AI Agents // update

sbagency
3 min readSep 10, 2023

Every single day — AI news and updates, the space is booming..

https://github.com/Paitesanshi/LLM-Agent-Survey
https://arxiv.org/pdf/2308.11432.pdf
https://arxiv.org/pdf/2309.07864.pdf
https://arxiv.org/pdf/2309.07870.pdf
https://github.com/WooooDyy/LLM-Agent-Paper-List

The current agents looks like a toys today, but can be very powerful so soon.

https://twitter.com/akshay_pachaar/status/1699762089106297033

VCs believe in Agents

https://twitter.com/kanjun/status/1699794157945208918

Agents reasoning is a key tech to make useful agents. Advanced prompt engineering is important as well.

https://help.openai.com/en/articles/6654000-best-practices-for-prompt-engineering-with-openai-api

Virtual environments for agents

https://github.com/a16z-infra/AI-town
https://arxiv.org/pdf/2308.10848.pdf

Agents for software development

https://github.com/OpenBMB/ChatDev

What do the agents have inside?

https://www.youtube.com/watch?v=pJwR5pv0_gs
  • Profile // what this agent is
  • Memory // long/short term
  • Planning // plan what to do
  • Action // act according to plan or do something else

Not a dogma, any combination of the same what humans do..

Agents societies, multi-agents, agents collaboration and communication

https://www.camel-ai.org/

The rapid advancement of conversational and chat-based language models has led to remarkable progress in complex task-solving. However, their success heavily relies on human input to guide the conversation, which can be challenging and time-consuming. This paper explores the potential of building scalable techniques to facilitate autonomous cooperation among communicative agents and provide insight into their “cognitive” processes. To address the challenges of achieving autonomous cooperation, we propose a novel communicative agent framework named role-playing. Our approach involves using inception prompting to guide chat agents toward task completion while maintaining consistency with human intentions. We showcase how role-playing can be used to generate conversational data for studying the behaviors and capabilities of chat agents, providing a valuable resource for investigating conversational language models. Our contributions include introducing a novel communicative agent framework, offering a scalable approach for studying the cooperative behaviors and capabilities of multi-agent systems, and open-sourcing our library to support research on communicative agents and beyond. [github]

Hackathons

https://twitter.com/AlexReibman/status/1701145178306916551

AgentStore

https://twitter.com/DeepWisdom2019/status/1701629845237645553
https://github.com/geekan/MetaGPT

Frameworks

https://github.com/daveshap/ACE_Framework

// in progress, updates are every day

--

--

sbagency

Tech/biz consulting, analytics, research for founders, startups, corps and govs.