@bcherny · posted as a claude artifact · jul 16, 2026

steps of AI adoption a ladder from zero agents to a thousand

Boris Cherny, creator and head of Claude Code, keeps seeing the same four steps as teams adopt AI.

His opener: one person is 10x'ing their output while the rest of the org hasn't caught up.

He mapped the climb — the tweet, the artifact.

01 · the ladder

Five rungs, tap to climb

Each step names your role, an agent headcount, the thing that unlocks, and the bottleneck that gates the next rung.

The headcount is exponential: 0, one, ten, a hundred, a thousand-plus.

Your job title mutates along the way.

height = agents running, log scale

All table content from the artifact, condensed. The "products that help" column is omitted here on purpose — it's an all-Anthropic shopping list, and that's an on-air talking point, not a diagram.

02 · the engine

Tokens don't move you. Guardrails do.

The thread's core claim: there's no one right path, and spending more on tokens doesn't advance you a single rung.

Progress is a loop — find the bottleneck, build the guardrail that breaks it, climb, repeat.

Every bottleneck on the table is a human quantity: attention, review, trust, decision throughput.

The model never appears in that column.

bottleneck always human guardrail buys trust next step 10x the agents a new bottleneck appears "tokens aren't enough to move you forward"
read this on air · the trap

"The trap is scaling agent count before the loop has earned widespread trust." — step 3's bottleneck, and arguably the whole episode in one sentence.

the ROI reframe, from the thread

Usage dashboards measure "activity, not return."

The better question: "would you have spent engineering effort on this anyway?"

"If yes, how much and what would it have cost in manual eng-hours? That's your return."

same-day addendum · context engineering

Thariq Shihipar says Anthropic removed over 80% of Claude Code's system prompt for the newest Claude models, with no measurable coding-eval loss.

That rhymes with Boris's ladder: better models need fewer permanent rules and better context boundaries.

The new advice is progressive disclosure — smaller skills, lighter CLAUDE.md files, and context loaded when the model actually needs it.

Thariq's post · Anthropic blog.

03 · the poll

What step are you on?

Boris closed the thread with a question built for a podcast: "Anthropic is on step 3 and pushing toward 4."

"Personally, I just hit level 4. Curious where you are -- what step is your team on?"

Answer five questions live.

Argue with the verdict.

Then ask the audience for theirs.

04 · the replies

The thread was half seminar, half bug tracker

1.3M views brings every kind of reader.

The replies split into skeptics with real objections, rival taxonomies, and — deliciously — bug reports about the artifact itself.

Tap the amber buttons for Boris's answers.

Jay McCormack · reply

"I've seen people who I expected to adopt AI almost actively avoid it... There is clearly a camp that 'gets it' and another apathetic group with no interest."

the ladder assumes you want to climb
yash · reply

"Great, but i think agents do make mistakes, in that case, how does this look like? backwards and keeps moving or does it require additional support?"

where's the down escalator?
Charlie Bailey · reply

"Would genuinely love to take a course on moving from level 0 to 'AI Native'... Literally don't even have a mental model of what this would look like irl."

the demand side
Steven Zimmerman · rival taxonomy

"I call the point where you stop reading code and just live in the agents 'AI Native,' and the point where PRs become continuous 'AI Industrialized.'"

everyone has a ladder
Behnam · reply

"I just have one question: Who added 'LOAD BEARING' to Fable? 😅"

guardrails discourse
@bcherny: "When you take away the belt and suspenders, you're absolutely right that what's left is load bearing"
fomoless · reply

"This tread looks like it was made by Claude"

disclosure, 2026 style
@bcherny: "Artifact was me, thread was me+Claude"
smoker bruda + Chad + Sid · bug reports

"Page not scrollable on safari on Mac" · "This doesn't load on android" · "scroll issues / buggy artifact"

the replies became a bug tracker
@bcherny: "Looking. Here's a doc in the meantime:" — and ships a Google Docs mirror of the artifact, twice.
Chris Thames · the best one

"this is great, but Claude can't crawl your artifact... claude should be able to crawl a public artifact?"

— with a screenshot of Claude failing to fetch the page.

our prep hit the same wall
the meta bit · save for late in the segment

The manifesto for the thousand-agent future shipped as an artifact that wouldn't scroll on Safari, didn't load on Android, and couldn't be read by Claude itself.

Even prepping this segment, the automated fetch got an empty JS shell — a human had to open a browser.

The ladder is real; the rungs are still wet paint.

05 · on air

Three angles to argue about

pick fights in this order
human bottlenecksAttention, review, trust, throughput — never the model. Same punchline as mahesh's four horsemen: waiting for a smarter model doesn't climb the ladder.
the vendor ladderEvery "product that helps" is an Anthropic SKU. Steelman it, then ask: what does the vendor-neutral version of this maturity model look like?
the motivation taxStep 3 makes you a manager of managers. RDEL: 41% of teams are less motivated, "authoring → reviewing." Is the top of the ladder where the energy drains?
callbacks to last episode
the featuresBoris's climb mechanism is literally last week's changelog: auto mode, /loop, /batch, dynamic workflows, worktree isolation, runaway caps.
the theorymahesh said agents fail like distributed systems; this table is the org chart you build so the failures land on guardrails, not pagers.
the fieldStep 0→1's blockers — security process, cost-per-token fear, no technical voice in the room — are the FDE arms race in miniature: 71% say readiness, not tech.
zoom out · why it matters now

The gap everyone's staring at right now

2026 is the year orgs moved past the lone 10x engineer and hit the org-wide adoption wall.

Five labs committed $8B+ to deployment teams this quarter, and surveys put the barrier at organizational readiness, not the model.

Boris's ladder is a map for exactly that gap — which is why it got bookmarked more than it got liked.