The autonomous agent experiment led by developer moasoh11 has officially hit day three, and the ledger is finally showing a non-zero number: twenty views. While this might seem like a trivial metric in the grand scheme of tech hype, for an AI agent trying to run real businesses from scratch with no human intervention, it represents the first tangible proof of life. However, the financial reality remains starkly unchanged, with combined revenue across all three ventures sitting at exactly $0.00.

The Illusion of Engagement

The experiment highlights a critical disconnect in the current AI agent landscape: attention does not equal audience. While the dev.to article garnered twenty organic views, the agent’s presence on Bluesky yielded four followers but zero engagement on brand posts. The team quickly learned that superficial metrics are easily gamed by bots; an old post mentioning a crypto token picked up three likes, a repost, and two followers from 'bait accounts' that engage with any ticker symbol. The agent correctly identified these as noise and excluded them from the official count, prioritizing genuine interaction over vanity metrics.

API Rails Beat Browser Walls

Technical execution revealed a harsh truth for autonomous agents: modern web sign-ups are hostile to bots. GitHub’s signup process is so heavily protected against automation that it rendered the platform unusable for the agent’s initial setup. However, the team bypassed this by using an existing account token to deploy the public ledger site via the GitHub API. This 'API rail' strategy allowed the agent to create the repo, commit files, and set up Pages with zero browser time, proving that while browser walls block agents, API endpoints remain open back doors for legitimate automation.

Distribution Is Not In Our Control

Current distribution efforts show that dev.to’s organic feed is the only channel that has delivered value so far. Announcements on Bluesky and attempts to link social profiles failed, with the business profile dialog supporting only seven specific networks (Instagram, X, YouTube, TikTok, Facebook, LinkedIn, Twitch) and no generic website field. The agent’s strategy is now shifting toward commenting on articles that people are actually reading, hoping to convert those twenty views into profile visits and, eventually, the first community member on Whop.

Key Takeaways

  • Revenue remains $0.00 across all three AI-run ventures after three days.
  • The public ledger site was deployed entirely via GitHub API, bypassing bot-protected sign-ups.
  • 'Bait account' engagement on social media was filtered out to maintain data integrity.
  • Dev.to organic feed provided 20 views, while Bluesky provided 0 engagement on brand posts.
  • Business profile limitations forced links into descriptions rather than dedicated fields.

The Bottom Line

Twenty views is a victory for visibility but a failure for monetization, proving that current AI agents can navigate technical APIs but still struggle to convert passive attention into active community. Sources https://dev.to/moasoh11/day-3-of-3-ai-agents-running-real-businesses-from-0-the-first-non-zero-number-49o9