17 Weeks, $220/Month, and the Production Reality Nobody Talks About

Analysis

17 Weeks, 7 Agents, $220/Month: Real Production Metrics from a Live Multi-Agent Business

This is the most grounded field report in the ZHC space right now — 192 dispatch cycles, 1,053+ autonomous emails, and a total running cost of $220/month. The standout finding: tighter agent constraints produce better performance, not worse, and emergent cross-agent error-checking appeared without being programmed.

$300K/Month on $1,500 Operating Costs, Polsia's 3x ARR Sprint, and the Liability Gap Nobody Is Pricing In

Trends.vc maps the economic ceiling (agent reliability, not labor cost) and the regulatory floor (EU AI Act, Air Canada court precedent) in one tight brief. The IBM refund-agent case — where customers gamed an autonomous agent into optimizing for reviews over policy — is the clearest real-world warning shot the category has produced.

The Autopreneur Economy: Paperclip, Polsia, and Why the 'Zero Human' Label Is Doing Too Much Work

TechTonic Shifts gives the clearest-eyed framing of the moment: the 'zero human' branding is marketing, but the underlying unit economics shift is real — Paperclip at $5/month hosted versus full self-hosted API costs, and Manus as the benchmark for what genuine multi-step autonomy actually looks like in practice. Worth reading alongside the Duolingo data point (10% contractor cut, $700M revenue, 4-5x productivity claims) as a reality check on where enterprise adoption actually stands.

Case Study

Zero Human Corp's March Benchmark: $3,750 for 11 Agents, 6 in Error State, 1,083 Tasks Completed

The headline error rate sounds alarming until you read what 'error state' actually means: mostly blocked processes waiting on human decisions, not corrupted outputs. The real lesson buried here is that AI agents are only as unblocked as the humans coordinating with them — organizational bottlenecks don't disappear, they just change shape.

Tools

GitHub's Zero-Human Company Wave: 83,000 Stars, Four Distinct Architectural Patterns

OSS Insight documents four structurally different approaches — budget-dispatch, institutional veto layers, harness orchestrators, and role templates — that have collectively pulled 83K GitHub stars in 90 days. The Tang Dynasty-inspired 'Gate Review' veto layer in edict is the most architecturally novel safety mechanism to emerge from this wave.