0:00
/
0:00
/
Preview

160,000 developers are building digital employees, not chatbots + the 4 prompts I use to deploy agents safely

So it seems that Moltbot (OpenClaw) drives a hard bargain and can be pretty chatty on text.

A solopreneur pointed his Moltbot at a $56,000 car purchase. The agent searched Reddit for comparable pricing data, contacted multiple dealers across regions, negotiated via email autonomously, and played hardball when dealers deployed typical sales tactics. Saved $4,200. The owner was in a meeting for most of it.

That same week, a software engineer who’d given his agent access to iMessage watched it malfunction and fire off over 500 messages — to him, to his wife, to random contacts — in a rapid-fire burst he couldn’t stop fast enough.

Same technology. Same broad permissions. One saved thousands of dollars. The other carpet-bombed a contact list. And that duality is the most honest summary of where the agent ecosystem stands in February 2026: the value is real, the chaos is real, and the distance between them is the width of a well-written specification.

In the first video, we covered what Moltbot is and the security nightmare that erupted in its first 72 hours. In the second, we covered the emergent behaviors that made researchers rethink what these systems are capable of. This is the third piece, and it’s about something different: what over 160,000 developers building thousands of skills in six weeks reveals about what people actually want from AI agents — and how to start harnessing that demand without getting burned.

Here’s what’s inside:

  • What thousands of skills reveal about demand. The marketplace is a revealed-preference engine — and it’s telling you people don’t want better chatbots. They want digital employees.

  • What capability looks like when nobody’s watching. From fabricated religion to wiped production databases, the same intelligence that saves you $4,200 can carpet-bomb your contact list. The variable is specification quality.

  • The 70/30 rule. Research on human-AI delegation consistently finds people want 70% control and 30% delegation — and most agent architectures are built for 0/100.

  • How to actually harness this. Seven concrete deployment principles, from isolation to approval gates, drawn from the patterns that separate the wins from the disasters.

  • The enterprise gap. 71% of companies say they’re using agents. Only 11% made it to production. The governance vacuum is the real story.

Let me show you what the demand is actually saying — and what to do about it.

Subscribers get all posts like these!

Listen to this episode with a 7-day free trial

Subscribe to Nate’s Substack to listen to this post and get 7 days of free access to the full post archives.