Anthropic posts are always great to read, which is wild if you think about it. Worthwhile on progress:
There isn’t full consensus among staff at Anthropic, but many believe that the Claude-written code was still worse in quality than human-written code at Anthropic in late 2025, and is roughly at parity today. We expect it to be better within the year.
And this
we ran a retrospective analysis, and found that an automated Claude review of every change to our codebase would have caught roughly a third of the bugs behind past incidents on claude.ai before they ever reached production.
# Jun 4, 2026
Hi, I'm Peter. (About me.)
I am an indie builder and consultant, and now live in Madrid after long stints in Medellin and New York. It's really nice here.
I've been blogging since 2002.
I help teams with technical AI consulting with a human centered perspective at helpful intelligence.
I run Sputnik Intelligence, data for your agents on millions of newsletters and podcasts, with full transcriptions and more. It's really good.
I also run Parse for Artisans, a document to markdown parsing service designed for Laravel.
I teach AI skills to UX-adjacent teams at Model Context Experience.
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.