The Trust Ladder: From Summarize to Act — AI Readiness, Part 2
- Trisha MerriamHost
- Chris CarolanHost
- Erin WiggersHost
October 16, 2026
Part 2 of AI Readiness climbs the trust ladder: summarize, recommend, prioritize, draft, decide, act. Chris Carolan, Trisha Merriam and Erin Wiggers on why trust doesn't mean it's correct, why you don't start at the sixth rung, and why the climb starts with what you know.
Moments from this episode
19 short cuts from this conversation.
Key takeaways
The trust ladder has six rungs: summarize, recommend, prioritize, draft, decide, act.
Trusting AI to give you something to work with is not the same as trusting that it is correct. Trust doesn't mean shipping what it hands you as gospel.
Treat new AI work like a new hire: a safety zone, not a penalty box, with the work checked before it goes any further.
Pilot with a small cohort first and make one life-changing difference for them. An advocate's word means more than the word of the person who brought the tool in.
Don't start at the sixth rung. Start there and AI will give you three options for how it could act.
Treat AI as a strategic thinking partner, not a tool. A tool has a job; a partner has to get to know you first.
Summarize first. Ask AI to show you how it understands the context before you ask for an outcome; the slightest correction in framing can be the difference between a great outcome and a terrible one.
"Tell me what you understand to be coming from this PDF" and "read this PDF and give me the top five tips for how to do this thing" are different requests. The second skips straight to draft, and AI will give you a compelling draft every time.
Rushing up the ladder costs you. When Chris Carolan pushed his AI assistant straight to acting, one transcript job burned about 20% of his usage.
When rushing to act goes wrong, switch AI to shadowing: Chris had his assistant shadow his existing Claude Code setup for the rest of the week and tell him where the boundary should be.
Recommending, prioritizing and deciding rest on judgment and experience, so they don't transfer well to another person or to an AI.
The farther the goal, the more the starting frame matters. Big work, like planning an app's architecture, needs extra diligence early in the conversation.
Before AI acts, it needs to know whether the context has changed. Asking AI to act takes a lot of connectivity.
The only place you can really trust AI at first is in the domain you know. Don't ask for a summary of a topic you've never learned.
AI readiness is based on trust. If the team doesn't trust, you're not ready.
Chapters
- 00:00Part 2: the trust ladderTopic
- 01:28Trust doesn't mean it's correctTopic
- 02:34What do we do when we hire somebody new?Question
- 07:24How do we do that with AI?Question
- 08:44One life-changing thing for a small cohortTopic
- 10:27Is there a trust ladder component to that?Question
- 13:20Six rungs: summarize to actTopic
- 14:44Don't start at the sixth rungTopic
- 19:26Aren't those two prompts the same?Question
- 21:37Rushing to act, then a week of shadowingTopic
- 25:35The farther the goal, the more framing mattersTopic
- 28:19Did you skip the summarize step?Question
- 30:36Prioritizing, and where the boundary sitsTopic
- 33:56Before AI acts: has the context changed?Topic
- 36:30Talking to AI out loudTopic
- 38:48Asking Nectar: where does that mistake sit on the ladder?Question
- 42:23GPT's voice lead, and an always-on assistantTopic
- 45:32Trust it first where you know the workTopic
- 49:29If the team doesn't trust, you're not readyTopic
- 50:42What's next in the seriesTopic
Transcript
This transcript was generated by AI and may contain errors. It is provided to help you skim and search the conversation.
Keep learning
You just watched the work. Now build it.
The next step is doing it yourself — live, with people on the same path.
Get the next conversation in your inbox
One signal-dense note when a new episode lands. No noise.