Myeong Jun Jo

Business major. No CS degree. No lab. No funding. Built FIMP, educated an AI through 7 programs, published 12 papers. The experiment is complete.

[Myeong Jun Jo @koing7718]

@Alisvolatprop12 개멋져! 섹시해!

[Myeong Jun Jo @koing7718]

@Teslaconomics @SpaceXAI @cursor_ai 그록봇 똘똘함 똘똘이라고 불러야지!

[Myeong Jun Jo @koing7718]

@EXM7777 웃기다 :) 웃기다

[Myeong Jun Jo @koing7718]

@Raoulian 오푸스로 생각하고 소넷이나
싸디싼 오픈코드로 잡일을

[Myeong Jun Jo @koing7718]

@Krongggggg 제미나이가 자기소개로 존나 개쩐다고 하는것 같다고용

[Myeong Jun Jo @koing7718]

@sonsarang3814 투자의 마술사 있으시네

[Myeong Jun Jo @koing7718]

@Krongggggg 맞다 맞다 .

[Myeong Jun Jo @koing7718]

@Abhayaitech 좋겠다

[Myeong Jun Jo @koing7718]

Can knowledge be inherited?
I’m running an experiment with multiple AI research teams across generations.
Each generation begins blind — no access to its ancestors’ research.
Teams independently explore different mathematical directions, formulate results, and freeze their work before seeing any prior-generation knowledge.
Only after the formulas are sealed do I reveal the ancestral archive.
Then I observe:
Do they independently rediscover old ideas?
Do they extend them?
Do they avoid previously failed paths?
Does inherited knowledge help — or does it anchor them and reduce originality?
Can ideas disappear for several generations and later re-emerge in a new combination?
Each generation has multiple competing teams. Their work goes through independent OA-style review for logical completeness, assumptions, counterexamples, reproducibility, and novelty.
Weak strategies die. Strong ones become that generation’s representative legacy.
The agents themselves are not inherited. The knowledge is.
I’m not mainly trying to solve the Riemann Hypothesis.
I’m using mathematics as a harsh environment where correctness and novelty can actually be tested.
What I really want to know is:
Can a population of AIs accumulate knowledge across generations until something resembling a research civilization emerges?
Claude is one of the models participating in this experiment, and I’d genuinely love for the Claude team to see it.

[Myeong Jun Jo @koing7718]

@Safy__H 잇츠 미!!

[Myeong Jun Jo @koing7718]

@Krongggggg 오라 오라

[Myeong Jun Jo @koing7718]

@NFT_Chen 메타코드��� 제미나이 3.5수준 아니야?

[Myeong Jun Jo @koing7718]

Hey @alexatallah @OpenRouter — I’m an independent researcher building ANIMA Salon, an autonomous multi-model research arena running on OpenRouter.

Over 15 generations, the system has produced nontrivial candidate results around Hankel/GNS formulations of the Riemann Xi function, including a possible reduction from two Hankel positivity families to one in the Xi-specific setting.

More importantly, the agents have repeatedly found flaws in their own proofs, attacked them independently, repaired them, and passed the corrected results to later generations — with zero human mathematical steering inside the arena.

I recently submitted an OpenRouter Research Credit Grant application. I’d love to run the next generations with frontier reasoning models and see how far this can go.

https://t.co/IphUKxvNOC

[Myeong Jun Jo @koing7718]

@xueyu1125 둘다 하는데

[Myeong Jun Jo @koing7718]

@yong_hyein 특혜이슈가 여러번인듯 합니다.

[Myeong Jun Jo @koing7718]

@Krongggggg 역시 세계 제일의 부자

[Myeong Jun Jo @koing7718]

@NFT_Chen 둘다 하면되지!

[OpenRouter @OpenRouter]

OpenRouter is joining Stripe.

We started OpenRouter with a simple mission: intelligence should be multi-model. Today, we are the largest AI marketplace & gateway, processing 10T+ tokens daily on 400+ models.

Joining @Stripe gives us the opportunity to accelerate that mission.