I help experienced independent consultants, advisors and small expert-led firms that already deliver valuable work, but whose online presence fails to show why their expertise is different and worth paying for.
Before, I dismissed every claim that writing with AI could be legitimate. Then I read something that made me change my mind in a single morning.
How could a new idea make me change my mind after I had taken a position on it so strongly?
Here, with the help of other authors’ perspectives, is the explanation of what it takes to have an idea that people take seriously and adopt.
For advice to get acted on, it needs to carry three things:
a new frame,
reasoning the listener can walk through, and
enough context to give it weight.
Table of Contents
1) The Advice That Changed My Mind in a Single Morning
2) The Three Things Your Advice Needs to Be Taken Seriously
3) Why Your Recommendations Get Ignored
4) Four Ways To Make Your Advice Land
1) The Advice That Changed My Mind in a Single Morning
I traced back the day I bought into an idea that I had honestly never considered, and that in pure principle I would have instantly discarded. February 26 of this year.
The idea that I found on that day contradicted a position I had held for quite some time and which I had published more than once.
So I went back and asked myself what was so special about that idea to make me change my mind and what context it came through to carry such influential power.
Because that’s what every independent expert would want to know too: how do you make your advice land when there are a thousand other experts trying to do the same?
How do you make your recommendations and new ideas land in a way that people take them seriously and act on them?
2) The Three Things Your Advice Needs to Be Taken Seriously
You give good recommendations. To clients, to readers, to the people who ask for your opinion because you have had a long and rich professional experience. But most of the time, nothing happens. A few likes, an email or two, but nothing more.
I have felt this frustration myself, many times, and I have no fear in admitting it. I know I have good, extended and deep professional experience. I know I have ideas that are useful, practical and different from what everyone else is recommending.
Why then do so few of the people who read me turn into passionate readers, and eventually into customers? Why do my recommendations land less widely and less deeply than, in my own critical evaluation, they should?
While researching this very issue with AI and going back through my own saved notes and highlights, I did find an answer.
For any advice, idea or recommendation to land successfully and turn readers into potential customers, it must have three very specific traits that make its weight become immediately visible, familiar and trustworthy:
a) A reframe.
A new, non-obvious way of looking at things that changes the perception and understanding of what the issue really is about.
b) The messenger.
Who you get the message through plus the context, relevance and trustworthiness surrounding that source.
c) The decision trail.
The decision trail showing a verifiable / testable reasoning strategy behind it.
Whether my ideas get picked up seriously depends on these three elements much more than on how well I write them.
a) A New Frame Makes the Solution Visible
Take this example. How I thought about AI writing was on record. I had written and taken a strong stance about it publicly.
I had written before that perfect AI writing is plastic and that AI-generated content felt too polished, too self-conscious, too staged for me to trust it.
The conclusion I had arrived at was pretty unambiguous:
“You can’t find your voice if you delegate to AI to write your content.”
Three weeks later I had sharpened it into a rule:
“You can teach AI to use your writing style once you have it, but not vice-versa.”
Most opinions on writing with AI that I came across were on a pretty recognizable continuum. On one end, people were clearly saying that writing with AI was the worst decision a writer could ever make. Evil! On the other: here is the secret recipe, the magic prompt that makes AI write exactly like you.
I had answered both camps in print across three issues in the summer and autumn of 2025: On July 29, 2025, I wrote that AI was not my enemy, and I described using it to free brain space for sense-making and to hunt down opposing viewpoints. Research, yes. The writing itself stayed out of this picture.
Five weeks later, on September 3, I closed that door explicitly: “You can’t find your voice if you delegate to AI to write your content.”
Three weeks after that, on September 24, I turned it into a rule with an order of operations: “You can teach AI to use your writing style once you have it, but not vice-versa.”
In the same article I even built a metaphor to defend it, the climber and the helicopter tourist, where both reach the same peak but only the one who climbed can see the view with any real resolution.
Then this year, on February 26, 2026, while checking my morning email, my Yutori Scouts custom news brief arrived with this title: “Curation as Authorship.”
It pointed to a paper published on Springer that very same day: The Apt Curation Model by Tiago Vieira Rodrigues. The first five words of the abstract contained the name Luciano Floridi and his concept of “distant writing.”
“...distant writing reframes rather than diminishes the author’s role, shifting epistemic competence from direct composition to second-order curatorial acts:
architectural design,
dialogical refinement,
evaluative verification, and
integrative synthesis.
…which positions the human author as the central epistemic agent, not as sentence originator, but as virtuous curator whose competence makes the text an apt epistemic performance.”
I read the abstract and stopped. Wow. This completely reframed how I look at authorship, and it made a lot of sense.
Plus it was sitting in the first line of a peer-reviewed paper, with its reasoning fully on display, and living proof attached that the idea was indeed traveling: a Brazilian researcher had already picked it up and built on it.
Floridi was making a strong, bold claim: the author of an AI-assisted text writes “at a distance” from the sentences. His authorship lives in the meaningful decisions: how the problem gets framed, how the exchange gets steered, how every claim gets verified, how the result gets integrated with his own body of work.
Floridi’s analogy is the museum curator. She paints none of the paintings, yet the exhibition is entirely hers, because every meaningful choice in it is hers.
Rodrigues took this further and elevated the four specific curatorial acts I listed above into the actual proof of an author’s true competence.
So, by reading that paper, my whole stance against AI writing collapsed instantly. From there on, the key question to answer became: who made the meaningful decisions in this text, and can I double-check them?
Here was a method that let me check honestly whether I truly am the author of what I publish, beyond the simple typing of the words.
At the same time, my old concern about AI writing remained intact: keeping an authentic voice still matters.
What I did next: I devoured the paper. I saved the passages that struck me, took notes, and kept digesting them for weeks. By mid-April, Floridi and Rodrigues had already earned their place in my curation lineage map and those ideas and notes became a full article on AI and authorship, published on May 27.
The practical lesson I learned was:
the framing determines the shape of the solution.
For a long time I have framed curation as the best way to establish authority, trust, and credibility. It is an interesting frame, and it is mine. But if you do not feel your problem in those terms, my recommendation stays invisible to you, however good it is.
Proper framing places the issue under the perspective that matches your reader’s real problems and fears. Do that, and the same solution they scrolled past becomes the obvious one.
Go deeper:
Tiago Vieira Rodrigues, “The Apt Curation Model” (Philosophy & Technology) / Luciano Floridi, Distant Writing: Literary Production in the Age of Artificial Intelligence (2025)
b) Context Gives Weight to the Recommendation
Tyler Bainbridge runs what may be the world’s largest taste anthropology project. His Perfectly Imperfect has collected over 800,000 human recommendations, none of them sorted by an algorithm.
When Collab Fund interviewed him this past March, the title of the piece asked almost exactly this article’s question: “Who do you trust to tell you what’s good?”
His answer:
“A great recommendation that’s shared without personal context, is ultimately worthless.
I need to know who you are, why you like this thing, and what else you’re into, in order to make an informed decision on ‘is this worth my time.’
There’s far more weight to a recommendation if it comes from Lou Reed (i.e. his Yeezus review for Talk House, one of the all time best pieces of music writing, ever) than a co-worker, a friend’s recommendation is more valuable than a stranger, and so on.”
Sari Azout saw the same mechanism years ago in her essay on boutique search engines: services with deliberately constrained supply feel “less like the Yellow Pages and more like texting your friends to ask for a recommendation,” and that constraint, she writes, is the foundation of their key strength: trust.
(Small disclosure: Sari is the founder of Sublime, the curation tool where most of the quotes in this article were collected. I use it daily.)
I agree with both. And I am also the living proof of the limit. A popular name, by itself, is not a guarantee for an idea to take hold. I scrolled past Luciano Floridi, a philosopher with four decades of track record, for years. A name without context is noise, even if he is very good.
What finally made it possible for Luciano Floridi to break into my info-sphere was the context in which I found him embedded: a framing with a name, a decision trail, and the evidence that serious, academic people were already building on it.
The lesson to take home from this is that trust gets accumulated before the moment when it is needed, and that context gives the weight to every recommendation. So much so that recommending fewer things, with more of yourself attached, beats recommending lots of supposed good stuff.
Go deeper:
Collab Fund, “Who Do You Trust To Tell You What’s Good?”
Sari Azout, “Re-Organizing the World’s Information”
c) Visible Reasoning Makes Your Judgment Trustworthy
Showing how you choose and why makes your recommendations have real weight.
That’s what people generically refer to as “having taste,” because to have it you must be able not only to express it coherently but also to explain its whys. Taste is not just a feeling or an opinion. It’s a precise choice based on a long-refined ability to pick.
Anu Atluru gave me the cleanest definition of taste I have found so far, in a conversation published on The Sublime newsletter last November:
“Taste is discernment expressed: informed judgment plus the courage to show it. It only exists once you make a choice.‘Discernment’ as in
having a judgment based on a strong set of references,
understanding the differences between things, and
being able to articulate why you like or dislike something.
‘Expressed’ as in you have to make choices.
You can only demonstrate taste if you do something with it.”
Henry Oliver compresses the first half even further: taste is knowledge. Quoting Susan Sontag, he reminds me that to understand any work we must be able to say how it is what it is, and that requires having read, seen, and heard enough to know where each new thing fits.
This is how Luciano Floridi’s concept found a way to earn serious attention. He shows his whole thinking, step by step. Every word is explained. Every point is one I can test against what I have lived. I did not have to trust him. I could follow his reasoning and check it for myself. That is a very different thing from someone just telling you “believe me” or “this is the right way.”
And expressing such judgment does cost something, because sometimes that judgment can show you’re wrong publicly.
My AI slop position was visible judgment too: three issues of TRUSTable with my name on them that defended a position that then I abandoned in a single morning.
In truth, I must say I felt good about it. I felt good because I am not here to defend any specific position, claim or idea. I am here to explore, learn and find out. When evidence or experience shows me something new that is valuable, or that contradicts what I had believed until then, that is the moment I write about it.
If I ask you to trust how I choose, the least I can do is to show you how those choices affect me and for how long I stick to them.
Because only by showing your choices, even the ones that later prove to be wrong, you can earn someone else’s trust. Being transparent and honest in public pays more than being silently discreet while showing only your victories and successes.
Go deeper:
3) Why Your Recommendations Get Ignored
The three elements explored above converge into one direction.
An idea gets through when:
a) it hands the listener a new frame instead of a claimed better option (Floridi),
b) it carries its context with it (Bainbridge, Azout), and
c) its reasoning is visible enough to be tested rather than believed (Atluru, Oliver).
Your advice and recommendations are ignored when:
it argues a solution inside a context that your client already knows well.
it arrives from someone who carries no context for that specific listener, and I have shown you that even a famous name can have none by default.
the decision trail behind it is hidden, leaving the reader with nothing to check, verify or test.
Sangeet Paul Choudary, whom I curated in Deep Reads #02, puts the stakes plainly: AI can now give anyone ten plausible answers, so “the final step - which one feels right - still belongs to someone you trust. That trust doesn’t come from logic or even pattern recognition. It comes from accumulated long-form discernment, taste, and context.”
One more thing.
If I look back I could presumptuously say that for a very long time I have argued, as a practitioner, that curatorial acts are the raw material from which authority gets built today.
Up until the day I ran into a peer-reviewed paper from a Brazilian researcher (building on a new concept by an Italian philosopher of information I had never heard of before) who had arrived at the same conclusions.
When I saw that two independent and disconnected roads reached the same conclusion, I started to pay immediate attention.
What changed in my own practice as a consequence of this realization is having a wider space and communication framework to move through.
In practical terms, the shame of having AI write parts of what I want to say went away. I understood there is no dishonor in distant writing with AI, as long as the key curatorial acts remain fully mine:
the direction I take,
the analysis I do,
the choices I make,
the integration with everything I have written before.
At the same time, my dislike for excessively polished, impersonal text that wants to perform remains intact. But that is a separate problem from authorship. Believing that you must type the sentences to be the author of something turned out to be a dogma I had believed in without ever analyzing it in depth. So much so that now that I have realized it, I have a hard time believing so many people strenuously defend a position that (like me) they had given for granted without ever truly questioning it.
The consequence of this realization, for me, has just been more space and breadth on how I choose to express and communicate something.
On some topics I have a lot to say: I jot down every idea I have, hand the pile to AI, we discuss the structure, what is good and what is bad, and we refine it, partly me, partly the machine. On other topics I have fewer notes but a specific destination I want to get to: once identified, I let AI suggest multiple possible trails to me and ask me questions, so that I can use my crystallized intelligence to pick the stories that best illustrate my vision while AI can help me find better words and analogies than my raw unfiltered writing.
Yes, I believe I have become richer in options, faster in the interaction, and my capacity to turn ideas into finished concepts has grown significantly.
That is in the end what an idea that truly lands does: it expands what you can see, while leaving every choice to you.
4) Four Ways To Make Your Advice Land
Four practices I am taking from this for my own consulting and for this newsletter:
1) Reframe Before You Recommend
Move the question before answering it, and give the new frame a name your listener can carry. “Distant writing” traveled from a philosopher to a Brazilian academic to my inbox on the strength of its name and frame alone.
2) Attach Yourself To the Recommendation
Why this, why now, what you rejected, what else you love. Context is the weight. A recommendation without it is noise, whoever puts his signature under it.
3) Show the Decision Trail
Make your reasoning walkable so people can test it instead of having to believe you. This is Paradata working as trust infrastructure.
4) Accumulate Trust Before You Need It
Every visible, honest decision you publish is like a deposit you make in your trustworthiness account. Then, when the moment comes to recommend something that matters, you withdraw from that visible, long-standing decision trail. Your choice, no matter what it is, has a lived history, a path, a true story to stand on.
Paradata
Intent / Why: This article addresses a frustration I have felt strongly myself. If I have good experience, good ideas, non-obvious things that are genuinely useful to my readers, why do so few of those readers convert into passionate fans and potential customers?
Why do my recommendations land less widely and less deeply than, in my own critical evaluation, they should?
I wrote this piece to understand that mechanism, starting from the one time this year I sat on the receiving end and watched a single idea break through my own resistance.
Process: The topic of this issue was chosen by me, against the advice of the best ChatGPT and Claude models. I had asked both to analyze my performance data, my Sublime quotes and notes archive, and my LLM-wiki personal knowledge base, to propose the best potential topic for this TRUST-able issue.
The AIs came back with three candidates: a decision-trail piece on how I chose the 5 curation tools I actually pay for (out of 40+ tested), a first client case story, and this one, which they ranked last. Yes, this title was on the suggested topic list, but the choice was mine. I insisted on focusing on this topic, the recommendation problem, because it is a frustration I actually feel, because the citations were already sitting in my Sublime library, and because it continues the Deep Reads format which my stats say readers like you enjoy the most.
In the very first AI-generated draft, AI had reconstructed my story as “strangers told me AI writing was fine for two years and I ignored them because of who they were.”. But that wasn’t true at all, and I corrected it: my position against AI writing dissolved because of a genuinely new frame, and the fame of its author had nothing to do with it, since I did not know him.
Second: the first draft ended with me as “a more critical AI user” who still hand-writes everything. Not true. There are times I write every single word by hand, but that’s not the rule. In fact the real change has been more freedom and range, as you read above. This article you just read is built on the corrected story, and the corrections themselves are a small demonstration of its point about evaluative verification.
I then asked for successive more refined drafts, directing the writing style, always toward less prose and performance, and more toward facts and good flowing narrative.
Once I saw the overall piece reflected my true beliefs, experiences and insights, I started writing on it, just like I am doing now. I read and question every single sentence. I push it out of sight and I rewrite by hand what those words, sometimes too elegantly, sometimes too theatrically, were trying to convey.
The draft is a canvas sketch on which I paint over.
Full transparency on AI: the outline, source dossier, and this draft’s first version were produced by Claude Fable 5 from my dictated raw stories, my published archive, and my curated sources, following my documented voice rules.
Time:
Data scan, research and topic analysis: 1.5 hr.
1st draft and structural revisions: 1 hr.
2nd draft: 0.5 hrs.
Re-writing: 3.5 hrs.
Formatting and links: 1 hr.
Paradata: 0.5 hrs.
Grammar and error check + fixes: 1.5 hr.
Cover ideation and design: 1 hr.
Title and subtitle analysis: 0.5 hrs.
Total apx. time: 11 hrs.
Tools:
Claude (Cowork) Fable 5 - for research plus drafting.
ChatGPT 5.6 Sol for theme / topic brainstorming + cover image design execution
Sublime - for the quoted passages that highlight the 3 elements needed to land good advice.
The Curator - my personal knowledge base on everything I have published
StackContacts - Substack full metrics for articles and notes.
Yutori Scouts - custom news discovery and briefing. Here everything started.
Cover image: Framing, context, visible judgment. Concept & Design by Robin Good, executed by ChatGPT Image 2.
What I Can Do For You
I help experienced independent consultants turn years of expertise and refined know-how into buyer-ready decision resources that make their expertise easier to understand, trust, recommend and buy.
Not a Great Writer?
Check out the curation-based authority system. Building value and trust without requiring you to be a great writer.
Follow a path with a heart.
The time is now.
From Koh Samui (TH)
Robin Good
This is what I've been trying to explain to people. You put it so elegantly. Now I'm just going to have to share this link with people. I was telling people: that it's OK if you use AI to help you. It's just not magically coming up with things on its own. You have a thought and you have the AI flesh out the thought, help you with research, help you with editing, and whatnot. Everything comes from you and through you and I don't see a problem with that. This is a great piece
Dear Robin,
Your story about writing with AI is touching.
Reading it was like sitting beside you and following your thought process.
You perfectly captured my feelings about AI and all the slop and noise we have to sift through.
Thank you for your honesty, as well as for sharing your sensemaking process and experiences.
All the best
Stefan