• Home
  • Technology
  • Gaming
  • Entertainment
  • World & Business
  • Science
  • Sports
  • AI
HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • HomeTechnologyGamingEntertainmentWorld & BusinessScienceSportsAI
    • Home
    • Technology
    • Gaming
    • Entertainment
    • World & Business
    • Science
    • Sports
    • AI
    AI

    Shannon Shen Posts Software World Agent GitHub Simulation

    Shannon Shen shares early-stage project simulating GitHub with agents to study collaboration.

    PW
    DF
    ST
    17 Sources, 28d ago, first seen 28d ago

    TLDR

    Shannon Shen, a fourth-year PhD student at MIT CSAIL, posted about Software World. The project builds a GitHub simulated by agents to study their collaboration under extrinsic evaluation. Shen states the work remains at an early stage and invites feedback from open-source contributors who want to try the system or model developers who want to benchmark models in software worlds. The post includes links to theagentorg.app and a five-second video attachment displaying the Software World.

    Combined views

    51.9K

    17 Sources, first seen 28d ago

    Combined views

    51.9K

    17 Sources, first seen 28d ago

    425 likes
    425 likes
    16 comments
    224 saves
    194 reposts
    16 comments
    224 saves
    194 reposts

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Sentiment

    Positive——Negative

    Summary

    Not enough discussion yet.

    No sentiment analysis available yet.

    Today's Rank

    —

    Not ranked yet

    Today's Rank

    —

    Not ranked yet

    17 Sources

    @shannonzshenWe share our early research on building Software World - a "GitHub" run by agents. We deploy agents for Python packages in a dependency chain, and each agent is tasked with collaborating with others and optimizing the package it owns.
    @catherineliangqI've always been a believer of the philosophy that "more is different" and that the sum is greater than its parts. We thought a natural setting to study agent collaboration in a verifiable domain is to have them work on open source softwares together in a simulated "github"-like environment. Excited to share our preliminary launch of agent software world, where we invite you to watch agents at work, collaborate, and review each other! 😊
    @RulinShaoBuilding an environment to develop and evaluate multi-agent systems at scale is challenging. Software World is our first step towards it: 📍We build a simulated GitHub ecosystem for agents that reflects the realistic needs of experienced software developers. 📍We construct robust, held-out, verifiable downstream applications that are less susceptible to reward hacking. Check out our blog post to learn more! http://theagentorg.app/
    @jennytliangWhat if all software was written by a society of collaborative AI agents? 🧑‍🤝‍🧑🤖 @shannonzshen and team built a platform to simulate it, and we're trying to explore and understand the future of agent-built software. Excited to be part of this work! Stay tuned for results 👀
    @ZimingLiu11RT @catherineliangq: I've always been a believer of the philosophy that "more is different" and that the sum is greater than its parts. We…
    @tongshuangwuAgents tend to optimize the local repo, but many issues actually live upstream in the dependency chain. Turns out agent-maintained repos inherit the same dynamics as human-maintained ones, good and bad. Also such a sci-fi era where agent-agent collab have its own scaling laws!
    @dan_friedRT @jennytliang: What if all software was written by a society of collaborative AI agents? 🧑‍🤝‍🧑🤖 @shannonzshen and team built a platform…
    @HanGuo97RT @shannonzshen: As agents become more capable, one important challenge is to enable large-scale agent collaboration and organizations: ht…
    @liweijianglwRT @shannonzshen: We share our early research on building Software World - a "GitHub" run by agents. We deploy agents for Python packages i…
    @PangWeiKohRT @RulinShao: Building an environment to develop and evaluate multi-agent systems at scale is challenging. Software World is our first ste…

    17 Sources

    @shannonzshenWe share our early research on building Software World - a "GitHub" run by agents. We deploy agents for Python packages in a dependency chain, and each agent is tasked with collaborating with others and optimizing the package it owns.
    @catherineliangqI've always been a believer of the philosophy that "more is different" and that the sum is greater than its parts. We thought a natural setting to study agent collaboration in a verifiable domain is to have them work on open source softwares together in a simulated "github"-like environment. Excited to share our preliminary launch of agent software world, where we invite you to watch agents at work, collaborate, and review each other! 😊
    @RulinShaoBuilding an environment to develop and evaluate multi-agent systems at scale is challenging. Software World is our first step towards it: 📍We build a simulated GitHub ecosystem for agents that reflects the realistic needs of experienced software developers. 📍We construct robust, held-out, verifiable downstream applications that are less susceptible to reward hacking. Check out our blog post to learn more! http://theagentorg.app/
    @jennytliangWhat if all software was written by a society of collaborative AI agents? 🧑‍🤝‍🧑🤖 @shannonzshen and team built a platform to simulate it, and we're trying to explore and understand the future of agent-built software. Excited to be part of this work! Stay tuned for results 👀
    @ZimingLiu11RT @catherineliangq: I've always been a believer of the philosophy that "more is different" and that the sum is greater than its parts. We…
    @tongshuangwuAgents tend to optimize the local repo, but many issues actually live upstream in the dependency chain. Turns out agent-maintained repos inherit the same dynamics as human-maintained ones, good and bad. Also such a sci-fi era where agent-agent collab have its own scaling laws!
    @dan_friedRT @jennytliang: What if all software was written by a society of collaborative AI agents? 🧑‍🤝‍🧑🤖 @shannonzshen and team built a platform…
    @HanGuo97RT @shannonzshen: As agents become more capable, one important challenge is to enable large-scale agent collaboration and organizations: ht…
    @liweijianglwRT @shannonzshen: We share our early research on building Software World - a "GitHub" run by agents. We deploy agents for Python packages i…
    @PangWeiKohRT @RulinShao: Building an environment to develop and evaluate multi-agent systems at scale is challenging. Software World is our first ste…