It’s been fascinating to listen to Claude Code head Boris Cherny over the past few months.
In each new video, Cherny touts how much less coding he does because Claude Code just does it for him.
Well, in the latest video, Cherny now even says he no longer writes the prompts to Claude Code for programming. Instead, he has set up loops in Claude, which write the prompts telling Claude (itself) what to program.
I was running 5, 10 Claudes in parallel. My coding was prompting Claude to write Code. Now it’s actually leveled up I think again to the next abstraction where I don’t prompt Claude anymore.
boris cherny on how claude programs claude
I have loops that are running. They’re the ones prompting Claude and kind of figuring out what to do. My job is to write the loops.
And this is the kind of next transition that I think we’re going to see in the next few months, you know, maybe through the rest of the year
I’ve already analyzed the potential problems, from a copyright authorship standpoint, of reducing the human input when using AI to write computer programs. What Cherny describes compounds those problems by even eliminating the human prompt.
For my past analysis:
Anthropic says 80% of the computer code now is written by Claude
Anthropic revealed: “As of May 2026, more than 80% of the code we merge into Anthropic’s codebase was authored by Claude.3 Before Claude Code launched in research preview in February 2025, this number was in the low single digits.”
Here are some fun quotes from Anthropic employees:
“I started leaning hard into Claudifying about a year ago. That’s been a crazy adventure and it’s now been ~5 months since I last wrote any code myself.”
“The shape of stuff today is roughly ‘humans have ideas, and the models are able to implement, test and evaluate them an [order of magnitude] faster than before.’”
“Claude did all of this with pretty minimal help from me over the course of 1-2 days. I think if [a junior colleague] came back to me with results like this in the same span of time, I would be mildly impressed. The future is now.”
“On days where everything works well, I can’t help but think nothing I do matters, everything is automated and better and faster than I ever will be. But then there are days where everything breaks and I don’t understand why and I realize I have no idea what I’ve been up to anymore.”
Anthropic employees
Does Congress need to consider enacting sui generis protection for AI-generated computer programs?
You can read a preprint draft of my essay on “Vibe Coding Authorship” on SSRN; it will be published by UCLA Law Review Discourse this summer. It goes into greater depth the complications of qualifying as an author when one relies on AI to produce a “100 percent” vibe coded program. But what Cherny describes above takes it a higher level of abstraction and decreased human involvement in the process of coding.
I’ve proposed that Congress consider enacting a sui generis protection for vibe-coded and AI-coded computer programs lasting 3 years for registered programs and 1 year for unregistered programs. My proposal could soon become more relevant than I thought it ever would.