When AI can produce any output on demand, the bottleneck, the rare, scarce thing becomes the human who can choose well.
The person who knows what to pick, what to leave out, what to connect, and what to apply to a specific context is the one who holds the value. Especially now, inside this ocean of AI-generated possibilities, where everything looks polished, plausible and publishable at first sight.
But the application of this emerging value, looks and is named differently, depending on who you are, what you do and where you are headed.
Sangeet Paul Choudary calls it “idea architecture.” Craig Perry calls it “wisdom.” Ruben Hassid calls it “taste.”
From my viewpoint, they’re all describing the same thing from different angles.
What they are saying is: we are going from production to selection, from generating output to evaluating what is relevant, from knowing to selectively choosing what’s worth applying.
The Idea Architecture
Sangeet Paul Choudary writes that most people using AI are still thinking in terms of outputs. They take one piece of content and repurpose it into a newsletter, a post, a short video, a thread, a carousel. Same stuff, different containers.
The new perspective he suggests considering is to stop treating your newsletter or book or course as the unit, and start treating the ideas inside them as modular components that can be recombined for different contexts and different readers into unique formats.
The human who owns the architecture, who knows what connects to what, and why, and for whom, becomes the one who can’t be easily replaced.
But there is one condition.
Choudary has 100,000 documented ideas built over 15 years in Workflowy. The unbundling / rebundling strategy he proposes clearly works here because there’s actual substance behind it. But for an expert who hasn’t spent years building a second brain or an explicit knowledge base, “idea architecture” can become a fancy name in which there isn’t much that can be re-arranged. In other words: There is not much to recombine if there is not much there.
So the first practical lesson is simple: if you want to compete in this new environment, you need to start building a visible reserve of thinking. Notes, articles, essays, research trails, saved examples, annotated links, personal knowledge bases. You don’t have to do anything difficult, but you definitely need to start doing it. You will learn how to make it better as you do it.
On the other hand, Choudary’s rebundling logic has a very practical and positive consequence for those who, in one way or another, have already started collecting and setting aside relevant stuff:
If you have a solid knowledge base, you do not need to keep producing brand new ideas every week just to keep publishing. You can find new ways to sequence, frame and present what you already know.
The same underlying understanding can become a newsletter, a mini-course, a consulting framework, a reference guide, a checklist, a map, a private briefing, a workshop.
The new effort goes into the architecture. Into the order. Into the connections. Into finding the perfect fit between the idea and the person who needs it. Not into generating fresh raw material every week.
Go deeper: Sangeet Paul Choudary - “How to Reimagine Your Work In The Age of AI”
Not Knowledge, But Experience-Backed Insight
I couldn’t agree more with Craig Perry’s claim that since information is free now, accumulating it is useless as a competitive position. Information is everywhere. Facts are cheap. Summaries are instant. Explanations can be generated on demand.
What AI cannot easily replicate is wisdom: the kind that comes from acting, failing, adjusting, trying again, and paying the small price of experience over many years.
What he’s saying is that today the reader doesn’t come to your newsletter to get new facts. SHe comes for your perspective, your viewpoint, the angle you have developed through your specific experience, made of mistakes, sudden discoveries and small wins.
“People don’t read your writing for the information you give them. They read your work because of your unique knowledge.“
That unique knowledge is judgment made visible on the page.
But visible judgment requires courage. You can select well and still hide the real value of what you are doing. A curated newsletter that doesn’t show why it chose this source and not that one, what it disagreed with, what it threw out, it’s just someone’s picks list. Maybe a useful one, but still a list.
The real opportunity opens when the selection process itself becomes part of the content. Why this source? Why now? Why this angle? What did you reject? What did you disagree with? What did you almost include, but finally left out? What changed your mind? What pattern did you see that was not visible at first sight?
That is where the reader starts seeing you think.
And that is where trust begins to form.
Go deeper: Craig Perry - “The Most Important High-Value Skill To Learn”
Taste Is The Moat (if you have it)
In his article “AI Is a Slot Machine,” Ruben Hassid frames AI as a slot machine: probabilistic, generating one of millions of possible outputs each time you pull the lever. Sometimes decent. Sometimes boring. Sometimes surprisingly good.
His conclusion is that the skilled AI user has learned to stop looking for the magic prompt that delivers the right answer in one shot. Instead, he uses what I have highlighted in the past as the “wise winnowing method”: generate 100 versions, sometimes 1,000, and then select the best one.
His ability to choose, his judgment, is again the winning new card.
As Ruben writes: “AI doesn’t have taste. You do. That’s the moat.”
One of his commenters nails it even deeper: “Our key skill becomes judgment: not generating but selecting.”
I agree with the direction. But I see a weak point.
This advice assumes that everyone has a strong developed ability to evaluate and assess critically what is better, what works, what fits and what doesn’t. But - at least in my experience - that’s not really the case. Most people do not have this refined ability to critically evaluate.
Having worked for a company for 20 years does not automatically give you refined judgment. A job title does not do it either. Sometimes the work you did never required close inspection, comparison, evaluation and continuous improvement. Sometimes you did the job well enough, but without the kind of attention that slowly sharpens your eye. Sometimes you never liked the work enough to care that much.
Time spent in a role does not automatically develop into a strong ability to evaluate and pick by itself.
So if your accumulated knowledge is thin, and your ability to judge is still undeveloped, even the wise winnowing method will leave you selecting mostly by personal preference. You will pick what you like. What feels good. What sounds polished. What gives you a little hit.
But you may not really know why.
To Ruben’s credit, he also makes a point I fully agree with: real taste requires action and risk. You cannot develop good judgment only by reading, watching, prompting or collecting. At some point you need to choose, publish, test, be wrong, adjust, and try again.
Because, and this is the really good news, taste can be trained.
There are specific practices that, if done consistently, can sharpen your ability to pick and select with criteria.
Here’s what I understood taste is, and how to actively cultivate it:
Go deeper: Ruben Hassid - “AI Is a Slot Machine”
Why This Matters for Experts
There is a deep shift taking place, and many experts may miss it because they are still looking at AI as a production tool.
The key shift is that what we have believed until now to be the key asset in determining the value and credibility of any expert - what he knew - is now giving way to something quite different: How he chooses, how he processes, how he thinks and reasons.
Now that knowledge can be retrieved, summarized, compared, remixed and rewritten in seconds, the value is moving somewhere else.
The key asset is becoming how he chooses. How he processes. How he connects. How he evaluates. How he decides what matters in a specific situation.
Knowledge has become the raw material.
Judgment - that is - the ability to pick and choose not based on aesthetics or superficial sensations, but according to a clear purpose, a specific audience, a real need, a context, a constraint, a desired outcome - is the skill that makes it become valuable again.
That is why curation matters so much now.
Curation Makes Your Judgment “Visible”
Why?
Because curation, done well, places you in a different position from the person who only creates individual pieces of content.
Think of art curators, book reviewers, film critics, DJs, librarians.
Their value is not that they add more information. Their value is that they help you navigate what already exists. They show you what deserves attention, what connects to what, what can be safely ignored, and why something matters now.
It’s the same difference that runs between an artist creating one powerful work and a curator designing an exhibition that helps you understand why certain works matter, how they connect, what they reveal, and why others were left out.
It’s the gap that exists between designing one beautiful house and planning a whole neighborhood so that people can move through it, understand it, and actually live better inside it.
To do either one, you need to know your field wide and deep. You need to have looked around. You need to have compared. You need to have changed your mind more than a few times. You need to know where the old roads are, where the dead ends are, and where someone has quietly opened a new highway.
It is through that very experiential knowledge that you can provide such a refined and valuable interpretation lens to others.
These people do not add more information, they help you navigate it, understand it, frame it in a way that allows you to see how it is relevant to you. That’s why they are so uniquely valuable.
When you show how you decided, what you read, what you kept, what you discarded and why, your process becomes a demonstration of your expertise.
Curation = Visible choice and selection at work. Judgement made visible.
Making judgment visible is the product.
How To Apply It In Your Field of Expertise
Here are 4 practical ways to train and show your ability to pick and choose with criteria in your own work.
.
1) Use Paradata
Add a section to every article you write, where you transparently show everything that is around the content you have just finished writing:
Why you wrote it. What was your goal.
What was your thinking process as you wrote it.
Who you wrote it for and why.
What steps did you take. What was your actual writing process.
What sources you used.
What you rejected.
What tools did you use and why.
How much time did each step take.
This does not need to become a bureaucratic appendix. It can be light, personal, useful. The point is to let readers see the thinking behind the final piece.
Paradata makes the backstage visible.
.
2) Make The Process Your Content
Start writing about your research, exploration and discovery process by mixing “learning in public” with “exploring in public”. Leverage the work you do that is normally invisible to story-tell your research and writing journey.
Many experts hide the most valuable part of their work: the search, the comparison, the doubt, the testing, the evaluation, the moment where one idea finally connects with another.
That invisible work is often more useful to the reader than the polished conclusion.
This is where your expertise becomes visible without you having to say “I am an expert.”
.
3) Show The Rough Parts (Scars)
Make your errors, mistakes, rejections and changes of heart visible. Show the friction. Show the parts that resisted. Show the idea you almost used and why you dropped it.
Do not hide the weak points you discovered when you had to make a decision. Do not polish away every sign of struggle. A perfect final answer often teaches less than a visible trail of decisions.
Readers trust people who show how they got there.
.
4) Connect To The Rest
Show how your ideas and your process connect to other ideas, people, tools, trends and examples.
Highlight where they harmonize and where they diverge.
Surface who came before you, and what their perspective was.
Who influenced this? What sparked it? What older idea does it update? What current trend does it push against? What other voices are part of the same conversation?
This is one of the most powerful things a curator can do: reveal connections that are not obvious from the surface.
Paradata
Intent / Why: This article came from my desire to exercise my ability to pick and choose among the many interesting articles being published by authors I like and respect, while challenging myself to surface valuable emerging trends that are not immediately visible.
This is also an expansion of the “Duck Constellation” curation format I have been testing: take different pieces that seem unrelated, look beneath the surface, and identify the hidden thread connecting them in a way that is useful for your readers.
The real question I am asking myself is this: if I truly care about my readers, the expert, the consultant, the serious scholar, the researcher, the journalist, the independent professional trying to build trust online, do I serve them better by writing about trust and curation, or by performing curation in front of them?
Maybe the curation analysis, the principles, the new emerging trends should become part of a premium, separate layer, while my general content should focus on systematically demonstrating in public what good curation is.
If you have read up to this point, I would truly value your opinion in the comments.
Process: For this article I had different ideas in mind.
One was to write about Bruce Lee as a powerful historical curator: not only the martial artist, but the researcher, explorer, philosopher and writer who studied different combat traditions, dance, fencing, boxing, psychology and personal expression to develop his own way.
I still think there is something strong there.
But after researching it, finding a possible working title and writing rough notes, I kept asking myself whether it was relevant enough for this specific issue.
I was not fully convinced. Plus I was too emotionally involved, as a long-time combat scholar and JKD practitioner I lose my objectivity when it comes to things I deeply love and have been involved with for extended periods of time.
I saved the notes and the outline and thought what other alternatives I had.
The first thing that came to mind - in the spirit of modeling more the cure I recommend (curation) instead of theorizing it - I pondered whether a new Deep Reads edition could have been a less risky and more relevant alternative (Deep Reads is a curation format I have recently introduced where one key idea is seen from multiple diverse perspectives. See one example here.
Given it was a Sunday, and I had more time than usual available to me, I went on a Substack reading spree and before I realized I had already found three articles - all published on that very day (!) - that felt very relevant to my readers.
Choudary was talking about idea architecture. Perry was talking about wisdom. Hassid was talking about taste. I felt they were all highly relevant topics, but wanted to find a common thread, that weaved through all of them, while connecting to one of my key themes.
So I convened both Claude and ChatGPT and asked them to help me see what I felt, but could not pinpoint.
In less than a minute I had their answers.
ChatGPT said it clearly: Robin, look, these look like different things. But they are actually three doors to the same room: AI makes producing easier, and as a natural, logical consequence, it makes choosing way more important.
Time: Reading and analyzing relevant articles: 1.5 hours. Deciding what topic to cover as I basically hand-wrote two rough drafts before committing to one: 3 hours. More writing, refining, and adding Paradata: 4 hours. Cover design: 1 hour. (explain what I do, why it takes so much time). Syntax and grammar review + revisions: 1 hour. Total apx time: 7.5 hours.
This is why curation takes time. The visible article is only the final layer. Underneath there is reading, saving, comparing, rejecting, connecting, testing titles, changing direction, checking whether the idea is strong enough, and then making the whole thing readable.
Tools:
Claude AI for finding patterns and underlying threads across different articles and authors.
ChatGPT for brainstorming, text revisions, cover design, syntax and grammar checking.
WUWT for rapid in-depth analysis of the articles I curated.
Cover image: Design by Robin Good executed by ChatGPT Imagen 2
What I Can Do For You
I help experts and consultants who are building a new online career gain the authority, credibility, and visibility they initially lack.
I do this by teaching them how to become top trusted curators in their area of expertise.
For this purpose I’ve created a focused 55-minute video workshop that walks you through:
Why curated content formats are so useful for building credibility and authority
When to use them and what requirements they have
An updated list of real-world examples of curated formats at work
The specific tools you need to curate
The actual key steps that transform researching and writing into curation
The 11 typical mistakes novice curators make
One-to-One Audit & Strategic Advice
For experts and consultants looking for ways to:
a) Create Value and Build Trust for Non-Writers (Curation)
I help you identify and master your own personal ways to gain authority and value by curating insights, research, news, resources and tools in your field of interest.
b) Improve Credibility and Trust
I review your content, positioning and goals to identify the best editorial strategies to build trust and credibility around your focus.
c) Positioning and Personal Branding
I analyze your market positioning to identify key strengths and weaknesses. I help you redefine a strategy to differentiate yourself from the competition while increasing the practical value you bring to your readers.
Available in two tiers: a) for those just starting and b) for those who have been publishing for more than a year, but are not seeing results.
Follow a path with a heart.
The time is now.
From Koh Samui (TH)
Robin Good
Thank you for writing this. Most clear breakdown on what is judgement, how it is defined by personal preference, which in turn is a function of depth of accumulated knowledge. No matter the sophistication of the the LLM, human factors(judgement or taste) still play a key role.