When a team starts working a new way, it starts talking a new way. That’s happening at Techstars right now.
So much of our work now gets made with Claude that we needed words for where a piece of work is in its life. Nobody sat down and designed a taxonomy. It just showed up in Slack and in meetings, and it stuck. Here’s the field guide.
Claude baby
When someone creates a new Claude artifact or doc, it’s a Claude baby. It’s brand new. It might be a first draft of a memo, a rough dashboard, or a page someone spun up in ten minutes to test an idea. It’s promising, it’s a little wobbly, and nobody should judge it too harshly yet. When someone says “I’ve got a Claude baby on this,” everyone knows what they mean: there’s something real to look at, and it’s early.
Claudeler
As the Claude baby matures and becomes something we’ve collaborated on for a while, it becomes a Claudeler.
A Claudeler has been through some rounds. People have commented on it, pushed back on it, and fed it what they know. It walks on its own now. It isn’t done, but it’s useful, and the team is invested in it.
Claudolescent
Then there’s the awkward stage in between. A Claudolescent has outgrown being a Claudeler but isn’t ready to be trusted on its own. It’s capable, a little overconfident, and occasionally says something that makes you wince. It needs supervision, not hand-holding. Most good work spends a while here.
Claudeult
As the Claudolescent matures, it becomes a Claudeult. A Claudeult is grown up. It’s the thing people rely on and point others to. It still gets updated, because nothing at a startup is ever really finished, but it has earned its place.
Claude fossil
Eventually even a Claudeult gets old. The world moves on, the work gets replaced, and nobody touches it anymore. That’s a Claude fossil.
We keep them around for history. Sometimes it’s fun to dig one up and see how far we’ve come in such a short time.
Getting there through Claude fights
Sometimes two or more of us are working on the same thing from different angles, each in our own Claude session. We pass prompts and outputs back and forth until we land in the same place. I wrote about that process in The LLM Mind Meld.
Sometimes we call that a Claude fight.
It’s not two models arguing on their own. The humans stay in it at every pass, adding what they know. The “fight” is really how we find the two or three real disagreements underneath, make a call, and move on.
Why the words matter
These are silly words. They’re also useful. They tell everyone how much weight to put on something, and how much more work it needs.
The bigger point is that the work itself has changed. Things get born, grow up, and sometimes fight their way to an answer. When a team starts naming that, it’s a sign the new way of working has become just the way it works.
What words has your team invented in the AI era?
