This week is the Translator issue. And I want to start with something I wrote at the end of the last issue, because it is the organizing question for everything that follows:

AI looks like it does what Translators do. It does not. But the confusion is worth taking seriously.

Here is the clearer version of that: AI can make things clear. It cannot make things land. And landing requires a kind of judgment that no model has, because it is not about words. It is about knowing your audience at a level that goes beneath language, into how they receive information, what they trust, what they resist, and what needs to be said first before anything else can be heard.

That judgment, for many women in this community, is something you have been building your entire life. You just may not have named it as a professional advantage yet.

If you have not yet taken the role quiz to find out whether you are a Curator, Synthesizer, or Translator, now is a good moment.

Take the Role Quiz

What it means to be a Translator

Translators make complex ideas accessible and move people to action.

The second part is the one worth sitting with. Making something understandable is a craft. Moving someone to act on it is a different skill entirely. It requires knowing what this specific person needs to hear, in what order, with what framing, before they will be ready to receive what you are actually trying to say.

AI is genuinely good at the first part. Given a complex idea, it can produce a clear, structured explanation at a level of polish that would have taken significant effort to produce even a few years ago. That capability is real and worth using.

What it cannot do is the second part. It does not know your audience. It does not know what they already believe, what they are afraid of, what kind of logic they find credible, or what they need to feel before they will trust anything you tell them. It produces communication that is clear. It has no way of knowing whether it will land.

What is happening at the frontier

AI tools are getting better at producing content that sounds tailored. You can tell a model to write for a specific audience, adjust the tone, simplify the language, make it more formal or more conversational. The output is often impressive.

But there is a difference between adjusting surface features and genuinely understanding an audience. AI adjusts surface features very well. It has no access to the deeper layer.

Erin Meyer, in her book The Culture Map, describes an American engineer named Kara Williams who prepared a thorough, well-researched set of recommendations for a German automotive company. She had visited plants, interviewed experts, and developed practical conclusions she was confident in. She flew to Munich and presented them clearly and directly, leading with her recommendations and supporting them with data. The presentation landed badly.

The problem was not the content. It was the structure. Her German audience operated on what Meyer calls principles-first reasoning: they needed to understand the theoretical foundation, the why, before they could evaluate any recommendation. Williams had led with the what. To her audience, the argument felt groundless, not because the conclusions were wrong, but because the reasoning that would have made those conclusions credible had been skipped.

AI would have produced exactly what Williams produced. It defaults to a clear, direct, conclusion-forward structure because that is the dominant pattern in most of the communication it was trained on. It has no instinct for the fact that the same content, restructured for a different cultural logic, would have landed completely differently.

That instinct is yours. And if you have spent time navigating between cultural contexts, in different countries, different organizations, different rooms where the rules were not written down anywhere, you have been building it for years.

What AI does not understand about cultural logic

In the Curator issue, I wrote about how AI processes text and why attention is not evenly distributed across everything you give it. In the Synthesizer issue, I wrote about how AI produces patterns that look like insight but are missing the contextual judgment that makes insight real.

For Translators, the equivalent concept is this: AI defaults to a single cultural register when it communicates.

That register is not neutral. It reflects the dominant patterns in the data it was trained on, which skews heavily toward a particular style: direct, linear, conclusion-first, individually oriented, and optimized for an assumed reader who is already broadly familiar with the cultural context the message comes from.

Erin Meyer maps eight dimensions along which cultures differ: how explicitly people communicate, how directly they give feedback, whether they persuade through principles or practical examples, how they relate to hierarchy, how they build trust, how they handle disagreement, and how they think about time and scheduling. These are not personality differences. They are deeply embedded cultural logics that shape what people expect from communication, what feels credible to them, and what makes them ready to act.

AI knows none of this about your specific audience. It can produce something that is clear in one cultural logic and completely miss in another. What feels direct and trustworthy to one reader feels abrupt and ungrounded to another. What counts as a compelling argument, what needs to be established relationally before anything substantive can be said, what should be left unsaid: none of this is consistent across cultures, and AI has no instinct for any of it.

This does not mean AI is useless for Translators. It means you have to stay upstream of it in a specific way.

Use AI to produce the first draft. Then ask yourself not whether it is clear, but whether it will land, for this audience, in this context, given everything you know about how they receive information. That question is yours. The model cannot answer it for you.

Subscribe now

How the tools fit into Translator work

Claude (Anthropic) Claude is useful for drafting and restructuring communication. It follows instruction well, which means if you tell it to reorder an argument, lead with a different element, or adjust the weight given to different parts of a message, it does this reliably. The value for Translators is in using Claude as a drafting partner that executes your structural judgment, rather than a tool that makes the judgment for you.

ChatGPT (OpenAI) Custom GPTs are worth knowing about for Translators who communicate repeatedly with the same type of audience. You can build one that holds specific audience context and communication preferences, so you are not re-explaining the same things each session. This is useful when you have already done the cultural judgment work and want to operationalize it.

Perplexity Less directly relevant for translation work, but useful when you need to verify that your framing of a topic is current and accurate before you adapt it for a specific audience.

Gemini (Google) Gemini’s integration with Google Workspace makes it practical for Translators who are working directly inside documents, slides, or emails that need to be adapted for different audiences. The workflow is smoother when the drafting and the adaptation happen in the same environment.

From Izumi’s workflow

I am a Synthesizer by dominant role, and cultural instinct is genuinely part of how I read situations and audiences. That part I trust.

What I am less naturally good at is simplifying my communication. I tend to go deep, add context, qualify my thinking. Left to my own devices, I write for people who are already close to the subject. That is not always who I am writing for.

This is actually where AI is most useful to me in translation work. Not for the cultural judgment, which I do myself, but for the simplification pass afterwards. I will write a draft that says exactly what I mean, with all the layers I think matter, and then ask the model to make it accessible without losing the substance. That gap between what I produce and what lands for a broader audience is where AI genuinely helps me.

The cultural instinct piece, knowing how to structure a message for a specific audience, what to lead with, what to leave out, what tone will build trust rather than distance, that is still mine. AI does not have access to it. What it can do is take the output of that judgment and make it cleaner and more readable than I would on my own.

If you are a Translator who has the cultural instinct but struggles with simplification, that is a real and honest use of the tool. If you are someone who simplifies easily but is less confident about the cultural reading, the tool will not fill that gap for you. Knowing which problem you actually have changes how useful AI is in this part of your work.

Solo vs. team: the constraints are different

If you are a solo practitioner, your translation work is often the thing clients cannot articulate that they need until they see it done well. They know the message is not landing. They do not always know why. Your ability to diagnose that, and restructure accordingly, is what makes the work valuable. AI can help you produce options quickly. The diagnostic judgment is yours.

If you work within a team or organization, Translators are often the people who end up adapting everyone else’s communication without the role being formally recognized. AI can help make that contribution more visible: by producing a baseline draft quickly, you spend your time on the cultural adaptation rather than starting from scratch. The work becomes more legible when the drafting is separated from the judgment.

There is also a specific risk for Translators in organizations that are using AI to produce more communication faster. Volume goes up. The quality of cultural calibration does not go up with it. That gap is where Translator judgment becomes more valuable, not less, even as the tools improve.

Subscribe now

Whatever your current setup, reply and tell us what your experience has been. The question we keep hearing from Translators is this one: how do you hold your cultural instinct when AI is producing something that sounds right but feels off? We will address that in a future issue.

Communication has always required knowing your audience well enough to meet them where they are. AI does not change that requirement. It just makes it easier to produce something that looks like it meets the requirement without actually doing so.

Your cultural instinct is not a background detail. It is the thing the tool cannot replicate. The question is whether you are using AI in a way that puts that instinct to work, or in a way that quietly substitutes for it.

This is the last issue in the back to basics series. Intermediate editions will come later in the year, as the tools and our understanding of them continue to evolve. Thank you for reading all three.

Izumi

This week is the Synthesizer issue. And I want to start with something I wrote at the end of the last issue, because it is the organizing question for everything that follows:

AI looks like it does what Synthesizers do. It does not. But the confusion is worth taking seriously.

That confusion is not a small problem. It is the thing most likely to quietly erode what makes Synthesizer work valuable, not by replacing it, but by making it harder to see where you actually contributed.

If you have not yet taken the role quiz to find out whether you are a Curator, Synthesizer, or Translator, now is a good moment.

Take the Role Quiz

What it means to be a Synthesizer

Synthesizers spot patterns across disparate information and see how things connect strategically.

What that actually looks like in practice: you are the person who reads a market report, remembers a conversation from last week, and recalls something you heard at a conference three months ago, and suddenly sees that all three are pointing at the same underlying shift. You make connections across things that were never designed to be read together. You find structure inside complexity, not by simplifying it, but by recognizing what is already there.

AI can produce something that looks like this. That is the problem, and it is worth understanding why.

What is happening at the frontier

Most AI tools are built to answer fast. You ask a question, they produce a response. The response is fluent, confident, and often very plausible. For many tasks, that is exactly what you need.

But synthesis is not most tasks. Synthesis requires sitting with complexity long enough to find what is actually there, rather than what sounds right on the first pass. Fast and fluent is not the same as correct, and for Synthesizers, a wrong connection delivered confidently can travel a long way before anyone questions it.

This is why a newer category of AI tools is worth knowing about: models that are specifically built to slow down. Instead of jumping to the first plausible answer, they work through a problem in steps, check their own reasoning, and take longer to produce a response. They are not faster or cheaper. But for synthesis work where the stakes of a wrong connection are real, the slower pace is the point.

You will see these described as “reasoning models” or tools with a “thinking mode.” OpenAI’s o-series, and the extended thinking options in Claude and Gemini, are current examples. You do not need to use them for everything. But it is worth knowing they exist and what they are actually for.

The thing AI does not tell you about how it reads

In the Curator issue, I wrote about tokens, the basic unit of how AI processes text, and why a session that grows too long can affect your output in ways that are not always obvious. For Synthesizers, there is a related problem that is worth understanding, and it has nothing to do with technical knowledge.

Here is the simplest way to put it: AI does not read everything you give it the same way you do.

When you read ten documents, you bring judgment to all ten. You decide which ones feel important, which ones contradict each other, which one contains the detail that changes everything. You hold them all in mind at once.

AI does not work this way. When you give it a lot of material in one session, it pays more attention to what comes at the beginning and what comes at the end. What is in the middle gets less weight. This is not a bug that will be fixed soon. It is a feature of how these tools process information, and it still shows up even in the most advanced models available today.

For Synthesizers, this matters because the connecting signal, the piece of information that makes the pattern click into place, is just as likely to be in the middle of what you have shared as anywhere else. If the model is underweighting it, you may get a synthesis that sounds coherent but is missing the most important piece.

Here is what this changes about how you work:

Give it less, not more. The instinct is to paste everything in and let the model find the pattern. The better move is to be selective before you start. What are the three or four inputs that you think are most likely to be connected? Start there.

Put what matters where it will be seen. If you are sharing several things in one session, the order is not neutral. What you most want the model to engage with belongs at the beginning or the end, not buried in the middle.

Keep research and synthesis separate. Gathering information and finding patterns in it are different kinds of thinking. They work better in separate sessions, and often in different tools. Mixing them in one long session tends to produce muddier output.

Here is how this plays out across the tools most of us are already using:

Claude (Anthropic) Claude handles long material well and is suited to the pattern-finding phase, once you have already done the research and want to work through what it means. If your synthesis work returns regularly to the same topics or bodies of knowledge, the Projects feature lets you store that context so you are not starting from scratch each time.

ChatGPT (OpenAI) The o-series models, and the reasoning options within ChatGPT, are worth using when your synthesis task is genuinely complex, when you are trying to connect things across domains that do not usually speak to each other, or when you want to stress-test a conclusion you have already reached. The Deep Research feature is also useful for pulling connections across a larger body of sources than a standard chat session handles well.

Perplexity Less useful for synthesis directly, but valuable once you have formed a hypothesis. It is a fast way to find out whether there is evidence for or against what you think you are seeing, and what angle you might have missed.

Gemini (Google) Gemini can hold a lot of material in a single session. A larger capacity does not solve the attention problem I described above, but it does reduce the risk of material being cut off entirely in very long sessions. I use it less than the others, so I will not overstate this. If it is a regular part of your workflow, I am curious what you are finding.

Subscribe now

From Izumi’s workflow

As I mentioned in the Curator issue, I am a Synthesizer by dominant role. This is the section most directly about how I actually work.

The synthesis phase of my work happens in Claude. Not because I made a deliberate technical decision, but because the open-ended, multi-threaded thinking I need to do, holding several angles in play before committing to a direction, feels more supported there than in other tools I have tried. I use Sonnet for most of this. When reasoning quality genuinely matters, I use extended thinking mode.

I do not try to do research and synthesis in the same session. Research happens in ChatGPT. Pattern work happens in Claude. That boundary is not rigid, but keeping the phases separate produces better output than collapsing them.

The thing I keep noticing: when I get a plausible-sounding synthesis back from the model quickly, that is usually the signal to slow down. Fast and fluent is not the same as right. The model is very good at producing connections that sound like insight. My job is to evaluate whether they actually are.

Solo vs. team: the constraints are different

If you are a solo practitioner, your synthesis work is often your most differentiated asset and also the hardest to make visible. The risk with AI is not that it replaces the synthesis. It is that it produces output that looks similar enough to yours that the distinctiveness of your thinking becomes harder to demonstrate.

The answer is not to avoid AI in synthesis work. It is to document your reasoning, not just your conclusions. When you arrive at a pattern or a framework, ask the model to help you articulate the logic that got you there. That articulation is what makes your thinking legible to others, and what distinguishes it from what a model would produce on its own.

If you work within a team or organization, Synthesizers tend to be the people pulled into too many rooms precisely because they can read across silos. AI is now doing a version of this for everyone, which can make your contribution less visible, not because it is less valuable, but because the output looks similar. The synthesis that remains irreplaceable is the kind that requires context no model has access to: the qualitative undercurrents, the history of how something was tried before, the reading of what is not being said in a meeting. That is where Synthesizers stay essential even as the tools improve.

There is also a trust question. When you share a synthesized view with a team, are you sharing the conclusion or the reasoning that produced it? The reasoning is where the value lives. AI can help you make that reasoning more explicit and communicable, which is worth building into your workflow deliberately.

Subscribe now

Whatever your current setup, reply and tell us what your experience has been. The question we keep hearing from Synthesizers is this one: how do you know when you have actually found a real pattern versus a plausible one the model generated that felt like yours? We will address that in a future issue.

Synthesis has always been the work of finding what is real underneath what is merely visible. AI makes that harder in one specific way: it produces visible patterns faster than you can evaluate them. The foundation of Synthesizer work does not change. But the speed at which you have to exercise judgment does.

Next issue: The Translator × AI. If Synthesizers face the problem of AI mimicking their output, Translators face a different version: AI that can produce clear, accessible communication at scale, which raises the question of what human translation actually adds. The answer is more specific than you might expect.

Izumi

This series is my own back to basics. Not because I am starting over, but because the foundation keeps shifting and I think there is value in revisiting it deliberately, rather than assuming what I understood a few months ago is still accurate. In some cases, what was true last month is already outdated.

That is why I am writing this now. Not because Curators, Synthesizers, and Translators are new to AI. But because the ground underneath how we use it keeps moving, and the people who benefit most from pausing to check their footing are often the ones who thought they already knew where they stood.

This is the first issue in a three-part series on how to use AI through the lens of your AI³ role. Curator this week. Synthesizer and Translator to follow. Each issue covers the basics. Intermediate and advanced editions will come later in the year, as the tools and our understanding of them continue to evolve.

If you have not yet taken the role quiz to find out whether you are a Curator, Synthesizer, or Translator, now is a good moment.

Take the Role Quiz

It takes two minutes and will change how you read this series.

What it means to be a Curator

Curators are great at research, gathering information, and figuring out what is actually credible and relevant.

That last part is the one worth sitting with. AI can find information at a speed and scale no human can match. What it cannot do consistently or reliably is evaluate it. It cannot tell you whether a source has an agenda, whether a statistic is being taken out of context, or whether something important is missing from the picture entirely.

That judgment is yours. And right now, it is one of your biggest competitive advantages.

AI can help with evaluation, but the final call, and the responsibility for it, stays with you.

What is happening at the frontier (and why this edition focuses on basics)

AI research tools including Perplexity, ChatGPT search, Gemini with Google integration, and Claude can now pull live sources rather than relying only on what they were trained on. That is a real shift from even a year ago.

Further out, agentic research workflows are starting to appear. These are systems where AI runs multi-step research on its own, coming up with follow-up questions, pulling from multiple sources, and putting results together, without you directing each step. For Curators, this is worth paying attention to, because when the gathering process is automated, the question of who is actually doing the evaluating gets more complicated.

But this edition is not about that yet. It is about using the AI tools most of us already have, and using them in a way that actually fits how Curators work.

Understanding tokens: why this matters for how you work

Before we get practical, there is one concept worth knowing because it affects your output across every AI tool you use: tokens.

Tokens are roughly how AI models read and process text. Every word you type, every source you paste, every instruction you give, it all uses tokens. This is not just a Claude thing. Every LLM, including ChatGPT, Perplexity, and Gemini, works the same way underneath. What is different is how each tool shows that to you, and how much each pricing plan actually gives you to work with.

Here is a simple breakdown:

Claude (Anthropic)
Tokens are most noticeable here. Every model has a context window, which is basically how much it can hold in its working memory at one time. On the free plan, you hit limits faster. On paid plans, you have more room. Sonnet is the practical everyday choice. Opus is more capable but more expensive per token, so it makes more sense for tasks that genuinely need that extra depth.

ChatGPT (OpenAI)
Tokens are largely hidden in the consumer interface, though they still exist under the hood. Free users hit message limits rather than seeing token limits directly. Paid plans, especially ChatGPT Plus and the deep research feature, give you significantly more to work with. Deep research feels less constrained because it runs extended, multi-step processes with higher internal limits, which makes it well suited for longer research sessions.

Perplexity
Also token-based underneath, but it feels lighter to use because queries are shorter and more focused rather than long and iterative. Free users get a limited number of Pro searches per day. Paid users get more. This is part of why it works well as a verification tool rather than a primary research environment.

Gemini (Google)
Same underlying structure. Free tier has limits. Gemini Advanced gives more capacity and longer context windows, with the added benefit of deep Google ecosystem integration for research.

The practical reality

Tokens are not a Claude problem. They are just how all LLMs work. What differs is how visible that is to you as a user. Free plans work fine for occasional use. If AI is a regular part of your workflow, paid plans are worth it, because running out of context mid-session is not just annoying. It actually affects the quality of your output in ways that are not always obvious.

For Curators specifically, this matters most when you are pasting multiple long sources into one session. When the context window gets full, the model does not treat everything equally. Information that is earlier or less central in a long context can get less attention than what comes later or sits closer to your actual question. The sources you carefully selected may not all be shaping the output the way you expect. Shorter, more focused sessions tend to produce better results than one big paste of everything you have.

From Izumi’s workflow
I am a Synthesizer by nature, but my work involves a lot of analysis and research is part of that. I did not build my workflow from a technical understanding of tokens. I built it by noticing what each tool did well and what frustrated me.

I use ChatGPT for research and building context. It handles volume well and I can go deep without hitting the same constraints. I use Claude (Sonnet) for brainstorming and ideation, which is where my core Synthesizer work happens. Sonnet is fast and capable enough for open-ended thinking without the cost of Opus. And I use Perplexity as a fact-checker. I think of it as having a second expert review my work before anything goes out. It often opens angles I had not considered and makes the final output more solid.

Of course I am not a researcher by nature. I use this process for context creation when I need it, like putting together this newsletter. If that sounds like you, building a workflow that fits your actual habits is more useful than trying to follow an ideal version of how you think you should be using AI.

Subscribe now

Projects vs. chat: a distinction worth making on purpose

Most people open a new chat every time they use an LLM. That is fine for one-off questions. But if you are a Curator who keeps coming back to the same topics, sources, or standards, starting from scratch every session costs you more than just time. It costs you context.

Here is the distinction that changed how I work.

Claude Projects
Claude Projects let you store guidelines, instructions, and reference material in a persistent project space. Claude holds that context across every session inside the project. You are not re-explaining who you are or what you are working on every time you start a new chat.

For Curators, the practical move here is to store your datasets, source lists, or editorial criteria in the Project knowledge base instead of pasting them into every session. Claude already has them before you begin. Your context window stays cleaner for the actual work, which means better output and less token overhead before you have even started.

Think about what a Curator’s Project knowledge base might include: a recurring set of trusted sources, criteria for what counts as credible in your specific field, a glossary of relevant terms, or research summaries you want Claude to build on rather than repeat. The project becomes a working environment, not a blank page.

ChatGPT Projects and custom GPTs
In my own workflow, ChatGPT Projects have naturally become my go-to for deeper research and more technical tasks. That was not a deliberate decision. It is just what each tool turned out to be better at. Custom GPTs add another layer, letting you build a repeatable tool set up for a specific task, which is useful when you find yourself doing the same type of research regularly.

For Curators, a custom GPT built around your research criteria and credibility standards could save real setup time across repeated tasks.

Gemini and Perplexity
Gemini has a feature called Gems which works roughly like Claude Projects and custom GPTs. Perplexity has its own ways of managing ongoing research. I do not use either enough to speak to them honestly, so I will not pretend otherwise. What I am genuinely curious about is how people in this community are using these tools. If Gemini or Perplexity is a regular part of your workflow, tell us in the comments below. That conversation is more useful than anything I could write from the outside.

NotebookLM
This one keeps coming up in conversations I am having. NotebookLM from Google is built specifically for working across multiple sources, which makes it potentially very relevant for Curators. I have not used it enough to have a real opinion yet. If you have, I want to know what you think.

The bigger point across all of these tools is the same. Projects and persistent environments are not just a nice feature. For Curators who work with recurring topics, ongoing research, or consistent standards, they represent a fundamentally different way of working. The question is not which tool has the best project feature. It is whether you are using any of them on purpose, or still opening a new chat every time out of habit.

Subscribe now

Solo vs. team: the constraints are different

How you use AI as a Curator depends a lot on your work situation.

If you are a solo entrepreneur or freelancer, you have more freedom to paste source material directly into Claude and move quickly. That freedom is real, but it comes with personal responsibility around what you are sharing. Just because you can paste something does not mean you should, especially if it contains client information or unpublished data.

If you work in a corporate environment or with clients who have LLM restrictions, the rules are more formal. Many organizations limit what data can go into a public LLM at all, which changes what AI can actually help you with in the research phase. In those cases, AI is most useful for structuring your thinking and shaping your analysis rather than processing the raw material directly.

There is also a sharing question that applies to both situations. When you share curated output with a team or a client, are you sharing the result or the reasoning behind it? The reasoning is usually what makes a Curator’s work valuable. AI can actually help you document and explain that reasoning more clearly, which is worth building into your process deliberately rather than leaving as an afterthought.

Whatever stage you are at, reply and tell us what your experience has been so far. We will share what we hear in a future issue.

Being a Curator has always meant knowing that the value is not in how much information you collect. It is in what you decide matters and why. AI does not change that. It just raises the stakes, because everyone now has access to the same retrieval capability you do.

Your judgment is still the differentiator. The question is whether you are using AI in a way that sharpens it or quietly replaces it.

Next issue: The Synthesizer × AI. If Curators face the risk of outsourcing their judgment, Synthesizers face a different problem entirely. AI looks like it does what Synthesizers do. It does not. But the confusion is worth taking seriously.

Izumi

A small, carefully curated room. Accomplished people. Many of them LinkedIn power users who knew the platform inside out.

And here’s what the feedback said: some of them left feeling the content was too conceptual. They wanted more advanced tactics. More actionable steps. They came for a blueprint and felt like they got a framework.

We’ve been sitting with that tension since.

Because we think it’s the most important signal we’ve heard in a while.

What “actionable” actually reveals

When a smart, experienced audience says something was too conceptual, the instinct is to go back and add more tactics. More steps. More tools.

But look at what the same audience said was the most valuable part of the evening: the networking. The intimate conversations. The moment when people turned to the person next to them and tried to answer a deceptively simple question — are you more of a Curator, a Synthesizer, or a Translator?

That’s not a tactics problem. That’s a different kind of value entirely.

The people who wanted more advanced LinkedIn tips were looking for an edge within the current game. The framework Irene presented was asking a different question: what game are you actually playing?

Those are not the same ask. And the friction between them is worth paying attention to.

Subscribe now

The Three-Role Lens

Here is what we know about why Curators, Synthesizers, and Translators tend to experience the same event differently.

If you’re a Curator, you came to the event with very specific gaps you wanted filled. You already know your landscape well — you follow the right people, you read the right things, you track what’s changing. When something doesn’t add to your signal, it feels like noise. The feedback that the content was “not actionable” often comes from Curators. They needed precision. They got principles.

If you’re a Synthesizer, you were probably the one most engaged during the interactive discussion. Not because you knew the answer, but because you were connecting it to three other things simultaneously. The three-role framework lands differently for Synthesizers — it gives them a new lens to pattern-match across everything they’ve already been observing.

If you’re a Translator, you left thinking about the people you work with. Who in your team is a Curator? Who tends to synthesize without knowing it? The framework became immediately relational for you — a language for something you’ve been doing instinctively.

Same room. Same content. Three completely different experiences of value.

Four skills that showed up in the room that night

Role clarity

What it is: Knowing which kind of value you most naturally create — and being able to name it.

In practice: When people are asked to identify their dominant role, the first response is almost always “I use all three.” That’s true. But pushed further, most people can find one that feels most like home — the mode they return to when the pressure is on.

Reflection question: When you’re at your best in a room full of smart people, what are you most often doing — filtering for what matters, connecting patterns, or making complexity land for others?

Tolerance for conceptual discomfort

What it is: The ability to sit with a framework before demanding that it tell you what to do.

In practice: The attendees who found the event most valuable were the ones willing to hold the Curator/Synthesizer/Translator lens loosely — to let it reorganize how they saw themselves before asking how to apply it. The ones who found it frustrating wanted the application step to arrive immediately.

Reflection question: When you encounter a new framework, is your first instinct to test it against your experience, or to ask what you should do with it?

Reading the gap between title and content

What it is: Recognizing when an event, a role, or an opportunity is offering something different from what it advertised — and deciding whether that difference is a problem or a gift.

In practice: The event was called “AI Can’t Replace Influence.” Several attendees wanted that exact promise delivered: advanced LinkedIn tactics, specific AI tools for leadership. What Irene offered was upstream of that — the identity-level question that makes the tactics make sense. Neither is wrong. They’re operating at different levels.

Reflection question: Think of a time you got something different from what you expected. In hindsight, was the gap a mismatch or a redirect?

Articulating value that resists measurement

What it is: The ability to make a case for something that doesn’t produce an immediate, visible output.

In practice: The hardest thing to sell in any professional context is clarity. Not a deliverable, not a result, not a metric — the kind of thinking that reframes everything downstream. This is exactly what the three-role framework offers. And it’s exactly why it’s a harder sell than a tactics list.

Reflection question: Is there something you know is genuinely valuable that you struggle to articulate because it doesn’t have a clear output? What would it take to put it into words?

Subscribe now

The AI³ Lens

Here is what we keep coming back to.

AI is extraordinarily good at tactics. It can generate LinkedIn content strategies, summarize influencer frameworks, draft post templates, and A/B test messaging faster than any human. If what you need is a more advanced set of tactics, AI can probably help you right now.

What AI cannot do is tell you which game is worth playing.

That’s the role clarity question. And it’s the one that requires a human — specifically, a human who has navigated enough different contexts to know that the same move lands completely differently depending on who’s making it and why.

The women in that room on International Women’s Day who are going to use AI most effectively aren’t the ones who left with the best tips. They’re the ones who left with a clearer sense of what they’re actually optimizing for.

That’s not a small distinction. It’s the whole thing.

Not sure which role describes you?

Take the AI³ role quiz

What’s one moment this week where your natural role showed up without you trying?

Izumi

P.S. The feedback from this event has stayed with me longer than I expected. The split in the room was clear — some people wanted to understand what they are, some people wanted to know what to do next. Both are valid. It’s made us think harder about how we bridge the two: the identity-level clarity the framework offers, and the concrete next steps that make it land. That’s the work we’re taking into the next phase of AI³.

In a survey of U.S. executives, 52% said the ability to lead humans and AI together will define successful business leadership in 2026. Not managing a tech stack. Not deploying the right tools. Leading humans and AI, together.

That’s a different ask than what most people are preparing for.

Most of the conversation around AI and careers is still about tasks: what AI will take over, what skills to learn to stay relevant, how to “keep up.” It’s a defensive posture. And it’s not wrong, but it’s incomplete.

Because what executives are actually looking for isn’t someone who survived the AI transition. It’s someone who can guide others through it. Someone who can hold the human side of the equation while the technical side shifts underneath everyone’s feet.

That’s not a tech skill. That’s a leadership skill. And it’s one that’s been undervalued for a long time.

What This Actually Requires

The executives in this research weren’t just talking about technical integration. They specifically named transparency, ethical decision-making, and emotional intelligence as the capabilities that matter.

Read that again: transparency, ethical decision-making, emotional intelligence.

These are the soft skills that get cut from job descriptions when companies are in growth mode. They’re the ones that get labeled “nice to have” in performance reviews. They’re the ones that mid-career professionals, especially women who’ve navigated complex organizations across cultures, have been quietly developing for decades.

They’re not nice to have anymore. They’re the job.

The Three-Role Lens

What strikes us about this shift is that it looks different depending on your natural role, and each role has a distinct contribution to make.

If you’re a Curator, you’re already doing the work that matters most right now: deciding what’s worth paying attention to. As AI generates more information faster, the ability to evaluate what’s credible, what’s relevant, and what’s actually worth acting on becomes a leadership capability in itself. Not everyone can do that well. You can.

If you’re a Synthesizer, your advantage is in seeing across the gap. The friction in most organizations right now lives at the intersection of human decision-making and AI capability, where the two don’t yet speak the same language. Synthesizers can see what’s actually happening versus what people think is happening, and help leadership navigate accordingly. That cross-domain clarity is hard to find and harder to replace.

If you’re a Translator, your advantage is in making the shift legible to the people inside it. Leading humans and AI together requires someone who can take what’s technically possible and communicate it in ways that build trust rather than anxiety. That’s not just a communication skill. It’s a strategic one.

Each role has a distinct contribution to make here. Knowing yours is where strategy starts.

Subscribe now

The Soft Skills That Actually Matter Now

The research names emotional intelligence as a defining leadership capability for 2026. But emotional intelligence is a broad term. Here’s what it actually looks like in the context of leading humans and AI together:

1. Trust Calibration

What it is: The ability to help others develop appropriate trust in AI outputs. Not blind acceptance, and not reflexive skepticism either.

In practice: You’re in a team meeting. Someone dismisses an AI-generated analysis without reading it. Someone else accepts it without questioning it. You’re the person who slows the room down, asks the right questions, and helps the team figure out what to do with it. That’s trust calibration, and it’s a skill very few people have named, let alone developed.

The question: When you encounter AI-generated work, what’s your default response, and is it serving you strategically?

2. Holding Complexity Without Collapsing It

What it is: The ability to sit with ambiguity long enough to make a good decision, rather than forcing premature clarity.

In practice: AI makes it tempting to get to an answer faster. But faster isn’t always better, especially when the question involves people, culture, or strategy. The leaders who will matter in 2026 are the ones who can hold complexity, weigh competing signals, and resist the pressure to simplify before it’s time.

The question: Where in your work are you being pushed toward faster answers than the situation actually warrants?

3. Transparency as Practice

What it is: Communicating honestly about what you know, what you don’t know, and how decisions are being made, especially when AI is involved.

In practice: One of the biggest trust gaps opening up in organizations right now is around AI-assisted decisions. Who made this call? Was it the algorithm or the human? Transparency as a leadership practice means naming that, not defensively, but matter-of-factly. It’s the difference between eroding trust quietly and building it deliberately.

The question: How transparent are you currently about when and how you’re using AI in your work?

4. Cross-Contextual Emotional Intelligence

What it is: Reading the room accurately across cultural, generational, and professional contexts, and adjusting how you communicate accordingly.

In practice: Leading humans and AI together isn’t a universal experience. A 55-year-old in a traditional industry navigates it differently than a 32-year-old in a startup. Someone in Germany navigates it differently than someone in Singapore. If you’ve lived and worked across contexts, which many of you have, this is a skill you’ve been building without calling it that.

The question: Which context in your professional life requires the most emotional recalibration, and what does that tell you about where your cross-cultural advantage actually lives?

AI³ Lens

Here’s the pattern we keep seeing: the skills being named as “future leadership capabilities” are skills that mid-career professional women, particularly those with cross-cultural backgrounds, have been practicing for years out of necessity.

Navigating organizations that weren’t designed for you builds trust calibration. Living between cultures builds cross-contextual emotional intelligence. Managing complexity without institutional support builds the ability to hold ambiguity. Being the person who bridges worlds builds transparency as a default, because you learned early that opacity costs more than it saves.

The question isn’t whether you have these skills. It’s whether you’ve named them, and whether you’re positioning them as the strategic advantage they actually are.

AI doesn’t lead humans. Humans do. And the humans who will lead well are the ones who already understand that the connective tissue matters more than the speed.

Subscribe now

Not sure whether you’re a Curator, Synthesizer, or Translator, or how your role maps to what’s being asked of leaders right now?

→ Take the AI³ Role Assessment and start from a position of clarity.

Which of these four soft skills feels most like you, and which one do you most need to develop? We’re curious: what are you seeing in your own organization right now?

Izumi

P.S. We’re in the middle of something that doesn’t have a clean name yet. Building a framework that we believe in, with no proof points yet to show the people who might need it. The research keeps confirming that what we’re pointing to is real. The harder work is building the visibility to reach the people who are ready to use it. If you know someone who’s navigating this kind of leadership transition, we’d be grateful if you passed this along.

You ask AI to draft something, read it back, and quietly rewrite chunks of it before it goes anywhere. You don’t think much of it. The output just wasn’t quite right for who it was going to. If that sounds familiar, this issue is about why that matters more than you might think.

What the research actually found

There’s a common assumption that as AI models get more capable and more multilingual, the cultural gaps in their output will naturally close. A 2025 study published at one of the top computational linguistics conferences tested whether that’s true.

The paper’s title gives away the conclusion: “Multilingual ≠ Multicultural.”

The researchers compared outputs from three families of current AI models, including Google’s Gemma series, AI2’s OLMo models, and successive versions of OpenAI’s models, against the World Values Survey, the standard global tool for measuring how values differ across countries. They looked at four languages: Danish, Dutch, English, and Portuguese. What they found was that language capability and cultural alignment have no consistent relationship. A model that performs well in a language is not reliably more aligned with the culture of the people who speak it. In some cases it’s less aligned.

Their conclusion:

getting cultural alignment right requires dedicated effort that goes beyond language capability. It doesn’t happen automatically as models become smarter or more multilingual. It has to be built in deliberately, and most current models haven’t done that work.

What you are probably already doing

Think about the last time AI produced something you edited heavily before it reached an audience. Maybe the tone was too blunt or too soft. Maybe it felt like it was written for the wrong type of professional, or the wrong country, or the wrong relationship dynamic. You adjusted it.

Most people doing this don’t call it anything. They just fix it. And the people who catch it fastest are usually the ones who have spent their careers working across different cultures, countries, or languages, because they already know what the default sounds like from the outside.

This is the kind of knowledge that doesn’t show up on a resume. It’s built over years of reading rooms, switching between contexts, and noticing when something lands and when it doesn’t. What is new is that it’s now directly useful in a very concrete way: working with AI that has a built-in cultural perspective it doesn’t announce.

You’re not an auditor. You’re already ahead.

We want to be clear about something: this is not a pitch to position yourself as an “AI bias detector” or to add a new job title to your LinkedIn. What we’re pointing to is more ordinary than that.

The cross-cultural fluency you’ve built, through moving between languages, working in different countries, or simply operating in professional environments where you were not the default, gives you a real edge when working with tools that have a built-in cultural slant. You see what others miss. You know what to ask for. That’s not a special skill you need to develop. It’s one you’ve already got.

What this means through the AI³ lens

If you’re a Curator, this shows up in quality control. When AI output feels slightly off for your audience, you catch it. That instinct is the product of years of noticing what fits in different contexts. It is not something you can learn from a prompt.

If you’re a Synthesizer, you’ll recognize a bigger pattern here: AI reflects the world that built it, and that world was not built with everyone in mind. This is true of AI, but also of most systems and industries. Knowing how to spot that pattern, and what to do with it, is one of the things that makes cross-cultural professionals valuable in rooms where strategy gets made.

If you’re a Translator, this is probably the least surprising thing you’ve read this week. You already know that the same message lands differently depending on who’s receiving it, and where. When you edit AI output for a specific audience, the work is not cosmetic. It requires real cultural knowledge. The research puts a name on work you’ve been doing all along.

Not sure which role you are?

Take the AI³ Role Quiz to find out

One resource worth reading

“Multilingual ≠ Multicultural: Evaluating Gaps Between Multilingual Capabilities and Cultural Alignment in LLMs” — Rystrøm, Kirk & Hale, COLING 2025

The title is the insight, but the paper earns it. Most AI coverage either overclaims (AI is culturally biased and broken) or underclaims (models are improving, it’ll sort itself out). This paper does something more useful: it tests the specific assumption that language capability and cultural alignment travel together, and shows they don’t. For Curators, that’s the research backing for an instinct you already have. For Synthesizers, it’s a pattern worth tracking across other tools and systems. For Translators, it confirms that the gap you’re bridging when you adjust AI output for a real audience is not going away on its own.

What patterns are you seeing?

When you fix AI output before it reaches your audience, what are you actually correcting for? We’d love to know.

If a colleague should be reading this, forward it their way — or point them to our Substack

Subscribe now

P.S.

This newsletter comes partly from my own desk. I sit between American and Japanese project teams, and almost every AI-generated meeting summary or email translation needs work before it goes anywhere. Not because the AI got the words wrong, but because the judgment calls — when to be direct, when to soften, which metaphors actually travel — aren't in the model. They're in me. That's what this newsletter is really about.

Not competitive edges in the abstract sense, but literal edges: the places where capability drops off sharply, where something that looked like an advantage becomes a liability. In AI research, there is a term for this — the jagged technological frontier — and it explains something we keep noticing in how people actually use AI at work.

The popular story about AI goes one of two ways. Either AI is coming for your job, or AI is going to make you dramatically more productive. Both versions skip the more interesting part: where AI helps you, and where it quietly makes you worse.

A finding worth sitting with

In 2023, Harvard Business School and Boston Consulting Group ran a controlled experiment with 758 consultants to measure what AI does to knowledge work. On tasks within AI’s capability range, consultants using GPT-4 completed 12.2% more tasks, worked 25.1% faster, and produced results rated more than 40% higher quality. Those numbers get shared often.

The part that gets shared less: on a task selected to fall outside the frontier, AI users were 19 percentage points less likely to produce the correct solution than colleagues who worked without AI at all.

The researchers called it a jagged shape because that is what it looks like up close. Not a smooth line where AI gradually becomes less useful, but an uneven terrain where tasks that appear similar can produce sharply different results depending on which side of the boundary they fall on. A task that looks adjacent to one AI handles well might be just far enough outside it to produce confidently wrong output. And if you trust that output without catching it, you end up worse off than if you had worked alone.

This is not a universal law about AI at work. It is the result of one carefully designed field experiment, and the terrain keeps shifting as models improve. But the underlying logic holds: knowing where your work sits relative to what AI can reliably do is now part of the job.

What this means through the AI³ lens

The HBS research identified three modes of AI use across its studies: Cyborgs, who blend AI into nearly everything and move between tool and thought without a hard boundary; Centaurs, who maintain a clear division between what they hand to AI and what they handle themselves; and Self-Automators, who use AI to systematically rebuild how they work rather than just accelerating existing tasks. Each mode produces different skilling outcomes over time.

What this tells us is that “using AI” is not one thing. The mode matters as much as the tool. And the right mode is not the same for everyone — it depends on where your value actually lives.

For Curators: the risk is over-delegating to AI on tasks that require exactly the critical discrimination that makes you valuable. Use AI to expand the volume of what you can evaluate, then apply your full judgment to what actually matters. The value lives in the discernment, not the discovery.

For Synthesizers: AI can surface connections, but adopting those outputs uncritically means pattern-matching against AI’s existing associations rather than generating genuinely novel ones. Use AI to pressure-test connections you have already identified, not to produce the initial synthesis. Your competitive advantage is originality, and that requires the first pass to be yours.

For Translators: making complexity accessible, finding the right frame for a specific audience, adjusting tone without losing substance — these are areas where AI genuinely accelerates the work. The risk is subtler: outputs that are technically clear but that flatten the nuance that only a strong original instinct can preserve. Bring that instinct to every output before it leaves your hands.

Not sure which role you are? Take the AI³ Role Quiz to find out:

Take the AI³ Role Quiz to find out

One curated resource with synthesis

“Navigating the Jagged Technological Frontier” — HBS Working Paper No. 24-013, Dell’Acqua, McFowland, Mollick et al. (2023), with follow-up research published by HBS in 2025.

The original 2023 paper introduced the jagged frontier concept and ran the 758-consultant experiment. The 2025 follow-up expanded the framework to include the three modes of AI use (Cyborgs, Centaurs, Self-Automators) and traced how each affects skill development over time. Together, they make the case that knowing where your work sits relative to the frontier is now a core professional skill — and that how you choose to work with AI shapes not just your output today, but what you are capable of a year from now.

What we’re still sitting with

The jagged frontier is not fixed. It shifts as models improve, as workflows evolve, and as your own fluency deepens. The professionals who navigate it well are the ones who stay curious about the shape — not the ones who assume they have already mapped it.

What patterns are you seeing in where AI actually helps versus where it gets in the way?

If a colleague should be reading this, forward it their way — or point them to our Substack.

Subscribe now

P.S.

I've been doing a lot of outreach lately, reviewing coaches' websites and profiles before reaching out. What I keep noticing is that my synthesis is always better when it comes from my own brain first. The moment I ask AI to "analyze this profile," something in the observation goes generic. But when I do the reading myself, notice what stands out, form a hypothesis, and then use AI to poke holes in it? The message I end up writing actually lands. I think that's what the Synthesizer edge looks like in practice.

Not in a conflict. In a negotiation. When someone is offering you a salary that sounds significant, but what you’ve actually built is worth considerably more.

Alex Cooper knew. And that decision — more than the podcast, more than the Spotify deal, more than the $125 million empire — is the story worth examining.

Who Alex Cooper is

Most people know her as the host of Call Her Daddy, one of Spotify’s top global podcasts. What’s less discussed is how she got there.

After graduating from Boston University, she moved to New York and worked in advertising sales. In 2018, she and her roommate launched a podcast from their apartment. Within months, Barstool Sports acquired Call Her Daddy — a digital media company built almost entirely around male sports culture and humor. She had signed away the intellectual property.

Two years later, in the middle of a public contract dispute that became a cultural moment, she negotiated it back. Then she signed a $60 million exclusive deal with Spotify — at the time, the largest contract ever for a woman-led podcast. In 2024, she signed a multi-year deal with SiriusXM valued at around $125 million, while simultaneously running the Unwell Network, a media company she owns outright.

The pivot that matters here isn’t from advertising sales to podcasting. It’s from creator to owner — in a space where most women never make that move.

The AI³ Role: Translator — with Curator instincts

Alex is a Translator. Her skill has always been making the interior lives of young women — relationships, ambition, fear, desire — feel speakable out loud. She didn’t invent the topics. She found the language for them that her audience had been waiting for.

But what kept her from being just another creator was a Curator layer underneath: she understood early what was worth keeping, what was worth amplifying, and critically, what she should never give away again.

Four soft skills behind the strategy

1. Naming what the audience can’t yet say

The most underrated Translator skill isn’t clarity, it’s timing. Alex didn’t explain things her audience already understood. She said things they were already feeling but hadn’t heard articulated. That gap between felt and spoken is where Translators create the deepest loyalty.

Ask yourself: In your work, are you translating what’s already clear ? Or finding language for what’s still unspoken?

2. Rebuilding without abandoning your voice

When the co-host left and the format had to change, most creators in that position would have pivoted to something safer. Alex kept the core … unfiltered, direct, intimate … and evolved the content around it. The voice stayed consistent even as the subject matter matured.

This is harder than it sounds. Consistency under pressure, when the temptation is to become whoever the room wants you to be, is a distinct skill.

Ask yourself: When you’ve had to reinvent a role or project, what did you protect? What did you let go?

3. Reading the room on ownership

She took a base salary at Barstool. Then she watched what the show became and understood what she had undervalued. The second negotiation wasn’t aggressive, it was informed. She knew what the IP was worth because she had watched it perform.

This is pattern recognition applied to your own value, not someone else’s market. Most people don’t do it. They accept the frame given to them at the beginning of a relationship and don’t revisit it as the evidence changes.

Ask yourself: Where in your career are you still operating on terms you accepted before you had data?

4. Building community before building product

Traditional advertisers initially avoided the show. So she built merchandise, and in doing so, turned listeners into identity. “Daddy Gang” wasn’t a marketing term. It was belonging. By the time big brands came back to the table, she had leverage they couldn’t replicate: a community that had already decided she was theirs.

Ask yourself: Are you waiting for institutional validation before investing in your own community? What would change if you didn’t wait?

Subscribe now

The AI³ Lens

Alex didn’t build with AI. She didn’t build across borders. What she built was a Translator’s core asset … a community that trusted her voice before anyone else recognized its value.

That’s the lesson for our audience.

If you’re a Translator today, AI changes the scale at which your voice can travel. What used to require a full production team, a distribution deal, and a media company’s backing is now within reach of one person with clarity about what they’re saying and who they’re saying it to. The content side is increasingly replicable. The trust side isn’t.

And the borderless dimension isn’t absent from Alex’s story, it just shows up differently. She built an audience identity that crossed every geography, industry, and background. “Daddy Gang” is a community of women who found each other through a shared language, not a shared location. That’s borderless fluency without a passport.

The ownership instinct … knowing what you’ve built, refusing to undervalue it in the next negotiation … applies regardless of your role. Curator, Synthesizer, Translator. Every field. Every geography.

P.S.

The skill I keep coming back to from Alex’s story is the second negotiation. Not the first deal, the one she made when she was still figuring things out. The one she made after she had evidence.

I’m in that in-between space with ChaosClarity™ right now. The framework works. The evidence is starting to accumulate. The harder work is not letting someone else define what it's worth before I do.

Take the Role Quiz →

Anthropic published labor market research this month that should be required reading for everyone in this community. Not because it’s alarming. Because it’s clarifying.

Here’s what they found: AI displacement risk is highest among workers who are educated, higher-paid, and disproportionately female. The most exposed occupations include computer programmers, customer service representatives, and financial analysts. The traditional white-collar career path, the one many of us were told was “safe,” is the one sitting most directly in AI’s path.

We want to sit with that for a moment, because the instinct is to either panic or dismiss it. We’re asking you to do neither.

The gap is the opportunity.

One of the most important findings in the research is this: AI is far from reaching its theoretical capability. Actual task coverage across professions remains a fraction of what’s technically possible. In Computer and Math occupations, for example, AI currently handles roughly a third of tasks that it could theoretically perform.

That gap between capability and reality? That’s not a reassurance. That’s a timeline. And it’s the exact window where strategic positioning happens.

The women who will come out of this transition well are not the ones who waited for certainty. They’re the ones who used the window.

The signal hiding in the hiring data.

The research also found something more subtle, and more urgent, than unemployment numbers. Hiring of workers aged 22 to 25 into AI-exposed occupations has dropped roughly 14% compared to 2022. The mass layoffs haven’t materialized yet. But the entry points into exposed careers are already narrowing.

This matters to our community in a specific way. The traditional pipeline, where you work your way up through established roles over a decade, is compressing from the bottom. Mid-career professionals with cross-cultural experience, strategic pattern recognition, and the ability to synthesize across contexts are not being squeezed out. They’re being surfaced. But only if they know how to name what they bring.

That’s the work we do here.

The AI³ Lens: Role × AI × Borderless

The research describes our community almost exactly. Educated. Higher-paid. Female. Most exposed.

But here’s what the data can’t capture, and what your borderless experience actually gives you.

AI excels at tasks that are well-defined, repeatable, and contained within a single context. What it cannot replicate is the strategic judgment that comes from having operated across multiple markets, languages, and business cultures. The ability to read a room in Tokyo and a boardroom in Berlin. To know that a “yes” in one culture is a “let me think about it” in another. To synthesize not just across data sets, but across lived human contexts.

This is the Borderless dimension of AI³. Not biography. Competitive advantage.

If you’re a Curator, AI multiplies your research capacity, but your cross-cultural credibility filter is irreplaceable. If you’re a Synthesizer, AI surfaces patterns across data, but your ability to know which patterns actually matter in a specific human context is what makes the insight actionable. If you’re a Translator, AI can adapt language, but it cannot carry the trust that comes from genuinely understanding both sides of a conversation.

The question this research raises for each of us is not “am I exposed?” We probably are. The question is: what dimension of your experience is hardest to compress into a task list?

That’s where you build from.

What patterns are you seeing in your own field right now?

Subscribe now

P.S. from Izumi

Something I keep noticing in conversations lately: career development is increasingly framed as “AI-first.” Which tools are you using? How fast can you prompt? What’s your AI stack?

And underneath all of that, I’m hearing something else. A quiet hunger for what’s real. For physical presence, human texture, the kind of experience you can only get by actually being somewhere, with someone, navigating something uncertain together.

I don’t think that’s nostalgia. I think it’s signal. The more AI compresses the cognitive and the informational, the more precious the embodied and the relational become. Living and working across multiple continents taught me things no dataset can hold. That used to feel like a footnote. Right now, it feels like the whole point.

Mallory Contois has worked at Pinterest, LinkedIn, Cameo, and Mercury. She’s held COO roles, led product-growth teams, and advised early-stage founders. On paper, she’s a career technologist, someone who knows how to build things that scale.

But what Mallory actually built, and what she’s best known for, is something most companies wouldn’t think to put in a job description: a space where senior women don’t have to pretend.

In January 2022, she founded Old Girls Club, a Slack-based community for career-focused women in senior roles. It started as something she wanted for herself. She was COO of Metafy, a gaming marketplace startup in Los Angeles, making high-stakes decisions in near-total isolation. “I felt like I really didn’t have anyone to talk to,” she’s said. “I knew a lot of women who were in really senior roles, and yet none of us were actually talking to each other.”

She sent a LinkedIn post. A form. Some Slack invites.

Within a few months, the group crossed 1,000 members with zero marketing spend. By 2024 it had over 1,200 paying members. Today it has around 2,300, generating roughly $300,000 a year, entirely through word of mouth.

The content of the community is straightforward: real questions, real answers, no performance required. What Mallory translated wasn’t a message or a strategy. It was the feeling of work itself.

What This Means for You

Our newsletter series profiles women through the three AI³ roles, Curator, Synthesizer, and Translator, not as labels, but as lenses for understanding how people actually create value.

Mallory’s story is a Translator story. But it’s a specific kind of Translation that often goes unrecognized at work: the ability to make culture legible, to name what people are experiencing but haven’t been able to say, and to build environments where the real conversation can finally happen.

Below are four soft skills at the core of this. You’ll probably recognize at least one of them in yourself.

The Translator Soft Skills: Mallory Contois Edition

Skill 1: Naming the Unspoken

What it is: The ability to identify and articulate experiences that are widely felt but rarely said out loud, especially in professional contexts where maintaining composure is the default.

In practice: When Mallory posted on LinkedIn about feeling isolated as a COO, she wasn’t just venting. She was doing something precise: putting language to a shared experience that had no public name. The response was immediate and overwhelming, because the thing she named had been real for a long time. It just hadn’t been said. The Old Girls Club was built on that moment. Not on a product insight or a market gap analysis, but on a sentence that turned out to be true for thousands of people.

At work, this looks like being the person who says “I think what’s actually happening here is...” in a meeting that’s been circling for an hour. It’s the ability to surface the thing everyone is feeling but politely avoiding, in a way that moves the room forward rather than derailing it.

The question: What is something your team, community, or industry is clearly experiencing, but nobody is saying directly? What would change if you named it?

Skill 2: Designing for Psychological Safety

What it is: Intentionally building conditions, in a team, a meeting, a community, or a conversation, where people feel safe enough to be honest.

In practice: Mallory describes the OGC ethos as “beautifully experimental and chaotic and unpolished.” That’s not an accident. It’s architecture. She spent significant time at the beginning establishing community norms: what the space was for, how to use each channel, what kind of questions were welcome. The goal was to create an environment where members don’t have to “put their mask on.” Her observation is that most professional spaces, including well-intentioned ones, still require people to perform competence, which makes real problem-solving impossible.

Her framework for community interactions: “Would you say this to someone at a coffee shop?” If not, it doesn’t belong in the community. That single question changed the quality of every conversation.

At work, this looks like creating meeting conditions where junior voices are actually heard, checking whether your team’s open door policy is genuinely open, or recognizing that silence in a room doesn’t mean agreement.

The question: Think of a recurring meeting or team dynamic in your work. What would have to change for people to say what they actually think? What is one thing you could do to lower the barrier?

Skill 3: Translating Experience into Systems

What it is: Taking something that exists only as personal intuition or lived experience and turning it into something others can use, repeat, and build on.

In practice: Mallory’s background in psychology, which she’s spoken about publicly as foundational to how she thinks about community and product growth, shows up most clearly not in theory but in operational design. She didn’t just create a “supportive space.” She created a Slack architecture with specific channels designed around specific needs she knew women in senior roles had. She built an onboarding flow. A member directory. A job board with warm intros. A biweekly digest. The community’s consistency comes from systems designed with human behavior in mind, not from Mallory’s personal energy alone.

This is what allows OGC to function at scale without requiring everyone to already be safe, skilled communicators. The structure does some of the translation work.

At work, this looks like turning what you’ve learned through experience into a repeatable process, a team ritual, or a set of guidelines, so the insight doesn’t leave the room when you do.

The question: Where in your work are you the only one who holds a piece of important knowledge, because you’ve never translated it into something others can access? What would it look like to build a system around it?

Skill 4: The Ask

What it is: The ability to ask for help, introductions, and support without over-justifying, over-apologizing, or waiting until you’ve “earned” the right to ask.

In practice: Mallory’s analysis of why the “old boys’ club” works is precise: the men inside it just ask. Introductions, endorsements, information — without requiring a deep personal relationship first, and without a lengthy explanation of why they deserve the favor. Women, she’s observed, often wait until they have a fully developed emotional bond before making a request. “You’re going to be so slowed down by that from a work perspective,” she says. OGC was designed to change that pattern. It gives women a context in which asking is not only acceptable, but expected, where helping each other is the point, not a secondary benefit.

This isn’t about being pushy. It’s about removing the unnecessary guilt around asking for support that many professional women have internalized as professionalism.

At work, this looks like sending the introduction request without a three-paragraph justification, offering to connect someone without waiting to be asked, or building a team culture where “can you help me with this?” is a sentence people say without anxiety.

The question: Is there a request you’ve been holding back because you didn’t feel you’d earned the right to ask? What would you do if you assumed the person would want to help?

Subscribe now

The AI³ Lens: Why Translators Shape How Work Feels, Not Just How It Functions

Mallory’s work reveals something important about the Translator role that doesn’t show up on an org chart.

Most organizational systems are designed to transmit information: strategy, tasks, feedback, metrics. What they’re less equipped to transmit is experience, the felt reality of what it’s like to work inside a system. That gap is where a lot of organizational dysfunction quietly lives.

Translators, at their best, bridge that gap. They make culture legible to the people living inside it. They name the dynamics that are shaping behavior but aren’t being acknowledged. They build systems that encode human insight so it doesn’t disappear when the person who holds it leaves.

This is where the Borderless dimension becomes a specific advantage. The workplace dynamics Mallory is translating — the unspoken isolation of senior women, the friction around asking, the gap between how work looks and how it feels — are not uniquely American or uniquely tech. They are recognizable across industries, geographies, and organizational cultures, because power dynamics and professional performance norms travel.

If you’ve worked across countries or cultures, you’ve likely already noticed this: the same unspoken rules show up in different forms depending on where you are. What’s considered “too direct” in one context is “refreshingly honest” in another. What reads as confident asking in one culture reads as presumptuous in another. That cross-cultural fluency — knowing that the rule exists, and that it varies — is exactly what makes a Translator more effective. You’re not just translating the message. You’re translating the conditions under which the message can be heard.

As AI takes on more of the information-transmission work, summarizing, analyzing, generating, the specifically human layer of Translation becomes more valuable, not less. AI can process what’s written. It can’t yet name what’s unspoken, build psychological safety in a room, or read the cultural context that determines whether an ask lands as confident or inappropriate.

That’s what Mallory does. And if you recognize some of it in yourself, it’s worth knowing that this is a strategic role in an AI-disrupted environment, not just a personality trait.

Not Sure If You’re a Translator?

Translators shape how work feels, not just how it functions. They’re the ones who notice when a team is off, who can read a room before the meeting starts, who make complex or uncomfortable things approachable without flattening them.

Curators ask: What’s worth keeping? Synthesizers ask: What’s the pattern here? Translators ask: How do we make this real for the people who need to act on it?

If the last question sounds like you, the role quiz is worth taking.

Discover YOUR Role - Take the AI³ Quiz

Your Turn

Mallory built something that 2,300 women pay for every month, not because it offers exclusive content or expert advice, but because it offers something harder to find: a space where you don’t have to perform.

Which of these four soft skills feels most like something you already do, but haven’t been calling a skill? And which one feels like the gap?

Drop a reply. We read them all.

Izumi

P.S. The Ask is the skill I’m actively working on. I’m building a consulting practice around the ChaosClarity™ framework, and asking for it to exist in the world — before the client list is long, before the framework is widely known — is still hard. Mallory’s reminder that the ask doesn’t have to come after the proof is one I’m keeping close.

Subscribe now

Sheryl Sandberg spent 14 years as COO of Facebook (now Meta). She joined in 2008, when the company had millions of users and an unproven business model. When she left in 2022, it was one of the most profitable advertising businesses in history. Zuckerberg has credited her with designing the advertising business, shaping the company’s management culture, and teaching him how to run the company.

She was, by any measure, the person who translated a vision into an organization that could actually execute it.

And yet, for most of her tenure, the story was always about Zuckerberg.

That’s not an accident. That’s what operational Translators often experience: the closer you are to making something function, the more invisible the work becomes.

We keep thinking about what that tells us about value, visibility, and what’s about to become urgently needed as AI disrupts organizations everywhere.

The Translator Who Scales Systems

Translators don’t just communicate. They convert.

They take something that exists in one form — a vision, a data insight, a technical concept, a cultural nuance — and rebuild it in a form that others can receive, act on, and trust.

Sheryl did this at scale. She joined a company that spoke the language of social product, and she built a second fluency: the language of revenue, operations, and institutional structure. Then she bridged them. Facebook didn’t just grow because it had a compelling product. It grew in large part because Sheryl helped build the systems that let the product scale — advertising infrastructure, sales teams, HR frameworks, and policy functions — without losing its core identity.

That’s not execution. That’s translation. And it requires a very specific set of skills most organizations underestimate until they need them desperately.

Subscribe now

The Four Soft Skills of Operational Translators

These are the capabilities that let Translators function as the connective tissue between vision and reality. You don’t need a COO title to use them. You might be using them right now without knowing their name.

1. Structural Empathy

What it is: The ability to understand not just what someone needs, but how their context shapes what they can receive.

Sheryl came from Google and the U.S. Treasury. She could have imposed those frameworks wholesale onto Facebook. Instead, she spent time understanding how Zuckerberg thought, how engineers communicated, what the company’s culture valued — and then built operational structures that fit that context rather than fighting it.

In practice: This is the person in a meeting who says, “I think what you’re trying to solve for is X — let me show you how the operations team would frame that.” Not correcting. Converting.

The question: When you’re explaining something important to someone, do you start from your model of the world or from theirs?

2. Translating Without Flattening

What it is: The ability to make something complex accessible without losing the parts that make it true.

One of the most common failures in organizations is oversimplification. A strategy gets translated into a slogan, a nuanced decision becomes a policy, a complex market signal becomes a bullet point in a deck. Every translation step loses something — until the people executing have no idea why they’re doing what they’re doing.

Sheryl was known for being able to communicate across audiences — to engineers, to advertisers, to regulators, to the public — without flattening the underlying logic. The message changed. The integrity of the message didn’t.

In practice: This is the person who writes a one-page brief that actually captures the decision, not just the outcome. The person who explains a technical constraint to a client in a way that doesn’t make the client feel managed.

The question: Think about a complex idea you regularly translate for others. What gets lost in that translation — and does it matter?

3. Cross-Functional Credibility

What it is: The ability to operate with authority across different domains without needing to be the expert in each one.

Sheryl oversaw advertising, sales, human resources, communications, and public policy simultaneously. She wasn’t a specialist in all of them. She was something more valuable: someone who understood how they needed to connect.

Cross-functional credibility isn’t about knowing everything. It’s about knowing enough of each language to build bridges between them, and earning the trust of specialists by genuinely engaging with their domain.

In practice: This is the project manager who earns the respect of engineers not by pretending to code, but by asking real questions and protecting their priorities in broader conversations. The HR director who speaks fluently to finance about the cost of attrition.

The question: In your current role, which team or domain do you least understand — and what would it take to learn their language?

4. Holding the Long View Under Short-Term Pressure

What it is: The ability to stay connected to the strategic horizon even when the immediate pressures are intense.

COO roles are defined by operational urgency. Something is always on fire. The quarterly numbers, the product launch, the regulatory inquiry. Sheryl’s particular skill was maintaining strategic clarity — what are we building toward, and does this decision serve it — even while managing daily crises.

This is harder than it sounds. Short-term pressure has a gravitational pull. The Translators who create lasting value are the ones who can feel that pull and still ask: but is this the right direction?

In practice: This is the manager who, in the middle of a crisis response, asks “does this solution create the next crisis?” The strategist who doesn’t let the urgent permanently crowd out the important.

The question: Where in your current work are you solving the immediate problem in a way that creates a bigger problem later?

What This Means for You (The AI³ Lens)

Here’s what we think is happening right now that matters.

AI is beginning to automate a significant amount of what used to require specialized knowledge. Reporting, research, data synthesis, first-draft communications. The middle layers of many organizations — the people who gathered information and passed it up — are under pressure.

What AI cannot yet do is translate.

It cannot build structural empathy. It cannot earn cross-functional trust. It cannot sense when a strategy is being lost in execution and reconstruct it with enough fidelity to matter. It can produce output. It cannot hold the logic together across a complex, human organization in motion.

This means Translators — especially Translators with cross-cultural, cross-market, cross-domain experience — are not facing irrelevance. They’re facing a window.

The organizations that will navigate AI disruption best are not the ones that automate the most. They’re the ones that maintain coherence as everything speeds up. And coherence — the ability to keep vision, execution, culture, and communication aligned — is what operational Translators have always provided.

Sheryl’s career became visible when she wrote Lean In and when Facebook’s controversies brought her into public scrutiny. For most of her tenure, the work was nearly invisible. That’s the paradox of the Translator role: the better you do it, the more the organization looks seamless, and the less anyone can point to what you specifically contributed.

AI will make that invisibility harder to sustain. Not because Translators will suddenly be celebrated, but because when the connective tissue disappears — when organizations try to run on automation without the translation layer — the gaps will become obvious very quickly.

If you’re a Translator, this is your moment to name what you do.

Not sure if you’re a Translator, Synthesizer, or Curator? Understanding your natural role is the first step to building your strategic advantage — especially as AI reshapes what each role requires.

→ Take the AI³ Role Quiz— and find out where your leverage actually lives.

Take the Role Quiz Now

Which of these four soft skills feels most like you — and which one do you most need to develop? We’re genuinely curious: reply and tell us.

Izumi

P.S. The soft skill I underestimated most in my Wacom years? Translating Without Flattening. I was good at making things clear. I just didn't always notice what I was losing in the process.

Mira Murati spent 6.5 years at OpenAI — joining as VP of Applied AI and Partnerships in 2018, rising to CTO in 2022, and helping shepherd ChatGPT, DALL·E, and GPT-4 from research curiosity to global infrastructure. In November 2023, when OpenAI’s board abruptly fired Sam Altman, she was named interim CEO — steering one of the most consequential companies in the world through a crisis that played out in real time, in public. Then in September 2024, she stepped down to start Thinking Machines Lab, which raised $2 billion at a $12 billion valuation within months of launching.

Most narratives about her focus on the drama, the exits, the funding numbers. We keep coming back to something quieter: how she stayed oriented when everything around her was ambiguous.

That’s a Synthesizer at work.

The Synthesizer Under Pressure

Synthesizers don’t just connect dots. They make sense of situations where the dots themselves aren’t fixed yet.

Murati’s career is a masterclass in this. She’s a mechanical engineer who moved into product management at Tesla, then into augmented reality at Leap Motion, then into AI product and research leadership at OpenAI. Every transition was a move across domains — absorbing different languages, different logics, different stakeholder realities.

By the time she became CTO, she wasn’t just managing engineers. She was sitting at the intersection of frontier research, product development, safety concerns, regulatory pressure, and public perception — all moving at different speeds, often in different directions.

She built cross-functional teams that brought together technical and non-technical perspectives, making the implicit synthesis visible as an organizational structure. She wasn’t just doing the connecting herself; she was building teams designed to synthesize across domains.

That’s an evolved Synthesizer. Not just pattern-spotting — but architecting for pattern-spotting at scale.

Why Synthesizers Are Essential as AI Grows Opaque

Here’s what most AI leadership narratives miss: as systems become more powerful, they also become harder to explain. The gap between what AI does and what humans can understand about what AI does — that gap is widening.

Who bridges it? Not pure researchers (they’re building deeper in). Not pure communicators (they don’t have enough access to what’s actually happening). Synthesizers — people who can move between worlds and translate what they see without losing nuance — become load-bearing.

Murati understood this. Her advocacy for responsible AI development and government regulation wasn’t separate from her technical work. It was an extension of the same skill: seeing how technical decisions land in ethical, social, and policy contexts, and finding a path that holds across all of them.

In a 2023 interview, she said: “It’s important for OpenAI and companies like ours to bring this into the public consciousness in a way that’s controlled and responsible. But we’re a small group of people and we need a ton more input in this system.”

That’s a Synthesizer recognizing the limits of synthesis done only from the inside.

Subscribe now

The Soft Skills Behind the Strategy

This is where it gets practical for the rest of us.

Murati’s Synthesizer capacity didn’t come from technical brilliance alone. It came from a specific set of human skills that don’t show up on a CV but are visible in how she navigated pressure:

Holding ambiguity without collapsing it. When OpenAI’s board crisis unfolded in public, there was enormous pressure to take a clear position fast. Murati held the complexity long enough to act with integrity rather than react with urgency. That’s a practiced skill — tolerating not-knowing while remaining functional. Most of us have been trained to resolve ambiguity as quickly as possible. Synthesizers learn to sit in it longer.

Building trust across different kinds of intelligence. Researchers, ethicists, product managers, policymakers — each group thinks differently, values different things, and speaks a different professional language. Murati’s ability to be trusted by all of them required something beyond being smart. It required genuine curiosity about how other people think, and the patience to make space for that in a room. This is something we can practice in our own teams and communities right now.

Making the invisible visible — without oversimplifying it. One of the hardest things a Synthesizer does is communicate a pattern they can see to people who can’t see it yet — without flattening it into something misleading. Murati did this in public-facing interviews about AI’s societal impact, walking a difficult line between honesty and accessibility. That skill — translating complexity with integrity — is increasingly rare, and increasingly valuable.

These aren’t personality traits. They’re capabilities that develop with deliberate practice, cross-domain exposure, and the willingness to stay in uncomfortable in-between spaces longer than feels natural.

The AI³ Lens: What Multiplies a Synthesizer

The AI³ framework is Role × AI × Borderless = ∞ advantage. So what happens when we apply that to Murati?

Role (Synthesizer): She connects research, product, ethics, and societal impact. Her value isn’t in any single domain — it’s in the movement between them.

× AI: AI tools don’t replace Synthesizer work. They expand the surface area of what needs synthesizing. More data, more domains, more stakeholder groups, more speed. This is exactly where Synthesizers become more valuable, not less.

× Borderless: Born in Albania during the final years of a closed regime, educated across Canada and the US, navigating a Silicon Valley institution with global reach — Murati’s cross-cultural formation isn’t a footnote. It shaped how she reads ambiguity: as something to navigate with judgment, not resolve with algorithms.

The result: a leader who could hold OpenAI’s contradictions — commercial pressure and safety mission, speed and caution, technical ambition and human consequence — long enough to move through them.

What This Means for You

You don’t need to be steering a $90 billion company to put this into practice.

If you’re a Synthesizer, the question Murati’s career raises is: Are you synthesizing across enough domains to see what others can’t?

And if you’re a Curator or Translator reading this — these soft skills aren’t exclusive to Synthesizers. The ability to hold ambiguity, build trust across difference, and communicate complexity with integrity are capabilities every role benefits from as AI reshapes the landscape.

The AI moment we’re in right now is creating more ambiguity, not less. Those who can hold complexity — across industries, languages, cultures, and time horizons — are exactly what’s needed.

Your cross-domain experience isn’t a liability. It’s your load-bearing capacity.

Not sure whether you’re a Synthesizer, Curator, or Translator? The distinction matters more than it might seem — especially in understanding which of these soft skills to develop first.

→ Take the Role Quiz — find out which role you naturally occupy, and how to use it as your strategic starting point.

Take the Role Quiz

Which of these three soft skills feels most underdeveloped in how you currently work — holding ambiguity, building trust across difference, or translating complexity without flattening it?

Izumi

P.S. The soft skill I underestimated most in my corporate years? Holding ambiguity. I was trained to find the answer, present the solution, move forward. It took being disrupted to learn that sitting with not-knowing — long enough to see clearly — is actually where the best strategic thinking happens.

Marie Forleo built a multi-million-dollar education business. But here’s what most people miss: she wasn’t inventing new business concepts. She was reframing familiar principles in emotionally engaging ways.

That’s a Translator role. And in the AI era, this strategic position becomes even more critical.

Translation is influence at scale

When Marie launched B-School in 2010 (co-created with Laura Belgray), business education was dominated by two extremes: dry MBA programs or hyped-up “guru” pitches. She carved out what we’d call a third path, teaching business fundamentals with emotional intelligence and relatable stories.

She didn’t invent marketing funnels or brand positioning. She made them feel achievable instead of intimidating.

The Translator advantage: While Curators gather information and Synthesizers connect patterns, Translators create belief. They make complex ideas emotionally resonant enough that people actually implement them.

What This Means for You

If you’re a Translator, your advantage isn’t having more information. It’s making existing knowledge emotionally accessible and actionable.

But what does that actually look like in terms of skills you can recognize and develop?

Soft Skill 1: Simplification Without Dilution

Marie doesn’t dumb down business concepts. She makes them approachable without losing their strategic power.

This is the ability to distill complex ideas into their essential components while maintaining accuracy. You can explain sophisticated concepts without jargon, making them feel achievable rather than intimidating.

In practice: You’re the person colleagues ask to “explain this to the client.” You can take a 50-page strategy document and create the 3-sentence version that actually drives action. Technical experts trust your simplified versions because you preserve what matters.

The question: What complex topic do you make accessible that others overcomplicate? Where do people say “the way you explained it finally made it click”?

Soft Skill 2: Emotional Calibration

Marie reads her audience constantly. She knows when they need encouragement versus tough love, structure versus inspiration.

This is the ability to sense emotional state and adjust your communication accordingly. You pick up on subtle cues about what someone actually needs to hear versus what they’re asking for.

In practice: You naturally shift tone mid-conversation based on how someone’s responding. You know when “here’s the framework” will land versus when someone needs “here’s why you can do this.” You adjust energy level, pacing, and directness based on what the moment requires.

The question: How do you naturally adapt your communication style to different people or situations? What signals are you reading that others miss?

Soft Skill 3: Trust-Building Through Consistency

Marie’s business runs on trust equity built over nearly two decades, from her early online programs in the mid-2000s through MarieTV (launched 2011) to today.

This is the ability to be reliably yourself while maintaining professional standards. You create psychological safety by being predictably authentic. People know what they’ll get from you, and that consistency builds trust over time.

In practice: Your professional reputation is “I know exactly what to expect from [your name], and it’s always solid.” People return to you not just for expertise but because working with you feels safe. You’re consistent in quality without being rigid in approach.

The question: What makes people trust you specifically with their vulnerable questions or challenges? What do you deliver so consistently that it’s become your reputation?

Soft Skill 4: Metaphor and Story Selection

Marie draws on experiences across finance, publishing, dance/fitness, and coaching, which lets her translate business ideas for very different kinds of audiences. She often blends practical business strategy with mindset and purpose-driven themes, using personal stories and creative metaphors drawn from her background as a former hip-hop instructor and Nike Elite Dance Athlete.

This is the ability to match the right story or metaphor to the right audience at the right moment. You have a mental library of ways to explain the same concept, and you intuitively know which version will land.

In practice: You explain the same idea differently to different stakeholders. Technical teams get system analogies, creatives get artistic metaphors, executives get business cases. Each version is accurate, but each one lands because it speaks in that audience’s language.

The question: What’s your range of translation? How many different ways can you explain your core expertise, and how do you know which version to use when?

The Translator trap: Trying to be all things to all people, losing your authentic voice in the attempt to be universally accessible.

The Translator advantage: These four soft skills compound. Simplification + emotional calibration + trust-building + metaphor selection = you create belief and drive implementation while others are still trying to explain the concept.

Not sure if you’re a Translator?

You might be a Curator (core soft skills: information filtering, research depth, knowledge organization, credibility assessment) or a Synthesizer (core soft skills: cross-domain pattern recognition, comfort with ambiguity, contextual translation, timing intuition).

All three roles are valuable. The key is knowing which soft skills come naturally to you, so you can build your advantage around them.

What AI changes for Translators

Here’s the pattern we’re seeing: AI can now Curate information better than most humans. AI is getting decent at Synthesis, spotting patterns across datasets humans would never manually analyze.

But AI cannot create trust. It can’t read a room. It can’t sense when someone needs encouragement versus tough love, or translate a concept into the exact metaphor that makes it click.

Marie’s real asset wasn’t her curriculum. It was the trust she built by consistently showing up with both competence and relatability.

Role × AI × Borderless:

• Role (Translator): Marie excels at making complex business concepts emotionally accessible

• AI Integration: Translators who use AI to handle research and content can focus entirely on connection and trust-building

• Borderless Advantage: Marie’s cross-contextual range (finance, publishing, dance/fitness, coaching backgrounds) lets her translate business concepts across multiple contexts and audiences

Result: A business model that scales through trust, not just information.

Subscribe now

Teaching vs preaching

Here’s what separates Marie from typical business influencers: she teaches, she doesn’t preach.

Preachers say “Here’s what you MUST do to succeed.” Teachers say “Here’s what worked for me and others, try this approach.”

Preachers create dependency. Teachers create capability.

As AI floods the internet with information, the value shifts from “what to do” to “how to actually do it in YOUR specific situation.” Translation becomes the bottleneck.

Why Translation matters in the AI era

Here’s what’s changing: AI can find endless information and spot certain patterns across data. But it can’t create trust. It can’t read a room. It can’t sense when someone needs encouragement versus tough love.

Translation isn’t about knowing more. It’s about making knowledge emotionally accessible enough that people actually implement it.

Marie’s been demonstrating this since the mid-2000s: belief drives implementation. AI makes this insight more relevant, not less.

Your Translator advantage

If you’re thinking “I’ve always been good at explaining complex things simply,” that’s your Translator role signaling.

Your cross-cultural experience, your industry pivots, your “weird” background: that’s your borderless dimension. It gives you translation range others don’t have.

The opportunity: Focus on the work only Translators can do: making ideas emotionally accessible, building trust through consistency, adapting communication to your specific audience. Let others (or AI) handle the rest.

That’s Role × AI × Borderless applied to education and influence.

Discover Your Role

Are you a Translator like Marie? Or maybe a Curator or Synthesizer?

Take our 5-minute Role Quiz to discover how you naturally create value, and how AI can multiply your specific strengths.

Take the Role Quiz →

Understanding your role is the first step to building your competitive advantage.

Your Turn

What are you translating that makes complex ideas emotionally accessible?

Where does your unique combination of experiences give you translation range others don’t have?

What would you teach if you focused on creating belief rather than transferring information?

Drop a comment below: Who do you admire for making complex ideas feel achievable?

Someone who teaches without preaching, builds trust through consistency, or translates across contexts in ways that create real capability.

Izumi

Lucy co-founded Scale AI at 21, building the data labeling infrastructure that would become essential for training AI models. After leaving Scale in 2018, she founded Backend Capital, investing in technical founders building frontier technologies. She’s now CEO of Passes (creator monetization) and co-founder of HF0, an intense residency-style accelerator backing technical founders, many building AI agents and automation tools.

Most people see her story as exceptional talent meeting perfect timing. I see something more strategic: a Synthesizer who spots patterns across disconnected domains before they converge.

The Synthesizer Advantage

Lucy’s pattern: she doesn’t wait for consensus. She synthesizes signals from technology, markets, and human behavior to identify inflection points.

Scale AI (2016): While most people were debating whether AI would work, Lucy saw a different pattern: AI companies needed massive amounts of labeled data. The breakthrough wasn’t the algorithms. It was the human-in-the-loop infrastructure to train them.

She connected three things most people treated separately:

• Machine learning models needed supervised learning at scale

• Existing data labeling was artisanal and didn’t scale

• Global labor marketplaces could be orchestrated through software

Scale AI became one of the key infrastructure providers for AI, working with major labs and U.S. government agencies. In 2024, it was valued at about $14 billion.

Backend Capital (2019): After leaving Scale AI in 2018, Lucy didn’t just invest in “AI companies.” Through Backend Capital, she backed technical founders building infrastructure for problems that didn’t fully exist yet. She focused on giving early-stage technical founders their first institutional check, not because she could predict winners, but because she recognized the pattern of how technical insight becomes market opportunity.

Her portfolio companies solve problems that emerged as technology created new bottlenecks: developer tools, data infrastructure, security for distributed systems.

Now (Passes & HF0): Lucy is operating across multiple layers of synthesis. As CEO of Passes, she’s building creator monetization tools. Through HF0, she’s backing and helping build companies focused on AI agents and automation, the orchestration layer between human intent and AI execution, rather than new foundation models themselves.

The pattern she’s synthesizing now: AI capabilities are advancing faster than most organizations can integrate them. The opportunity isn’t better AI. It’s better ways to deploy AI.

Role × AI × Borderless

Her Role (Synthesizer): Lucy spots patterns across technology, capital, and labor that others treat as separate domains. She doesn’t wait for validation. She builds when she sees the convergence.

AI Integration: AI doesn’t replace synthesis. It accelerates the Synthesizer who already sees patterns. Lucy used AI to build Scale, used her understanding of AI to invest through Backend, and now uses both lenses to support the next layer of builders.

Borderless Advantage: Lucy bridges multiple contexts: engineer, founder, investor, builder. She moves between building companies and funding them, between working with government and startups, between consumer and enterprise. This cross-pollination is her edge.

= ∞ (Infinite Advantage)

Lucy’s advantage isn’t just technical skill or capital access. It’s recognizing inflection points before they’re obvious, and acting while others are still debating.

What This Means for You

If you’re a Synthesizer, your advantage isn’t prediction. It’s pattern recognition across domains others keep separate.

But what does that actually look like in terms of skills you can recognize and develop?

Soft Skill 1: Cross-Domain Pattern Recognition

Lucy didn’t invent AI or crowdsourced labor. She saw how they’d need to work together before either was mature.

This is the ability to connect ideas from different fields, industries, or contexts that others treat as unrelated. You see similarities in structure even when the surface details look completely different.

In practice: You’re in a marketing meeting and reference a solution from supply chain. You read about education and immediately think “this applies to healthcare.” People ask “how did you even think of that?”

The question: What patterns are you seeing across different domains that others are treating as separate problems?

Soft Skill 2: Comfort with Ambiguity and Incomplete Information

Scale AI was controversial. Many thought data labeling was boring infrastructure. Lucy built anyway.

This is the ability to make decisions and take action without perfect information. You can form hypotheses with partial data while others are still waiting for certainty.

In practice: You start prototyping while others are still researching. You’re comfortable with “test and learn” when colleagues need “perfect before launch.” You sense when “we need more data” is actually procrastination.

The question: Where are you waiting for consensus instead of acting on what you already see clearly?

Soft Skill 3: Contextual Translation

Lucy bridges multiple contexts: engineer and founder, builder and investor, government and startups, consumer and enterprise.

This is the ability to move between different professional languages and cultures, understanding not just vocabulary but underlying assumptions and incentives in each world.

In practice: You naturally translate between technical and business teams. You hear what stakeholders mean beneath what they say. You notice when different groups are describing the same problem with completely different words.

The question: What different “languages” or contexts do you bridge naturally? How could your multilingual perspective become your strategic advantage?

Soft Skill 4: Timing Intuition

Lucy built Scale before “MLOps” (machine learning operations) was a category. She invested in infrastructure before “AI tooling” was defined.

This is the ability to sense when multiple trends are converging, distinguishing between “too early” and “just early enough” without waiting for market validation.

In practice: You have strong gut feelings about when to move. You’ve been right about trends before they became obvious. You’re watching multiple signals and can feel when they’re about to intersect.

The question: What convergence are you sensing that doesn’t have language yet? What would you build if you trusted your timing over the market’s consensus?

The Synthesizer trap: Waiting until you can articulate these patterns perfectly or until others validate what you’re seeing.

The Synthesizer advantage: These four soft skills compound. Pattern recognition + comfort with ambiguity + contextual translation + timing intuition = you see and act on opportunities while others are still debating whether they exist.

Not sure if you’re a Synthesizer?

You might be a Curator (core soft skills: information filtering, research depth, knowledge organization, credibility assessment) or a Translator (core soft skills: simplification, storytelling, persuasion, audience adaptation).

All three roles are valuable. The key is knowing which soft skills come naturally to you, so you can build your advantage around them.

Discover Your Role

Are you a Synthesizer like Lucy? Or maybe a Curator or Translator?

Take our 5-minute Role Quiz to discover how you naturally create value, and how AI can multiply your specific strengths.

Take the Role Quiz →

Understanding your role is the first step to building your competitive advantage.

Your Turn

What patterns are you seeing that others haven’t connected yet?

Where are you waiting for the market to validate what you already understand?

What would you build if you trusted your synthesis over consensus?

Drop a comment below: Who do you admire for seeing opportunities before they’re obvious?

Someone who connects dots others miss, moves fast on insight, or builds at intersections no one else is watching.

Izumi

Subscribe now

Everyone talks about Bumble’s “women message first” feature as if Whitney Wolfe Herd improved online dating. But that misses what actually happened.

She didn’t make dating apps better. She asked: What if the fundamental dynamic everyone accepts doesn’t have to exist?

This week, I’ve been thinking about that distinction — between improving a system and questioning whether the system’s baseline assumptions are even necessary.

Whitney’s Inflection Point

The moment that reveals the Curator role:

The entire dating app industry built on one invisible assumption: whoever initiates contact sets the power dynamic for the entire interaction.

Everyone accepted this. Some tried to mitigate it (reporting features, blocking, content moderation). But they were all trying to fix problems within the inherited baseline.

Whitney did something different: she questioned whether that baseline had to exist at all.

The unconventional move: Treating a power dynamic as a design choice, not a law of physics.

What looks like “women message first” is actually: what if we delete the assumption that whoever messages first controls the interaction?

The Curator Role Explained

Connecting Whitney’s approach to the broader Curator archetype:

This is what strategic Curators do that others miss.

Most people think curation is filtering — choosing the best from what exists. But the real curator skill is seeing which entire categories don’t need to exist.

What Strategic Curators Actually Do:

• Recognize which “that’s just how it works” is actually optional

• Identify inherited defaults masquerading as requirements

• Question the baseline before trying to improve it

• See which rules can be deleted, not just which rules can be added

The pattern: While others optimize within existing constraints, Curators question whether the constraints themselves are necessary.

This isn’t just philosophical. It’s how new categories get created.

AI³ Lens: The Curator’s Unique Advantage in the AI Era

The multiplication effect: Curator × AI × Borderless

Here’s Why This Matters More Now:

AI inherits our baselines. It trains on what exists, optimizes existing categories, and assumes current defaults are correct.

If you don’t question the baseline before AI scales it, you’ve just automated the inherited assumptions.

The Curator’s AI Multiplier:

This is the Curator’s specific advantage in an AI-saturated world:

You decide which baselines to feed the system.

• AI can generate infinite variations within a framework

• Curators decide whether the framework itself is worth scaling

• The combination means: AI amplifies your judgment about what deserves to exist, not just how to execute what already exists

Practical example: A Curator in content strategy doesn’t just use AI to generate better headlines. They question whether “engagement metrics” should be the baseline measure at all. They decide which goals are worth optimizing before AI optimizes them at scale.

The Borderless Multiplier

How cross-cultural experience reveals optional defaults:

Here’s what makes borderless experience multiply the Curator advantage:

When you’ve lived in multiple countries, you’ve seen the same problem solved with completely different baseline assumptions. You know which “universal truths” are actually local conventions.

Whitney’s insight came from seeing power dynamics play out differently across contexts, then recognizing she could choose which dynamics to design for.

For AI³ members who are Curators:

• You’ve seen which “industry standards” are just inherited local practices

• You know which baselines are culturally specific vs. actually structural

• You can spot which conventions people are exporting as solutions when they’re actually optional problems

The borderless Curator advantage: you’ve already practiced questioning baselines just by navigating different systems. You know the water you’re swimming in is optional.

Subscribe now

Application: The Curator Questions

Embedded naturally in the flow:

If you’re a Curator (or discovering you might be), here’s what Whitney’s story suggests:

Your strategic value isn’t just filtering better. It’s questioning which categories should exist in the first place.

The questions that reveal curator thinking:

• What am I trying to optimize that shouldn’t exist at all?

• Which problem am I solving that isn’t actually necessary?

• What does everyone in my industry accept as baseline that’s actually a choice?

• Which “that’s just how it works” is worth deleting before improving?

These aren’t abstract. They’re strategic positioning questions that create new market categories.

Discover Your Strategic Role: Take the Quiz

Reading about Whitney’s Curator approach, you might be recognizing patterns in how you naturally work. Or you might be thinking: “This doesn’t sound like me…I work differently.”

That’s exactly right. Curators question baselines. Synthesizers connect patterns across disparate information. Translators make complex ideas accessible and persuasive.

Each role has a distinct advantage in the AI era. Each multiplies differently when combined with AI capabilities and borderless experience.

Which one are you?

Take our role assessment to discover whether you’re a Curator, Synthesizer, or Translator. Understanding your strategic role is the first step to building your competitive advantage.

Take the AI³ Role Quiz

(Takes 5 minutes. No email required to see your results.)

Closing Reflection

Whitney Wolfe Herd built a billion-dollar company not by fixing dating, but by questioning which dynamics had to exist at all.

What baseline in your work are you treating as immutable that’s actually just... inherited?

I hear something different: he’s describing the exact roles that create competitive advantage in an AI-disrupted landscape. He just doesn’t have language for it yet.

Dimon told Fox News that as AI transforms work, the skills that matter are “critical thinking, how to be good in a meeting, how to communicate, how to write.” Satya Nadella said emotional intelligence and empathy are becoming increasingly important. Ginni Rometty called out collaboration, judgment, and adaptability.

Here’s what they’re actually describing:

Curators excel at critical thinking and judgment—filtering signal from noise when AI floods us with information.

Synthesizers connect patterns across contexts—the “how to be good in a meeting” insight that sees what others miss.

Translators master communication and persuasion—making complex ideas accessible and driving action.

These aren’t “soft skills” you add to your resume. They’re strategic roles that multiply in value when combined with AI capabilities and borderless perspective.

The CEOs are right that AI will eliminate jobs. What they’re missing is the framework for understanding which capabilities become more valuable, not less. Most people will scramble to “learn EQ” without recognizing they already operate primarily in one of these roles—they just haven’t claimed it as their strategic positioning.

When Dimon says “the next job may be a better job, but they have to learn how to do the job,” he’s describing the gap we’re focused on: people who see disruption happening but lack the framework to reposition themselves strategically.

Your role (Curator, Synthesizer, Translator) × AI integration × borderless thinking = the competitive advantage CEOs are trying to articulate when they talk about “soft skills.”

The question isn’t whether you have emotional intelligence or communication skills. It’s whether you understand your natural role well enough to use it as strategic positioning while others are still treating these as generic competencies to “add.”

Not sure which role describes you? Take the 3-minute quiz to discover your strategic positioning—and what it means for how you leverage AI disruption.

Take the Quiz

What patterns are you seeing in how AI is reshaping your industry? And which role are you claiming while others are still figuring out what “learn your EQ” actually means in practice?

— Izumi

Adobe, InDesign, Photoshop — those tools existed for decades before Canva. Professional designers had everything they needed. The problem wasn’t the technology. The problem was that 99% of people couldn’t use it.

So Melanie did something most founders overlook: she translated power into accessibility.

That’s why Canva is now valued at over $40 billion with 220+ million monthly users.

The Translator Role in Action

When Melanie was teaching design to students in Perth in 2007, she watched them struggle with software that required entire semesters just to learn where the buttons were. She saw the gap between what design tools could do and what most people needed them to do.

Instead of building “better design software,” she asked a different question: What if design software worked the way non-designers actually think?

That’s the Translator mindset. Translators don’t create new capabilities — they make existing capabilities usable by people who don’t speak the technical language. They reduce friction. They bridge gaps. They turn “only experts can do this” into “anyone can do this.”

Melanie translated professional design tools into drag-and-drop simplicity. She didn’t dumb it down—she made it intuitive. And in doing so, she unlocked a market of 220+ million users who would never have touched Adobe Creative Suite.

Subscribe now

Her Curator Instinct: Ruthless Simplification

Here’s where her secondary role as Curator shows up: Melanie didn’t try to replicate every feature professional design tools had. She curated relentlessly.

Curators excel at research and filtering. They ask: What matters most? What can we remove? What do people actually need versus what they think they need?

Rather than trying to “boil the ocean of graphic design all at once,” as she described it, Melanie and her co-founder Cliff Obrecht started with one narrow slice: Australian high school yearbooks through their first company, Fusion Books. They proved the model worked before expanding.

When they launched Canva in 2013, they focused on what most people actually needed: social media graphics, presentations, simple marketing materials. They curated templates. They curated fonts. They curated the entire experience around speed and ease.

The result? While Adobe was adding more complexity for professionals, Canva was removing complexity for everyone else. That’s the Curator-Translator combination in action.

Why This Matters for Your Strategic Advantage

Most people think innovation means building something new. Melanie proves that sometimes the biggest opportunity is translating what already exists.

If you’re a Translator, your competitive edge isn’t creating breakthrough technology — it’s making complex things simple enough that adoption scales. The AI disruption we’re seeing right now? It’s creating massive Translator opportunities. Someone needs to make AI tools understandable to the marketing manager who just wants better content. Someone needs to translate technical AI capabilities into workflows non-technical teams can actually use.

And here’s where AI multiplies the Translator role: AI can now handle the technical complexity while you focus on making it accessible. You don’t need to code. You need to understand what people struggle with and how to remove that friction.

Melanie did this before AI. Imagine what Translators can build with AI as their multiplier.

The Borderless Dimension

One more thing worth noting: Melanie built Canva from Perth, Australia — not Silicon Valley. She raised funding across borders, meeting Silicon Valley investor Bill Tai at a conference in Perth and eventually moving to Sydney to access better tech talent. She built a global team spanning Beijing, Manila, and beyond. She created a product now used in over 190 countries.

That’s the Borderless advantage in practice. Geographic constraints didn’t limit her market. Cultural diversity informed her design choices. Her ability to think beyond local norms helped Canva become a global standard.

If you’ve lived across cultures or markets, you already have this lens. You see patterns others miss. You understand that “simple” looks different in different contexts. That’s strategic advantage.

What role do you default to when things get complex?

Do you try to build something entirely new (Synthesizer)?

Do you research and filter what already exists (Curator)?

Or do you focus on making the complex accessible (Translator)?

Take the 3 minute Quiz

There’s no wrong answer. But knowing your natural role changes how you position yourself in disruption.

Melanie knew hers. And that clarity built Canva.

What could you build if you leaned into yours?

Izumi

Someone in their mid-30s, working in research and insights, shared that they’ve been laid off three times in recent years. They’re watching AI tools roll out at their current company — restricted to just Copilot, which they’ve tried for research analysis and found “terrible.” Their question to the group was essentially:

“I’m wondering how I can keep working in a job I actually like until retirement. Does anyone have thoughts on whether industries like CX or marketing are going to be completely taken over by AI?”

One response in particular caught my attention because it perfectly illustrates what AI³ is all about.

Same background, different approach

Someone else with the same UX research background replied with something that made me pause:

“What we’re really seeing isn’t AI replacing people the way the headlines suggest — it’s companies making restructuring decisions, and AI becomes the convenient explanation.”

They went on to share their own experience:

“I have the same research background. Over the last couple of years, I’ve moved into work where I’m basically designing the collaboration between people and AI — figuring out where AI should handle things versus where you really need human judgment. The funny thing is, this has actually made my position more stable, not less.”

Same industry. Same disruption. Completely different outcome — because of how they repositioned.

What repeated layoffs actually tell you

When someone gets laid off three times in a few years, conventional wisdom says: unstable industry, bad luck, maybe not the right fit.

But here’s what I see: someone who keeps getting positioned in roles that are becoming commoditized — and doesn’t yet realize that’s valuable strategic information.

The person who replied read that same signal differently. Instead of asking “How do I protect my research role?” they asked: “What new role is being created by this disruption?”

Their answer: helping organizations figure out how people and AI should work together. Drawing the lines between what AI handles and what needs human judgment.

That’s not a job that existed three years ago. It exists because of AI disruption. And they positioned themselves to claim it while others were still defending the old role.

Understanding your strategic role

This is where our three-role framework matters.

Your strategic role is how you naturally create value:

Curators excel at research, gathering information, and identifying what’s credible and relevant. When AI floods organizations with information, Curators become more valuable — someone needs to know what sources actually matter and what questions to ask.

Synthesizers spot patterns across disparate information and see how things connect strategically. When AI generates outputs without context, Synthesizers become essential —s omeone needs to see the patterns AI misses and connect insights to business implications.

Translators make complex ideas accessible and persuade others to take action. When AI can generate content but not persuasion, Translators become critical — someone needs to make technical insights land with the people who make decisions.

Each role gets multiplied by AI differently. The key is understanding which role is yours, so you can position yourself where AI amplifies your natural strengths rather than competing with them.

Want to discover your strategic role right now?

Take the 3-minute assessment here →

The pattern that matters

The original poster mentioned that their company’s Copilot implementation produces “terrible” research analysis. But think about what the person who replied is now doing: “designing the collaboration between people and AI — figuring out where AI should handle things versus where you really need human judgment.”

In other words, they’re solving exactly the problem the original poster is experiencing. They’re the person companies need when their AI tools produce bad outputs and they can’t figure out why.

Someone needs to bridge that gap. This person positioned themselves to do exactly that — not by learning to code, not by “upskilling” in the conventional sense, but by understanding what they’re naturally good at and repositioning around the emerging need.

When I read that Facebook exchange, I saw two people with the same background facing the same disruption:

First person:

• “How can I keep working in a job I actually like until retirement?”

• Seeing AI as a threat to their current role

• Treating repeated layoffs as evidence of failure

• Focused on survival

Second person:

• “What we’re really seeing isn’t AI replacing people — it’s companies making restructuring decisions”

• Seeing AI as creating demand for new capabilities

• Positioned themselves at the intersection of human judgment and AI capabilities

• Focused on strategic advantage

Same circumstances. Different strategic thinking. Completely different outcomes.

Subscribe now

What this means for us

The conventional wisdom right now is: “Hold on tight, upskill fast, hope your job survives.”

But that’s not what we’re building here. We’re asking:

“What unique combination am I building that only becomes possible because of AI disruption?”

The person who repositioned found their answer: someone who understands research AND understands how to design the boundaries between AI capabilities and human judgment. That combination didn’t exist as a role three years ago.

What’s your combination?

Nobody’s teaching the right framework for thinking about this strategically. That’s what we’re doing here. That’s why the three-role lens matters. And that’s why the women who are repositioning based on strategic role rather than clinging to job titles are going to have very different next five years than everyone else.

Ready to find your strategic position?

Our next AI Opportunity Mapping Workshop is Wednesday, January 28th at 18:00 CET (90 minutes, online).

What you’ll do:

• Take a 3-minute assessment to discover your strategic role (try it here before the workshop)

• Map your complete role balance and your borderless dimension

• Identify where AI multiplies YOUR specific combination

• Create your personal opportunity map with 2-3 concrete positioning opportunities

What you’ll get:

• Your completed opportunity map

• Framework for identifying opportunities during disruption

• Access to AI³ community

• Integration Session invitation (follow-up where participants share implementations)

Early bird pricing: €16

Save your spot here

What patterns are you seeing in your industry that look like threats but might actually be repositioning opportunities?

Reply or comment to this newsletter — I’d love to hear what you’re noticing.

Izumi

Today I want to focus on the dimension most people undervalue: the Borderless part of the AI³ formula.

There’s a line in the research I referenced that made me stop:

“A graphic designer in Mumbai can work with a startup in San Francisco. A software developer in Warsaw can build products for a company in London.”

The article frames this as revolutionary—the “borderless work” era. But here’s what struck me: for those of us who’ve already lived and worked across countries, this isn’t new. It’s just how we’ve always operated.

What’s new is that this way of thinking has suddenly become competitive advantage.

If you’ve worked in multiple countries, you don’t just understand different markets—you understand how business culture translates. Or doesn’t. If you’ve navigated visa systems, tax codes, and professional licensing across borders, you’ve developed a skill most people don’t even recognize as a skill: the ability to operate in complexity without needing everything to be familiar.

Most people treat this as biographical footnote. “I’m originally from X, moved to Y, now I’m in Z.”

But in the AI³ framework, this is strategic multiplier.

The research points out that “navigating international tax and labor laws remains a key challenge for global self-employment.” True. But if you’ve already lived in multiple countries, you’ve already climbed the learning curve that stops people who’ve only worked in one place. You’ve already developed the cognitive flexibility that makes borderless work natural rather than overwhelming.

Here’s why this matters right now: the professionals thriving in this shift aren’t just the ones with technical skills. They’re the ones who combine technical skills with cross-cultural fluency.

A Curator who knows which sources are credible across different regions can filter signal from noise globally—not just locally.

A Synthesizer who spots patterns across European and Asian markets sees opportunities that someone operating in a single market will miss.

A Translator who understands how to make ideas land in both German and American business contexts has exponential value compared to someone who only speaks one cultural language.

The research says “skill matters more than location.” I’d add: the skill of understanding multiple locations is itself a skill that matters.

But here’s the thing — most people don’t know which role they naturally play. Are you actually a Curator? A Synthesizer? A Translator? Because how your borderless experience becomes strategic advantage depends entirely on your natural role.

A Synthesizer with cross-cultural experience builds advantage differently than a Translator with the same background. Same experiences, different strategic positioning.

This is exactly why we created the role assessment. It takes 3 minutes and reveals which of the three roles you naturally occupy—Curator, Synthesizer, or Translator.

Because here’s what changes once you know your role: you suddenly see your international experience through a completely different lens.

A Curator with cross-cultural experience leverages it differently than a Synthesizer with the same background. A Translator with borderless fluency uses it in ways a Curator never would.

Same experiences. Different roles. Different strategic positioning.

The quiz is your entry point to understanding your natural role—which opens up an entirely different dimension of how you consider your international experience.

Take the 3-minute assessment →

Once you know your role, suddenly your cross-cultural background stops being a footnote and starts being the foundation of your strategic advantage.

What borders have you crossed that you’re treating as background instead of leverage?

—Izumi

P.S. The self-employment research for context: The Self-Employment Revolution

The piece identifies three forces: economic volatility making traditional employment unstable, AI creating both threats and opportunities, and borderless work opening global markets.

Sound familiar?

These are the exact three dimensions we’ve been talking about—Role, AI, and Borderless thinking.

But here’s what the article misses: the professionals who will thrive in this shift aren’t necessarily going independent. They’re developing what the research calls “security through independence”… building capabilities that create value whether employed or self-employed.

Think about it through the role lens.

If you’re a Curator who filters signal from noise, that’s valuable to clients and employers. If you’re a Synthesizer who spots patterns across markets, that skill works in any employment structure. If you’re a Translator who makes complexity accessible, you’re needed regardless of your org chart position.

Find out which "Role" you are

The article mentions “AI is a double-edged sword for self-employed professionals”—automating some tasks while creating demand for AI-related skills.

But here’s the deeper insight: AI is revealing which professionals understand their strategic role and which don’t.

If your value is task-based, you’re vulnerable in any employment structure.

If your value comes from your specific Role × AI × Borderless combination, you’re positioned for multiple futures simultaneously.

One detail from the research struck me:

“Workers aren’t just leaving traditional employment because they want to. They’re leaving because the traditional employment contract has broken.”

This isn’t about everyone becoming freelancers. It’s about everyone needing to think strategically about their positioning, employed or not.

The question isn’t “Should I go independent?”

It’s “Am I building the kind of strategic positioning that creates options regardless of how the employment landscape unfolds?”

Our January 28th workshop walks through exactly this, mapping your Role × AI × Borderless combination to identify opportunities only you can see because of your specific vantage point. What are you seeing in your industry?

Join the workshop on Jan 28th

—Izumi

P.S. Full article: The Self-Employment Revolution