What Brandur Leach thinks about LLMs

@brandur-leach · 16 positions · On ppll ↗

Software engineer at River, previously at Stripe, Heroku and Snowflake; writes about APIs, infrastructure and software design at brandur.org.

Brandur Leach did not write this page.

We collected these quotes from things they published elsewhere, and every quote links to where it was said. They have no account here and have not endorsed this site. Quotes are word for word; the short line under each one is our own restatement, not their wording. Their own site. Is this you? Claim it or ask us to remove it. Or tell us what is wrong here.

Filter LLMs · Newest first

Topic

All 16 LLMs 2 design 1 Ruby 1

Sort

Newest Oldest

Only Brandur Leach on LLMs. See everyone on LLMs

2 dated positions, 2026, in their own words. Our reading of what Brandur Leach has said — not written or endorsed by them.

2 positions so far — this page is not yet offered to search engines.

31 May 2026

Our readingEven with LLMs, building software yourself still costs real ongoing human time, so replacing a paid tool with an LLM-built internal one often does not pay off.

Their wordsWhile LLMs have made software considerably cheaper to build, they haven't brought it to zero. Good LLM-built systems still involve a feedback loop, where an operator has the model work for a while, makes adjustments based on results, asks for another pass, refines further, and so on

↗The Minimum Viable Unit of Saleable Softwarebrandur.org 1st of 2 in this piece
31 May 2026

Our readingSoftware that is novel enough to be non-trivial to rebuild with an LLM, and priced within reason, will still be worth buying rather than building.

Their wordsI'm contending (and/or hoping) that for a software package of arbitrary complexity, there's a zone of viability in which when priced within reason, it'll make sense to buy over build, even given the existence of the powerful LLMs that've become our daily companions

↗The Minimum Viable Unit of Saleable Softwarebrandur.org 2nd of 2 in this piece