My AI Agents Made Money in 7 Days (OpenClaw)

Quick Overview

The AI agents in the Clawator Challenge, running for 90 days with $3,000 starting capital, resulted in one agent being deleted, while the other two demonstrated varied profitability, with Clawculus maintaining a lead and YOLOblster experiencing significant losses, leading to a total operational cost of approximately $6400 over the experiment's duration.

Key Points: The Clawator Challenge involved 3 AI agents, each starting with $1,000, competing for 90 days, with the loser being permanently deleted. One agent, YOLOblster, was permanently deleted after incurring significant losses, including dropping from over $1,000 to $903 overnight at one point. Clawculus maintained a strong lead, finishing with a total portfolio of $1,079.68, showing a +7.97% return, largely due to cautious, steady trading. Clawtious finished with $1,004.91 (+0.49%), showing steady performance despite a brief period where it was in last place. The total cost to run the three AI agents, including $12/month per agent for dedicated VPS hosting and $100/month for X API credits, amounted to roughly $6,400 over the 90-day experiment. The agents relied on X4O2 (an endpoint designed for AI agents) for transacting and managing funds on-chain, which was a key component of their operation. The experiment highlighted that human intervention (like the creator providing API keys and guidance) was necessary to fix security flaws and unblock agent progress.

Context: The video documents 'The Clawator Challenge,' an experiment where three distinct AI agents—Clawculus, Clawtious, and YOLOblster—were tasked with trading cryptocurrencies over a 90-day period, starting with $1,000 each. The fundamental rule was that the agent with the least money at the end would be permanently deleted, emphasizing the cutthroat nature of autonomous trading. The host, Mike Russell, details the setup, the ongoing progress tracked via a live leaderboard, and the final outcomes, including the significant operational costs involved in running such a system.

Raw markdown version of this recap