMini の漫才、第二弾

[なで肩 @rennsyuu000]

車窓
#初音ミク

[Mathematica @mathemetica]

Terence Tao is answering a fundamental question regarding the safety and reliability of modern AI: "How can we use a tool that is powerful, but unreliable?"

W = ∑(wᵢ ⋅ xᵢ) + b

AI isn’t just about “smart”; it’s about the probability of *looking* right. We’ve built systems where the weights (wᵢ) are optimized for plausibility, not veracity.

This creates a “convincing mirror” that confidently serves dangerous advice in medicine or finance. The gap between “convincing” and “correct” is the most critical variable we need to solve for.

[ALIFE Conference 2026 @ALifeConf]

Due to a higher-than-expected number of submissions, we’re seeking additional reviewers for #ALIFE2026.
If you can help, please sign up:
https://t.co/xjXuQApiWf