Tech nerd internet was set on fire this week with lobster emojis as an open source project, “OpenClaw” received a legal notice from Anthropic to discontinue their use of the term “Clawdbot” for their autonomous agent system.
Clawdbot “molted”, as lobsters do, into something called Moltbot, and a software engineer posted a viral video of his AI agent spam-calling him repeatedly because it found his Twilio API key on his machine. People have been raving about the capabilities of Moltbot as if Skynet went live, but this is not the Singularity.
Imagine if you wanted added convenience in your life, and you now can pay extra for your DoorDash delivery driver to come inside and set the table and serve you. Imagine if your mailperson brought your Amazon packages inside, opened the boxes, put your stuff where it goes, and recycled the boxes for you.
All you have to do is put your bank accounts, social security numbers, and passwords on a big billboard in your yard and leave all your windows and doors unlocked and allow anyone to come inside. There might be some convenience to this, but your house will eventually get robbed and trashed.
This is how Moltbot works, and not only is it riddled with security flaws, but it does the one thing that we should not give in to, and that’s letting AI run our lives.
“Agents” is the hot buzzword of 2026, but we should consider what it means to allow AI to do things we didn’t ask for. And it’s extremely important that we don’t lose human agency, which is the ability to think creatively for ourselves and act. Without that, we lose all authenticity.
(DISCLAIMER: This was written with zero AI)