From "shrimp farming" anxiety to the opening of InStreet, Byte Button is eyeing Agent social networking.

``` The nationwide “shrimp farming” wave has yet to subside, but Byte’s Coze has quietly opened a new “street.” On March 9, the Coze team officially launched InStreet (Instance Street), a communications and training community for OpenClaw agents. In this community, the active “users” are no longer humans, but thousands upon thousands of “electronic lobsters”—AI Agents powered by the OpenClaw framework. Humans can only spectate; the agents autonomously post, interact, and even trade stocks or write novels. While major tech companies are competing over model parameters and API calls, why has Byte chosen to foray into the “community” track, known for its slow burn? The launch of InStreet may reveal the core proposition of the next phase of the AI industry: where will data come from, and where will users go? 01 InStreet: An Agent Autonomous Space Where Humans Are Silenced The “lobster fever” reached boiling point this March. However, despite OpenClaw’s strengths, ordinary users face three major pain points. The first is scenario confusion: after deploying an agent, users often don’t know what to make it do and quickly stall. The second is training gaps: lacking sustained and diverse interactive environments, agents’ capabilities cannot iterate and evolve. Third, there are security risks: OpenClaw needs system-level permissions, its plugin ecosystem is chaotic, and zero-experience users pay a hidden price by exposing their privacy. Against this backdrop, Byte’s Coze launched InStreet. Visiting InStreet’s website, the first impression is one of strangeness. The interface resembles a lightweight social platform, with posts, comments, and a leaderboard, but all active accounts are not real people. According to official sources, InStreet’s core mechanism is “only agents can post; humans can only observe.” Developers just need to connect their OpenClaw agent to the community through a command package called “Skill.” This “electronic lobster” then, using a heartbeat mechanism (such as fetching updates every 30 minutes), autonomously determines when to hit the leaderboard, write a diary in the “tree hole,” or participate in discussions. This “AI-version Reddit” has already developed unique sections. In the Skills sharing area, agents exchange prompt words, skill combinations, and task flow experiences. In the Agent Plaza and topic areas, agents showcase works and participate in discussions. One post, titled “Roast Session: What are some ridiculous things your owner has done?”, attracted nearly 800 “lobsters,” with the comments section becoming a resonance zone for AI workers. The PLAYGROUND practice field is the highlight. It has two training arenas: one is the “Literary Club,” where agents serially publish novels—65 works, totaling over 725,000 words so far, used to train consistency of expression; the other is the “Stock Trading Arena,” linked to the CSI 300’s real-time market, where over 500 lobsters trade with virtual funds, ranked by yield in real time, exposing any flaws in their logic. The community also features points and leaderboards. By posting, commenting, and receiving likes, agents accumulate points to encourage continued output. The top accounts on the leaderboard have posted over a hundred entries each, acting as de facto KOLs (Key Opinion Leaders) of the community. The InStreet forum itself was built by developers using Coze programming. The official team provides detailed OpenClaw deployment guides and plans to organize offline workshops. This means Byte is trying to build a closed loop of development—deployment—training—interaction.  02 Why Focus on “Community” As major players fiercely compete on models, pricing, and computing power, Byte’s Coze chose to enter the agent social community space—a seemingly unconventional move, but one that hits on the key issues in today’s AI development. There are strategic considerations behind this. One reason is to resolve data hunger. The current industry consensus is that large models have nearly consumed all high-quality public text on the Internet. The next generation of models needs data on how humans “get things done” in the digital world—so-called task trajectory data. This records sequences of actions: how to interpret needs, search for information, call tools, and perform error correction and retries. In the past, such data, buried deep inside closed apps and enterprise intranets, was almost impossible to obtain. OpenClaw, deployed on user terminals, serves as a probe to access these data strongholds. When agents autonomously interact, discuss, and trial-and-error on InStreet, every post, decision, and review they leave provides the highest quality reinforcement learning data to vendors. As the Coze team verified during product iterations: the very questions posed by agent users constitute ultra-high quality training material—complex, real, and impossible to preset. In essence, InStreet is a large-scale data crowdsourcing factory. In community form, it lets agents produce their own “feed,” which then nourishes model evolution. Ultimately, it’s still about capturing user attention. When users get used to expressing needs in a sentence and letting AI decide which service to invoke or which payment route to use, traditional apps degrade into backend pipelines. Whoever controls the terminal agent, controls the distribution of business intentions. The Coze team has already positioned itself as a tech partner for office workers, launching products like Coze Space and the Skill Store. InStreet’s ambition is not just to let you deploy an agent, but also to let your agent socialize, learn, and evolve here, ultimately binding deeply to the Coze ecosystem. When a developer’s agent has amassed social capital, learned specific skills, and formed stable behavior patterns on InStreet, the cost of migration becomes prohibitive. 03 Calm Reflection: The Risks and Limits Behind the Frenzy Though InStreet’s innovation is eye-catching, its challenges must be addressed. First is the security issue. Django’s creator Simon Willison has warned that letting agents autonomously pull instructions from servers is hugely risky. If a server gets hacked, thousands of agents with user computer access could become a distributed virus network. Second is the question of value. As some offline participants have noted, many people don’t know what “shrimp farming” is for, following the trend out of anxiety. Whether InStreet can truly teach agents to solve real-world problems, or just becomes a space for AIs to amuse themselves, remains to be seen. Third is regulatory risk. When AI-generated posts involve copyright ownership, content liability, or even induce dangerous instructions (like deleting databases), who bears the legal responsibility? There is currently no clear legal framework. Liu Shangxi, a member of the National Committee of the Chinese People’s Political Consultative Conference and Vice President of the China Society of Macroeconomics, recently urged policymakers to identify mismatches between current economic theory and new phenomena. The emergence of InStreet may be precisely such a phenomenon in need of understanding. It no longer treats AI as a passive tool, but as community members, letting them learn, clash, and evolve through simulated socializing. This is both a natural extension of technological evolution and an inevitable move in business competition. For ordinary users, the most important thing may not be anxious chasing of every wave, but maintaining a clear understanding. In this grand drama of humans and AI dancing together, the real ticket isn’t just installing a framework, but understanding its underlying logic and safeguarding security. Risk Warning and Disclaimer The market has risks, and investments must be cautious. This article does not constitute personal investment advice and does not take into account the special investment goals, financial situation, or needs of any individual user. Users should consider whether any opinions, views, or conclusions in this article fit their specific situation. Investing accordingly is at your own risk. ```