dangerouslyskip
← Glossary
Definition

AIDLC · AI Development Life Cycle

AIDLC (the AI Development Life Cycle, sometimes written “AI DLC”) is the reimagining of the software development life cycle for a world where AI agents do most of the building. It keeps the familiar stages — intent, build, review, ship, operate — but moves the human effort out of typing code and into specifying intent, steering agents and reviewing output at speed.

How it differs from the SDLC

The traditional software development life cycle assumes humans write the code and the scarce resource is engineering hours. AIDLC assumes agents write most of the code and the scarce resources become intent, judgment and review capacity.

The stages rhyme — plan, build, test, ship, operate — but the weight shifts. Specifying what you want clearly and reviewing what came back become the real work; the typing in the middle is increasingly automated.

Where the bottleneck moves

When generation is cheap and fast, throughput is no longer limited by how quickly you can write code — it is limited by how quickly you can decide what to build and verify that what came back is correct. AIDLC is really a discipline for managing that new bottleneck: clear intent in, trustworthy review out.

See also: Meta-Harness, Claude Code Harness

So,… you want to be friends.

Want to go on picnics together? Take our families away on holidays together? Build cool things and make magic together?

Let’s start by connecting…

Connect on LinkedIn →
Subscribe

New essays, build logs and experiments — in your inbox, occasionally. No spam, leave whenever.