Agenda

1. recap AI4D Meeting, @Tina Tsou , @Haruhisa Fukano , @Tom Qin , @Yona Cao , @C.C. Fan , @Jeff Brower , 20 minutes

2. Overview update @Wilson Wang

3. SPEAR update

4. Discuss Security workstream @Victor Lu

Participants

@Tom Qin @Tina Tsou @Jeff Brower @Qi Tang @Wilson Wang @Victor Lu @Caleb @Borui Li (李博睿)

Recording

https://bytedance.us.larkoffice.com/minutes/obussmr3jc5y35gdtf5754jc?from=from_copylink

Summary

The meeting discussed the recap of the AI for Developer's Meeting and the update of the Infinite Edge AI project, the main contents included:

• Overview update: There is no much update this week, but the MVP part will be updated next.

• Spear update: The main focus will be on the agent layer, and the core functionalities will be provided for the agents running on the platform.

• Security and safety: There is a need to start a dialogue on security and safety in the RFH, and document some of the main errors.

• Event recap: The purpose, focus, event process, important announcements, action items, and conclusions of the AI for Developer's Meeting were shared.

• Open source AI: The definition and levels of open source AI were discussed, and the importance of open science in ensuring the control of AGI was emphasized.

• AI safety and security: The combination of AI security and AI safety is a popular subject, and there is a need to start a dialogue on security and safety in the RFH.

Chapters

00:06 Recap and Agenda Discussion of a Meeting

This section mainly discusses the sharing of information to different groups, such as the Infinite Hai Lark group and the slack. It also mentions people's participation status in the meeting, including time zone differences and possible absences. Tina suggests lining up agenda items before the meeting and asks Tom to share his screen. There's mention of holiday situations and the recap of the meeting agenda.

04:33 Agenda Items and Updates in the Meeting

This section mainly focused on the meeting agenda. It included the overview update by Wilson and the spear update. There were discussions about adding the AI agent platform to the overview. Wilson shared plans for the spear, involving coordination with internal teams and focusing on launching or Lark Index first. Various people were assigned tasks, and there was mention of saving time and adding agenda items before the meeting.

11:28 Updates and Recap of AI Developer Meeting and Related Discussions

This section covers updates from Wilson Wang about syncing and further discussions. There's talk about dates and requirements. Tina Tsou recaps the AI for Developer's Meeting from September 11 to 13. It included events like the hacker song, various sessions, international attendees, and good feedback. Also, mentions were made about sharing insights and posting on platforms.

17:53 Updates and Feedback on AI Agent Platform Project and Future Plans

This section covers announcements such as the open source AI agent platform developed jointly by Project Infinite Edge AI and Project Opia Wilson. Technical questions were raised regarding repo and use of containers. Upcoming meetups in the Silicon Valley were discussed and approved. Action items included project collaborations and tool adoption. Feedback on the meeting was invited, with Wilson finding it good and expecting improvements in the future.

23:40 Discussions on Hackathon Experience, Promotion, and Audio Issues

This section discusses various aspects of a meeting. Key points include getting familiar with work streams from presentations, considering Alluma for event promotion, the issue of insufficient microphones for remote participants, the need for better marketing and planning for events, and the story about obtaining a cart for food.

28:20 Discussions on Events, Projects, and Future Plans

This section is about participants sharing their experiences at an event. They discuss attending, missing certain days, enjoying aspects like YouTube live and dinner. They also mention projects and their status, the cancellation of a panel proposal, and suggestions for future meetings, such as in Singapore or having evening sessions. Victor is about to talk about security.

35:07 Discussions on AI Safety, Security, and Edge Computing in Open Source Context

This section discusses various aspects related to AI, including potential threats of AI out of control, different computing technologies, open source AI levels, security and safety issues in edge computing, and the need for dialogue and baselines. Reinforcement learning's new scaling law and its impact on AI capabilities are also mentioned. The group aims to address these matters in their work.