Juggling two dogs, tangled leashes and a sloshing coffee during this morning’s sprint around the neighbourhood, I caught an interview with Tod Maffin on CBC’s The Current.

He was talking about the launch of Main Street Canada, a site that connects Canadians with small businesses affected by U.S. tariffs.

A few minutes into the conversation, an old feminist line came to mind: “The personal is political.”

In 2026, a related truism feels newly urgent: the personal is economic.

The ordinary choices we make like where to spend our hard earned dollars, what to recommend, which businesses to trust with our money all help decide who gets to earn, own and remain independent.

This matters a lot right now.

The acknowledged political rupture to the south has turned economic dependence from a policy discussion into something we now feel. Tariffs can destroy industries. Long-standing supply relationships now liabilities. Decisions made in Washington are landing in the payroll accounts of business owners on Vancouver Island and in Sarnia alike.

Buy Canadian is no longer just an expression of national pride. It feels like a new economic imperative.

Like many other Canadians, we’re changing our purchasing and travel habits to reflect this stark new reality.

Each decision we make determines where our money lands. Personal conviction becomes economic action.

Ten people can change a month

Tod Maffin is an acerbic, witty and sharp Canadian commentator I recently started following on Instagram, and a former CBC presenter. A recent video of his about buying Canadian clearly hit a nerve: roughly five million views.

Now he and his wife have launched Main Street Canada showcasing small Canadian businesses affected by U.S. tariffs.

It joins the raft of lists Canadians are circulating among friends: ditch Heinz, buy French’s; so long Domino’s and helllllo Boston Pizza. You get the point.

Small-business owners submit short videos showing what they sell. Maffin puts those videos in front of people who might buy.

After he featured a small Canadian belt manufacturer, five orders arrived in a single day - the most the business has ever received in 24 hours.

Five orders may not sound like a pathway to economic sovereignty. For that owner, they changed the day.

Jacqueline Corbett made the stakes even plainer. Her Canadian children’s clothing company, Mini Mioche, had relied on the U.S. for about half its business. When tariffs hit, those sales tanked.

She doesn’t need millions of Canadians to rally around her company. She needs ten people a day to make a different purchase.

“Ten people is not an insurmountable challenge for a country of this size,” she told The Current. But ten purchases can fortify her business.

Small businesses employ 5.8 million Canadians. Inside any one of them, viability and staying power can come down to some pretty small numbers.

The lived economics of a small business often operate at a startlingly intimate scale.

Ten purchases. Or five clients. Or one referral that reaches the right person.

One purchase may be small but its effect extends beyond the sale - from payroll, to rent, to paying another supplier … to the owner’s ability to keep going.

When I need a speaker or friend gift, I browse Gather Books, owned by the ebullient Julie or pick up a piece of jewelry from the talented Helen of Lazy Maisie. When we hosted our first two Generous AI Society tables, we turned to Peak Provisions, owned by sommelier and entrepreneur Martha. A huge part of what I love when I can support a local business, or an entrepreneur whose story inspires me, is knowing the human on the other side of the transaction.

In each case, these choices put money into businesses I value, owned by women in my community. In each case, a personal decision becomes revenue that can help meet payroll and keep a small business viable.

The same kind of math applies to independent professionals. A friend with a small consulting practice does not need hundreds of clients. Five people paying a fair monthly fee allow her to sustain an engaging and independent working life.

For each client, it is one modest contract. For her, five of those contracts create a viable business.

The economics of discretion

It’s sometimes hard to feel much personal agency in markets dominated by near-oligarchic companies whose advertising, algorithms and platforms shape what we notice and purchase.

AI adds another layer of fog: more cheap output and automated persuasion, with little clarity about who actually made the thing or whether an actual business is even on the other side of the screen.

AI is also generating and amplifying fake reviews, endorsements and impersonations—sigh—making it even harder for real businesses to earn our attention and trust.

That makes deliberate choice more important.

We exercise economic discretion when we decide what to buy, who to hire, which product to recommend and whose name to put forward.

Individually, these may be small decisions. Repeated across households, companies and communities, they influence who receives revenue, who builds ownership and who has enough demand to continue.

No single personal choice fixes a macroeconomic nightmare. It can impact which business gets to keep going.

We choose who gets the opportunity

We direct economic opportunity even when no purchase is being made.

We do it when we choose a consultant, recommend an accountant, introduce a founder to an investor or add an independent business to a supplier list.

A purchase routes money. An introduction, recommendation or invitation can determine who gets the opportunity to earn it.

This wider lens matters when we talk about women supporting women. Women are making these decisions as customers, business owners, executives, investors, board members and trusted connectors—in a country where majority women-owned businesses account for roughly one in five private-sector businesses.

Access to customers and opportunity begins long before a final purchasing decision. It begins with who comes to mind, whose name gets passed along and who gets a chance to be seriously considered.

A founder is introduced to an investor. A consultant is recommended for a contract. An independent business is added to the supplier list.

Someone makes the call and says, “You should talk to her.”

The work of organizations like The Forum, WeBC and Community Futures remind us that vibrant small business ecosystems require more than conscientious shopping. They require access to customers, capital, advice, networks and opportunity.

Groceries, restaurants, clothing and gifts from local artisans are visible expressions of buying Canadian. They are obviously not the only economically consequential purchase decisions we make.

The same question follows us into software and technology. Maffin points to Canadians deleting Google Maps after the Lake Ontario naming controversy.

His larger point: you don’t need to control a corporate procurement budget to make a technology choice. We decide which search engines, streaming services, software - and maps - earn our use and our data.

Yes, companies and institutions can become the first serious customers for Canadian startups, giving them revenue, proof and a foothold. But individuals are customers too. It may be harder to spot the Canadian option in a software stack than in a grocery aisle, but our tech choices are economic choices all the same.

Subscribe now

Personal choices and public consequences

Buy Canada doesn’t let government off the hook. Government action and personal action are not competing strategies for building a more resilient economy. We need both.

The ability to exercise purchase discretion is unevenly distributed. Those feeding growing children while grocery bills grow with them know that price, availability and convenience are not trivial concerns. We cannot all absorb a higher price or spend hours investigating the ownership of everything we buy.

So the maple leaf appearing on shelves, menus and websites becomes more than decorative. Many of us are actively looking for it as a small way of asserting sovereignty in the ordinary decisions of daily life.

Made here, owned here, sold here and controlled here are not interchangeable. We can care about those distinctions without stripping the purchase of its personal or patriotic meaning.

Policy matters. So do thousands of everyday choices about what we buy, what we try and whether we look beyond the biggest brand name on the screen.

You’ll find me in local stores or mumbling in the grocery aisles as I inspect the produce’s provenance.

I’ll be buying the book from Julie, hosting gatherings at Martha’s, trying the Canadian technology and putting a good woman’s name forward when someone asks, “Who should I talk to?”

None of this is grand. It i

s shopping, hiring and recommending with my eyes open.

The personal is economic because our choices accumulate. Where I have a real choice, I want to understand whose future I’m helping to keep possible.

Keep reading …

The Red Thread is where I think in public about women, work, AI, entrepreneurship and the messy, creative overlap between professional and personal life.

Subscribe for weekly articles to your inbox!

Subscribe now

The latest Big Tech earnings made me wonder how much of that urgency is also being driven by capital commitments already made by these companies: hundreds of billions invested in data centres, chips and power that now need customers.

Through broad index funds, I own a minute part of many of the companies financing the labs, supplying the chips and building the infrastructure. I also pay for AI tools and use them for hours most weekdays. I am invested, financially and otherwise.

Walking the dogs this week, I listened to and interview , the NYU finance professor known as the “dean of valuation,” on Prof G Markets. The episode was called “Big Tech Has No Idea How AI Pays Off.” That title does some heavy lifting. It’s also a fair summary of Damodaran’s central concern: Microsoft, Amazon, Alphabet and Meta have yet to explain what kind of business all this spending will eventually produce.

Will it be a high-volume operation with thin margins? Or a premium one with fewer customers? Who will keep paying for these products and services, and how much?

The rampant spending is already changing the nature of these companies. Previously they were valued as asset-light businesses capable of producing more revenue without adding an equivalent amount of physical infrastructure. Now they are buying chips, securing power, building data centres and taking on the depreciation and operating costs that come with them.

They are starting to look positively industrial.

Meta’s latest quarter gives a sense of the shift. The company generated nearly $32 billion in operating cash flow and spent $31 billion on capital expenditure. In 2026 alone it expects to spend between $130 billion and $145 billion on capital expenditures.

This is mighty expensive infrastructure to build before it is clear which products and services customers will pay for at scale.

A big buildout with a short customer list

As we live through this AI boom, computing power is discussed as though scarcity were the main problem. There are now some signs that story may be changing.

Meta is reportedly developing a cloud business that could sell access to its excess computing capacity. SpaceX, which built its Memphis data centres for its own AI ambitions, has already struck deals to rent computing power to Anthropic.

Selling spare capacity may simply be smart business. But it does raise a new question: after racing to secure enough computing power, who will use everything being built?

The list of large customers appears surprisingly short.

Amazon, Microsoft and Alphabet operate the world’s largest cloud-computing platforms. The biggest buyers of their AI compute capacity are OpenAI and Anthropic, which need it to train and run their models.

Elson cites a Barclays model estimating that OpenAI and Anthropic together could account for nearly 75 per cent of AWS’s $25 Billion AI-related revenue in 2026. While Amazon does not provide enough information to confirm that estimate, the possible concentration is still striking.

Much of the revenue being counted as AI demand may be coming from two AI labs buying the computing capacity they need to serve a market they are still building.

The concentration becomes more interesting when you look at who is investing in whom.

Amazon has invested heavily in Anthropic, also one of AWS’s largest AI customers. When Anthropic raised money at a higher valuation, Amazon got to record $50+ billion in income, primarily because its Anthropic holdings had become more valuable. Anthropic can therefore improve Amazon’s financial results in two ways: by paying AWS for computing power and by becoming more valuable as an investment.

I learned this system is unusually self-reinforcing: Amazon helps finance Anthropic, Anthropic buys computing power from AWS, and Anthropic’s rising valuation increases Amazon’s reported profit. 🤔

The list may be short, but these are some of the largest companies in the world, and the business they do with one another is enormous. The unresolved question is whether demand for AI products and services will spread far enough beyond this group to support all the computing capacity now being built.

When slow AI adoption becomes a business problem

Damodaran sees fear of being ‘the company that failed to make the bet’ as one force behind these huge investments.

Once hundreds of billions have been committed, slow adoption becomes a business problem. Investment at this scale create their own pressure to speed it up. The financial case cannot depend on AI remaining a useful tool for some of the people, some of the time.

To justify infrastructure this expensive, AI needs to become far more widely used.

You can see how that pressure travels from earnings calls into corporate plans, budgets and performance expectations. Companies start setting AI targets and looking for productivity gains. Workers hear a familiar warning: learn AI or get left behind, even when the value of a particular use remains uncertain.

Some of the urgency may be a valid response to a useful new technology. But some is also likely to be the demand side of an investment already made.

Through The Generous AI Society, which I am co-creating as a space for women to discuss what AI is worth in their work and lives, I meet people at every stage of adoption. Some are just using AI as a fancier search engine. Others are building their own agents, apps and more.

However, a good many have experimented with AI without finding a use valuable enough to bring into their regular work. They are not necessarily resisting it. They may not yet see where it would help, or they may be doing perfectly well without it.

The latest Statistics Canada survey found that while 19 per cent of Canadian businesses had used AI to produce goods or deliver services during the previous year, a full 40 per cent said it was not relevant to their business.

The survey cannot tell us how many people have experimented casually or why a business has not adopted AI. It can show the distance between current use and the ubiquity implied by the spending.

Those numbers may look very different a year from now. At the moment, they leave a large gap between the demand that exists versus the demand the buildout appears to require.

When Big Tech needs AI to become ubiquitous

Simple adoption numbers often hide who receives the benefit and who carries the risk.

When a company buys an AI system expecting productivity gains, the employee required to use it usually faces higher output expectations and less control over how their work gets done.

Cory Doctorow’s The Reverse Centaur’s Guide to Life After AI offers useful language for thinking about the kind of adoption we want.

A centaur is a person who uses a machine to become more capable. A reverse centaur is a person made to work at the machine’s pace.

A corporate rollout can create the second arrangement and still look successful on a company dashboard. Both arrangements count as AI use. Only one keeps the machine working for the human.

Some of the pressure now reaching workers and users starts much further upstream, with hundreds of billions of dollars committed to infrastructure before the companies spending it can explain the business that will justify it.

Investors may benefit as long as markets continue to reward the bet, and companies may find real productivity gains. Individuals might gain useful capacity, face greater pressure or experience both at once.

Those of us who work for ourselves may feel the pressure to use AI especially acutely. We wear many hats, and AI can give us capacity we would otherwise have to buy or go without. We may also have more agency than many employees over where we use it, when we use it and when we put it aside.

The same agency is not necessarily available to a professional whose employer chooses the system and raises expectations, or to a factory worker whose job is displaced by it.

I still use AI every day and expect to keep doing so. Damodaran’s questions and Doctorow’s book have made me pause over the message that AI will soon be everywhere.

It may well become ubiquitous because people find it useful. But after so much capital has been committed, widespread use has also become part of the business case.

Keep reading …

The Red Thread is for smart, savvy women building, earning, learning, caregiving, and adapting in today’s work world and an on-rushing future.

I write broadly. Including thoughts about AI, women and work, entrepreneurship, and the messy and creative overlap between professional and personal life.

Subscribe for weekly articles to your inbox!

Subscribe now

I started writing in my red notebook on January 5.

I carefully selected a pen to write with as a mechanism to.slow.myself.down. Like many of you, my brain is constantly moving at warp speed and I wanted needed a process that would make me sit with a thought longer, struggle with a sentence structure, play with a turn of phrase that doesn’t involve the words ‘quietly’ or ‘honestly’.

I started strong, journalling every day in January, then almost every day in February, then most days in March, and then …. twice in April. 😳

My journal writing just … petered out. The breadth and speed of the machines seduced me, and I went all in on writing with/adjacent to/about AI. Instead of slowing down, I feel a freneticism to produce faster. I’m using the machines more and more - learning, building, creating in new and exciting ways. And also … I feel the erosion of my ability to just think.

Last week between building agents, saving a million Substack articles, updating a website and launching two events, I wrote a Note about needing to go touch grass: step back from reading everybody else’s work and take a big step back from the machines.

I have failed utterly and abysmally on both counts.

Then I read Anna Jahn’s Op Ed in the Globe and Mail. Jahn runs the Centre for Media, Technology and Democracy at McGill and also has two undergraduate kids who have made opposite bets on AI: one refuses to touch it, the other uses it daily and wonders whether any learning is happening at all.

The common thread? Neither has had structured guidance from their university on how to handle AI. In Jan’s words, they are “improvising their education alone.”

Her article cites historian Niall Ferguson’s proposed solution for Higher Ed: divide university life between the Cloister and the Starship.

• The cloister becomes a “quarantined space” where all devices are excluded. Inside it, students read printed books, debate and discuss, write with pen and paper and submit to oral and written examinations..

• The Starship is everything outside the cloister. There Fergusson assumes all study outside will be conducted with LLMs. A zone where machines are not just permitted, but presumed.

The separation of the cloister and starship is meant to protect deep and independent learning while using AI tools seriously and teaching them properly.

I started reading it over a morning coffee through the lens of my daughter’s future. Then about her present, as she goes into grade 7. And then ultimately, about the application of all this to myself.

I realized: I need a cloister.

So I opened Claude, pulled up my research agent Rosie (props to for Waston, who inspired Rosie), and spent the next few hours talking to the machines about cloisters and AI. (The irony is not lost on me.)

What I wanted to know was: does a version of the cloister exist for adults at work?

I put the question to Rosie, and sent her it looking for to see whose job it is to build or protect the a ‘corporate cloister’.

It doesn’t seem to be anybody’s job 🤷‍♀️. The priority runs hard the other way with most companies now expecting AI use by default.

As Shopify’s CEO mandated last year "reflexive AI usage is now a baseline expectation." AI usage is part of performance reviews, and teams are told not to ask for more headcount until they could show why AI could not do the work.

I sent Rosie off to investigate more deeply.

The agent did not simply search for companies offering AI-free hours. It realized that I had asked the wrong version of the question.

Rosie found four different ways spaces without AI or automation exist already: laws requiring human decisions, union contracts protecting the right to refuse AI, safety regulations requiring people to maintain manual skills, and security rules excluding connected technology altogether.

She distinguished binding rules from recommendations. She identified who each protection was actually designed to protect. She told me which arguments were already crowded, which findings appeared less explored and which claims she could not verify well enough to use.

I was blown away. And I was also looking at way more research content and than I could reasonably absorb, never mind turn into good writing.

That was when the cloister stopped feeling like an interesting educational metaphor. The starship is now sitting on my desk.

The agent had done exactly what I built it to do. It had ranged further than I could have ranged alone. It had been more systematic, more tireless and, in some respects, more disciplined.

Now came the part it could not do for me.

Which of these ideas actually deserved my attention? Which one connected to something I had observed or cared about? Where, in all that horsepower, was the judgment? That question took me back to my daughter.

Our kids will grow up able to ask LLM’s for a clear explanation of almost anything. They can ask follow-up questions without worrying about slowing down the class. Or explore a subject beyond what appears in a textbook or lesson plan.

They are also going to get the synthesis before they’ve experiences the learning and maybe the confusion.

Confusion is uncomfortable, but it is not wasted time. It’s where you realize you didn’t get the question, or maybe that your sources conflict or observe a tidy answer does not quite fit.

How does a student learn to find the question when the synthesis arrives first?

The same problem follows us into work.

A junior employee can now use AI to produce an analysis, recommendation or presentation that looks far more senior than the thinking behind it.

The immediate output may be better. The client may get a cleaner document. The organization may save time.

But junior work was never valuable only because someone needed to prepare the first draft. It was part of how people became less junior.

You do the research. You decide what belongs. You miss something obvious. Someone with more experience asks why you trusted one source or dismissed another or made a recommendation that wasn’t properly substantiated.

If AI performs more of that awkward first work, how does the junior employee develop judgment rather than simply produce senior-looking documents?

Experienced workers face another version of the same question.

We already have a lot of the knowledge, context, pattern recognition and memories of decisions that went well or badly. AI can make all of that more powerful. It can find contrary evidence, challenge an assumption and connect our question to something we would never have found alone.

The problem is partly cognitive overload: the sheer volume of information and possibility coming back at us. It is also partly cognitive offloading: the parts of the mental work we stop doing because the machine can do them faster.

Cognitive offloading itself not automatically a problem. The question is whether we’re offloading a task we no longer need to perform or a capacity we still want to retain.

A lot of the discussions about responsible AI rest on the idea that a human must remain “in the loop.”

But. That means very little if she has not developed enough knowledge to recognize when the machine is wrong, practiced forming her own view before seeing the machine’s answer, or retained enough confidence to disagree with twenty pages of ‘evidence’.

And even those capacities mean little if she has no real authority to act on the disagreement.

Just because a process has ‘proof of human’ does not mean it contains independent, hard-earned human judgment.

Subscribe now

One finding from Rosie’s research has stayed with me.

Commercial aviation has spent decades dealing with automation weakening the manual skills pilots may need when automation fails. Aviation is not a tidy analogy for AI + knowledge work, and I’m aware of this.

The useful distinction is between structured practice and arbitrary abstinence.

Aviation identifies specific manual capabilities that need to survive and requires pilots to practise them. It does not simply prescribe a set number of hours away from automation and hope competence follows.

That made me realize I had been asking what an ‘adult cloister’ should look like before being very clear on what it needs to protect.

Ferguson’s recommendations travel pretty well.

• Printed books, not summaries.

• Problems argued out loud.

• Writing by hand.

• And examination by a person.

I can do the first three myself.

The fourth is harder.

A university cloister has walls, rules and an institution behind it. Mine has me, sitting at the same desk as the starship, with nobody stopping me from opening another tab.

I can make a first attempt before opening an LLM. I can read original sources rather than relying entirely on summaries. I can write in my red notebook. I can protect time to think before asking for more information.

But personal discipline is not the same as a protected structure.

The examination by another person may be the most important part: someone who will question my premise, argue with the answer and notice when I have mistaken polish for thought. A reading group. A trusted colleague. A room where the point is not to generate more, but to decide what is worth keeping.

I do not yet know exactly what my cloister will look like.

I do know why I need one.

Not because I distrust the starship. I am astonished by what it can do.

I need a cloister because I don’t want to lose my independent thinking and hard-earned judgment from a life filled with experiences. I do not want the machine to get the first word every time.

Keep reading …

The Red Thread is for smart, savvy women building, earning, learning, caregiving, and adapting in today’s work world and an on-rushing future.

I write broadly. Including thoughts about AI, women and work, entrepreneurship, and the messy and creative overlap between professional and personal life.

Subscribe for weekly articles to your inbox!

Subscribe now

Today, I deleted roughly a thousand photos from Instagram. Not archived. Deleted.I did it manually, because Instagram hasn’t exactly designed the exit to be easy.

As I did so, I watched a filmstrip of my life play backwards.

My daughter jumping from a dock, aged seven. My elderly mother standing at the edge of the ocean with her old dog and her little granddaughter. The launch of a business with my best friend. A COVID-era photograph involving hair foils and face masks. Our wedding and our fifteenth anniversary.

Moves between towns and schools. Family and friends. Travel, birthday cakes, beaches, dinners, ordinary afternoons.

It is a life filled with so many moments of joy. Looking at it made me a little nostalgic, but mostly grateful.

I didn’t post the photos thoughtlessly originally. I shared them because they were honest reflections of moments I was happy to be living. Sharing was part of the pleasure. (I am a Gemini, after all.)

I do not regret sharing them.

I no longer agree to the terms under which those files can be used.

Subscribe now

Today’s delete-a-palooza started with an article by from AI Family Network about the risks of “sharenting” in 2026.

Sharenting is that compulsive habit many of us parents have to post photographs, videos and fun details about about our families especially on FB and IG. An everyday way to share the special moments — and all the mundane ones too.

Sharenting always carried risks. But something fundamental shifted between 2016 and 2026. And boy does Manisha draw the line clearly.

2016:

2026:

Sobering.

Her article included two video shorts. The first is about 40 seconds. The second is two minutes and 48 seconds.

Together, they ask three and a half minutes of you.

Please watch both.

The first, Pause Before You Post, left me in a cold sweat. Just another perfectly ordinary family posting about perfectly ordinary stuff giving strangers more information than they should ever have.

The second film, A Message from Ella, brought tears to my eyes. Yes, actually.

It’s under 3 minutes. It’s compelling. And please, watch it all the way through.

Using just a single childhood photograph, the film creates an AI-generated, adult version of nine-year-old Ella to show her parents what her image could become once it leaves their hands.

I am not going to tell you much more than that. The film needs to do its own work.

I closed the video. I opened Instagram.

Then I started deleting.

By the time I finished, I permanently deleted nearly a thousand photos and more than 90 videos, shared over nearly two decades.

When I began sharing pieces of my life online, my account was private. I understood that I was putting photographs onto a platform owned by somebody else. I understood, at least in the vague way most of us did, that there were terms and conditions attached.

I did not think I was giving organizations permission to turn those photographs into material from which new versions of the people I love could be made.

That possibility was not part of my understanding when I began posting.

The AI piece changes the meaning of the act.

A photograph is no longer only an image sitting on a feed, waiting to be looked at. It is an input: something to analyze, categorize, copy, train with, reference, alter or use to generate something new.

The rules changed after many of us had already spent years supplying the material. The timing is ironic. Less than two weeks ago, I changed my Instagram account from private to public.

I did it because I am trying to extend the reach of my writing and grow a couple of bodies of work. Visibility is treated as part of the job now. Show your work. Show your thinking. Show enough of yourself that people can know and trust you.

On July 7, Meta launched a feature inside Instagram that let anyone @-mention a public account and pull that account’s photos into an AI-generated image — opted in by default, whether or not the account holder had asked for it. Three days later, after a major backlash over consent, Meta pulled it. “We’ve heard the feedback that this feature missed the mark,” the company said, “so it’s no longer available.”

This reversal does not restore my comfort.

Someone still decided that publicly visible could reasonably mean available for synthetic reuse, unless users found the right setting and opted out.

I never agreed to that. Most of you haven’t either.

Look, I know that deleting my photos from Instagram does not guarantee they’re gone from every database, archive or system that may already have accessed them. I am not confusing deletion with erasure.

The arrangement may well be rigged. The terms may allow more than most of us ever understood. The files may already have travelled further than I can see. Still, the fact that I cannot retrieve every possible copy does not require me to keep offering the originals.

The photographs have not disappeared from my life. They are on my phone and my hard drive. I own my copies. The memories belong to my family. Instagram was never the memory. It was one place I displayed it.

I am not anti-AI. I use it daily, in my work and my writing. I’m building because I want more women involved in the conversation about how AI enters our work and our lives.

That does not require unquestioning loyalty to every use of it.

Enthusiasm is not blanket consent.

I showed my daughter my Instagram account: the empty Stories, the near-empty grid and the handful of recent posts about my writing and work. We talked about why I had deleted the photographs.

I posted the majority of them before she could have any meaningful say in the matter. It felt important that she understood why I had now made a different decision for both of us.

I do feel some regret when I look at what remains. The account now shows such a small sliver of me. The whole person has been replaced by the professional one: essays, projects, events and work. I have always liked sharing the broader, messier picture — the dogs and the dinners alongside whatever I am building.

For now, that is the trade I am willing to make.

Deleting your own treasure trove may not be right for you. You might archive instead. You might tighten your audience, strip out names, faces, school uniforms, and locations. You might decide the platform’s safeguards are enough.

Manisha offers a useful 30-second PHOTO check in her article: pause, consider permission, share only what is needed, tighten the audience and consider keeping it offline or one-to-one.

This is not only for parents and guardians.

Ella’s story made the danger brutally clear to me because I have a daughter. What are you willing to share when an image may not simply be seen by people, but used by machines?

My own rules have changed. I will continue to use social media for my work. I may occasionally share something personal, but not with identifiable faces or names. Family photographs will go to the people they are intended for through texts, emails and smaller private circles.

Tonight, my Instagram account looks emptier. I feel a little saddened by that. I also feel freer.

I shared those photographs with joy. I deleted them with clarity.

Keep reading …

The Red Thread is for smart, savvy women building, earning, learning, caregiving, and adapting in today’s work world and an on-rushing future.

I write broadly. Including thoughts about AI, women and work, entrepreneurship, and the messy and creative overlap between professional and personal life.

Subscribe for weekly articles to your inbox!

Subscribe now

I had coffee this week with someone who knew me from my Whistler years in ‘before times’. We didn’t bother to swap business cards or trade job titles.

Back then, those details would have been simple. I ran regional sales, she led customer service training. That was true for both of us once. It was also several careers ago.

After we parted ways, I counted. Sales for Intrawest. Managed an 80-person sales team in a Glengarry Glen Ross-type boiler room, then quit without a backward glance. Helped run a federal campaign — three months, 22,000 votes, and the certainty I would never (ever) do that again. Opened hotels in Montreal and Whistler. Led MarCom for a national charity. Raised money and launched an e-commerce startup with my best friend. And most recently, 10 years co-running a brand strategy agency for startups and venture firms. Plus sitting on three boards few ask about, and years of mentoring I rarely speak of.

Try writing one job title for that.

The world still wants a clean answer to “what do you do?”.

After a few decades of work, the answers are less like a ladder and start looking more like a sprawl. And one title can’t cover a sprawl.

So we give people the version they can place fastest, and leave the rest of the details out. Not so much false modesty as triage.

This is where “multi-hyphenate” stops clarifying the work and starts muddying the person.

Founder-VP-advisor-author-director-speaker-planner-mother-sister-daughter- aaaaah.

The hyphens are what we reach for when an intentionally varied career keeps reading as a scattered one, and we want to prove it was deliberate all along.

But punctuation can only do so much.

The problem is (clearly) not how many things we’ve done. It is that individual labels are not built to hold this much accumulated experience.

The real task is making the through-line clear enough that people know what to come to us for, where our value sits, and how to talk about us when we are not in the room.

That last part matters, a lot.

Why? Because the cost of outdated labels show up where we can’t see them.

You only get referred for a fraction of the work what you actually do. The board seat conversation doesn’t reach you, because your one word said “consultant.” The advisory work routes to someone louder. The introduction that might have changed your quarter goes around you, because a single old label told people which room to keep you in.

And even the full list — every role, title, company, board seat, project, and unlikely detour stacked end to end — is only our working identity. It is never the whole of us. It doesn’t hold the family we go home to or have chosen, the parent we’re helping now, or the reasons we chose this particular life in the first place. No org chart has a box for this.

I don’t mean titles are useless. A clean label does heavy lifting — it tells people which room to put us in, what to assume, who to trust with what. And stepping away from that is hard. Whether you’ve had one steady career track, or you’ve made peace with a string of hyphens, the pull is to keep the shorthand and hope it still fits.

Naming and claiming our own thread asks more of us than a title ever has.

Which leaves one piece of real work.

Find the through-line that lets a title list become a story instead of an inventory.

Mine took me over two decades to see. Across all the rooms, companies, campaigns, boards, brands, and strange career turns, I have been doing the same work: untangling complexity, pulling clarity forward, and making it useful and legible.

An old title doesn't just sit on a Linkedin profile. It's out there working — sending some things your way, steering others off, shaping what people picture you doing next. Worth considering what you want it to say about you - and whatever is next.

Keep reading …

The Red Thread is for smart, savvy women building, earning, learning, caregiving, and adapting in today’s work world and an on-rushing future.

I write broadly. Including thoughts about AI, women and work, entrepreneurship, and the messy and creative overlap between professional and personal life.

Subscribe for weekly articles to your inbox!

Subscribe now

Recently, I’ve been in some of the most honest conversations about artificial intelligence I’ve had yet. No panels. No webinars. No Zoom.

Actual rooms. Actual women.

My conviction after these conversations is stronger than ever. Women are not hesitating their fervent embrace of AI because they are less confident. Or hesitant. Or fearful. Or need encouragement, access, or a friendlier on-ramp with fewer frightening acronyms and more pastel.

Many are pausing because they see the consequences of rapid and careless AI rollout from where they stand: disrupting the workplace, dispensing false information, offering fake intimacy, augmenting social isolation, reinforcing stereotypes, and well … the list goes on.

The catch-up narrative is out of date

We’ve already had the “women can’t be left behind” moment.

In April, Reese Witherspoon told her millions of followers that the AI revolution had begun and encouraged women to learn the tools so they would not fall behind. The post sparked a quick backlash in part because women are tired of being told the answer to a structural problem is one more personal upskilling project.

A few weeks later, Mel Robbins entered similar territory through a Microsoft Copilot partnership. This time, AI adoption appears as the new empowerment pitch, wrapped in urgency, served with prompts, and aimed at women who are apparently one productivity tool away from professional safety.

Yes, I believe women should use AI. We should understand it. We should test it, question it, and get much more fluent in what these tools can and cannot do.

But the catch-up framing is now behind the facts. Pew reported this month that the gender gap in chatbot use has essentially closed: 50% of men and 47% of women say they use AI chatbots, and ChatGPT use is identical at 44%.

But the societal narrative persists in asking whether women are adopting fast enough. The better question is where women are in those rooms where AI is being procured, permitted, taught, governed, normalized, and absorbed into everyday life.

The women I’ve spoken with are not short on appetite, curiosity, or creativity as regards AI usage. Some are power users. Some are toe-dippers. The conversations were not shaped by technical fluency so much as proximity to the consequences of AI.

Subscribe now

Too fast is not the same as afraid

Pew found that women are more likely than men to say AI is advancing too quickly: 68%, compared with 58%.

The lazy reading of this is ‘anxiety’.

Read as fear, that number says women are nervous and need encouragement. Reflected more accurately, I think it says women are questioning pervasiveness, ethics, trust and value.

Around the table, I heard hesitation grounded in experience, responsibility, and a very reasonable suspicion of tools that arrive wrapped in convenience and leave us to clean up society.

The future is onrushing. AI is being dropped into workplaces before managers understand what good use looks like, into schools before parents have caught up to what their kids are doing with it, into government policy and surveillance, into hiring and health and search and daily decisions.

It is being sold as convenience, inevitability, competitive advantage, personal assistant, efficiency engine, and intellectual partner.

A woman looking at that and saying “hold on” is not the problem. Her pause seeks information.

What the pause is actually doing

The women hitting pause are not standing outside the AI conversation. They are inside it, checking its work — catching the wrong answers and seeing the bias underneath them.

The UN Women Report on AI reports that a study of 133 AI systems found 44% showed gender bias, with more than a quarter showing both gender and racial bias. Its digital technologies lead, Jayathma Wickramanayake, described models that pull bias from decades of text written about people in a world where women were filed under ‘home and family’ and men under ‘business and career’. The bias in AI, trained on biased resources, now runs at scale often unquestioned by users.

Catching the bias on your own screen, one prompt at a time, is vigilance — useful, unpaid, and endless. It changes nothing upstream. The reason to be in the room is not to clean up after the tools more efficiently. It is to decide what the tools are for before they’re embedded and inevitable.

AI platforms and policy are being built without the people closest to the consequences. Only 24 countries reference gender in their national AI strategies, and just 18 include substantive gender-responsive measures. Of the global AI workforce, women make up only 30% .

The blueprint is being drawn now, and yes women must be at the table.

A seat at which table?

“Women need a seat at the table” has been true for so long it has worn into wallpaper. It sounds good but - please, which table? Here are four where the terms are being set right now.

• Procurement. Someone is deciding whether Copilot or another system gets installed across everyone’s tools, what it can access, who checks its output, and whether not using it quietly becomes a performance problem. Being in the room means asking those questions before the rollout, not after.

• Education. Teachers, administrators, and parents are deciding what counts as cheating, what counts as support, and how we teach kids to question an answer before they outsource the uncertainty.

• Government. Be in the room where data centres get approved — water, power, noise, community benefit, who owns what. ‘At the table’ means the permit is a decision, not a foregone conclusion with a public meeting attached for show.

• Policy. Job exposure, privacy, bias, deepfakes, child safety — treated as lived realities, not abstract innovation concerns.

  If this is resonating, consider sharing with a friend or colleague!

  Share

Some friction is contact

An elder at one of the tables named something the data does not capture. Her worry is civic, not technical. She talked about self-checkouts, banking machines, online ordering, delivery — the way you can move through a whole day now without speaking to one person. No teller. No clerk who knows your name.

AI is not arriving into a socially rich world. It is arriving after years of ordinary contact being stripped out for convenience, and now we are layering synthetic conversation on top of the silence we already built.

That does not mean we reject it.

Nobody in those rooms was purely anti-AI. Even the most skeptical comments were not “burn it down.” They were about bad information, fake intimacy, synthetic selfhood, job loss, surveillance, isolation, and the way automation has already removed a lot of ordinary human contact.

Some friction is waste.

A form that takes three hours and tells you nothing can now go. A grant application that punishes a tiny nonprofit for not having a writing department can be made less onerous. A medical appointment you arrive at unprepared because you didn’t have the language can be supported.

Use the tool.

But some friction is contact.

The phone call. The exchange at the counter. The in-person room where women compare notes and find they were not the only ones who paused. That is not inefficiency. It is how we stay human enough to know what the tools are for.

Before the terms get set

The pause is not a failure to adopt. It is a demand to look at the conditions of adoption before the rollout becomes inevitable and the cleanup becomes invisible.

So use the tools. Thoughtfully and practically, to make invisible work less crushing where you can to prepare, draft, compare, translate, advocate. Use them in the thin slice of time that was supposedly going to become your creative practice.

But use is not the same as blind consent - and a pause is not the problem.

The pause is not a failure to move. It is the moment before the terms get set — the one where we sit at the table while the blueprint is still wet.

Keep reading …

The Red Thread is for smart, savvy women building, earning, learning, caregiving, and adapting in today’s work world and an on-rushing future.

I write broadly. Including thoughts about AI, women and work, entrepreneurship, and the messy and creative overlap between professional and personal life.

Subscribe for weekly articles to your inbox!

I’ve been building The Red Thread for about 3 months now. Not long, but enough to know I need it to hold for the future: Articles, Notes, Field notes, Videos, future offers, and the occasional personal confession. Ha.

When I looked at the visual identity I created way back in May 2026 ;) it wasn’t quite meeting those needs. It was (IMHO) elegant, personal, considered. But it was starting to trap the work inside the just my image.

So I rebuilt it. Not with a designer — with AI, my own background in brand strategy, and a priceless resources from who built the blueprint on building a Substack visual identity system.

I’m pretty happy with the results but am still in full tweak, iterate, and change mode. That said - I’m the proud owner of a full brand bible: palette, type, illustration modes, annotation marks, avatar direction, application templates.

Here’s what changed, and why.

What was working — and what wasn’t

The original identity had taste. Ink line work, warm paper, author portrait. Human, considered, not generic sludge.

The problem wasn’t that it was wrong. It was becoming too fixed. The avatar was doing too much work. The scenes felt slightly nostalgic — writer, desk, window, notebook. Lovely, but passive.

Heavily indexed on “woman reflecting” when the publication is increasingly about women deciding, building, judging, earning, using tools, comparing notes.

The identity was saying: Katrina is thinking about this. It needed to say: This is a place where the work, the room, the tools, and the reader’s own life can be felt.

The main move

Separate me from the primary visual identity of the publication. I’m still present. The portrait still has a role. But it’s no longer the main anchor.

The Visual ID has moved away from “look at the author” to “look at the work, the room, and the evidence.” More ways in for readers. More surface area for the publication — including room for a commercial thread without the whole thing suddenly feeling salesy. And without turning everything into a portrait of me thinking beautifully near a mug.

Subscribe now

What changed visually

The palette is more modern.

• Ink carries structure.

• Thread Red marks the thought — a circle, an underline, a “pay attention here.”

• Hot Pink appears once, as a small pulse.

• Chalk and Mist offer a warm background but not so ‘bookstore chic’.

The illustration direction is getting sharper: contemporary, bright, airy, selective. Fine linework, open space, current technology, experience and competence. No heavy graphite, no sepia, no vintage desk, no symbolic overload.

The type pairing does real work.

• Fraunces gives the publication voice and editorial confidence.

• Inter gives it clarity and usefulness.

Together I’m hoping they give the brand space to move between a personal essay and a practical asset without looking like it changed personality.

What AI actually did

AI did not make the brand. It did not bring the taste. It did not know when something felt too sepia, too staged, too much like a woman gently imprisoned inside an artisanal notebook. That was judgment.

But AI (and the generosity of Substack) gave me capacity and know-how to work like a tiny creative studio — testing avatar direction, essay covers, Note cards, object-led images, room scenes, product surfaces, fast enough that the testing was actually useful.

With correction. With irritation. With taste. With many rounds of “absolutely not.”

No designer. Just one very opinionated client.

Oh. And if you’re curious or inspired to try this for yourself... don’t take my word for it, bust a move over to 😊 and check out her Branding with AI tools.

Subscribe to The Red Thread

Subscribe now

Essays and field notes on work, AI, growth, care, money, and the judgment it takes to build under constraint.

I’m excited about building a portfolio career.

I’m also old enough to be suspicious of any version of work that promises to be easy.

This is one of the strange gifts of midlife. You can still be drawn to possibility, but you stop believing the version of it that comes in a tidy Substack post.

You hear words like freedom, flexibility, optionality, and independence, and some part of you thinks: gorgeous. Another part asks who is organizing summer camps, when property taxes are due, and whether anyone in this house has thought about dinner.

After years co-running a small agency, I have been building what comes next. Not in the reinvention-montage way. No linen shirt on a beach. No sudden clarity in a beautiful notebook. More like: prompts, conversations, walks, small waves of panic, better questions, and the slow realization that I may not be looking for another single container.

Turns out I’m building a portfolio. I didn’t know that when I started out. I’m giving myself permission to call it that now.

Advisory engagements. A substack practice. Some board work. Creating a room where women can talk about AI, work, judgment, care, ambition, and money without pretending any of those things politely stay in their own lanes.

It’s all so interesting. It is also … a lot — and I’m the one doing it to myself.

What the story leaves out

There is a confident new story going around about women and work, and most of it is true. Many women are opting out of corporate at scale and on purpose.

In a recent Gusto report, women-owned startups accounted for 49% of new businesses, up from 29% in 2019. At the same time, ‘fractional’ work has gone from niche when I first started using it 10 years ago, to mainstream career architecture — by late 2024, LinkedIn profiles using the word “fractional” had climbed past 140,000, up from around 2,000 two years earlier - though Axios, which ran the count, noted plenty of those are aspirational.

The framing around all of this is upbeat: not an exodus, a reallocation. Not a retreat, a recalculation.

I believe it. I’m part of it. It’s also the kind of narrative that stops working on you the moment you are the one ‘doing’ it.

Because the story ends at the leap and sometimes skips the landing. It’s strong on the decision to leave and quiet on the years afterward, when the romance burns off and you are left holding trying to build the actual thing. I’m not here to talk anyone out of it. The recalculation is sound and the variables are real — but at this age, a few of them move differently than the brochure suggests.

Three of them matter more than the rest. Income. Identity. Infrastructure.

Income

Not in the abstract. The actual math of a household — a mortgage, school costs, taxes, insurance, the egregiously expensive vet bill that suddenly appeared.

I’m not new to this part. I’ve worked for myself for the better part of fifteen years. I know the variability — revenue arriving in clumps, the late invoice, the month that tilts. That is the cost of independence everyone warns you about, and frankly it is the one I find least frightening (most days), because I’ve already made my peace with it.

The harder question is the one nobody hands you a number for: how do you get paid what your judgment is actually worth, when there is no firm setting the rate and no title doing the negotiating for you?

Inside an organization, your price is largely decided above your head. Out here, you set it, defend it, and live with it. Most women I know underprice on the way out — not because they don’t know their value, but because the institution used to translate that value into a salary band, and now the translation is your job.

Pricing stops being an HR conversation and becomes a positioning one. That’s a different muscle, and it’s worth building deliberately rather than discovering by accident three underpriced contracts in.

Identity

This is the one that ambushes people, and it ambushes senior people hardest.

You think you are leaving a job structure. You are also leaving a version of yourself that other people understood.

Small agency leader. SVP at a place with a recognizable logo. The one who knows what to do in the meeting. A title does an enormous amount of translation work — it tells a room who you are before you open your mouth. So does a firm. So does an institutional email address. Strip those away and you have to become legible all over again, in your own name, to people who no longer have a shortcut for placing you.

This is not vanity, and it is not fragility. The self that organized your working life — the scope, the authority, the standing that told a room who you were — doesn’t taper off gently when you leave. It goes at once. And the more of your sense of self was built on the role, the harder you feel its absence.

The cruel twist is that the women this hits hardest are often the most accomplished. The senior operator who spent twenty years becoming synonymous with her institution has the most translation to redo. “What do you do?” is a genuinely harder question when the answer is no longer a logo. And the rebuild is slow in a way that money is not — you can land a contract in a month; becoming legible as yourself, in your own name rather than your last employer’s, takes a year or two of saying the same true thing in public until it sticks.

Worth knowing before you go, so you don’t mistake the wobble for a mistake. It isn’t a sign you chose wrong. It is the predictable cost of having been good at the last thing.

Infrastructure

Here is where the romance is thinnest and the real work lives.

The story sells the freedom and skips the scaffolding. But a portfolio without infrastructure isn’t freedom — it’s just many jobs sharing one nervous system. The infrastructure is the unglamorous machinery that keeps the whole thing standing: how you get paid and chase what’s owed, how you stay visible enough that the right work finds you, how you decide what you’re for and — harder — what you’re not.

This is also where the real risk of a portfolio career hides, and it isn’t instability. It’s diffusion.

Everything can be interesting. Everything can be adjacent. Everything can be part of the larger body of work if you squint and use the right font. Advisory connects to writing. Writing connects to community. Community connects to events. Events connect to sponsorship. Sponsorship connects to products. Products connect to workshops. Workshops connect back to advisory.

Whew. Also: please stand up, then eat lunch.

A portfolio becomes a more elegant form of overextension if you aren’t ruthless about what each piece is for. This is especially true for women practiced at being competent across domains — we can hold the room, fix the deck, remember the birthday, call the pharmacy, and make it all look calmer than it is. Competence is a trap precisely because it makes too many configurations possible.

AI runs through all of this. Not as a tool I reach for at the end, but as where I start. Most tasks, most thinking, most half-formed ideas begin with it now. That’s the capacity it buys, and I’m not interested in pretending otherwise. The discipline is the infrastructure question, because the risk isn’t using it — it’s letting it rapidly turn every thread into more output when the whole point is restraint.

Used without thinking, AI will lightspeed your production of well, more of everything, which for a portfolio career is precisely the wrong kind of help. Used with judgment, it absorbs the operational weight of running several threads at once so the person holding them can stay a person.

Before you add anything new to the plate

Freedom requires design. Not a vibe. Not just a personal-brand moodboard. An architecture. These are the questions I’d ask before adding anything — the ones I wish someone had handed me earlier:

• What does each piece actually earn? Not just money — reputation, access, learning, joy. Name it. A thread that earns none of these is a hobby wearing a business costume.

• What can only you do? It’s the hardest one here, and if the answer doesn’t come fast, that’s the work, not a failure of nerve. Whatever it is, that’s what you protect. Everything else is potentially delegable, automatable, or droppable, and you should know which.

• What looks like opportunity but is really someone outsourcing their ambiguity into your calendar? This is the most expensive category, because it arrives flattering.

• What’s the line that makes you say no? Decide it cold, in advance, before an interesting room talks you into a configuration your week can’t hold.

None of these are glamorous. They are better than glamorous. They are useful — and they’re the difference between a portfolio that holds a life and one that quietly bills it for parts.

What I’m building

My life doesn’t wait politely outside the (home) office door. There’s a tween who needs me. An elderly parent whose care arrives as admin, grief, and small heartbreaks disguised as phone calls. A mid-life body making its own editorial comments. A bank account with thoughts.

And still, there is ambition — though I’m not sure that’s even the right word anymore. I don’t want a smaller life. I want my version of it: enough income to be free, a community to create within, purpose in the work I choose, and room for a whole life around both. That’s not a lesser ambition. It’s a more aligned one.

A portfolio career is an exciting opportunity. But only if it’s built with the variables in plain sight: income that respects what your judgment is worth, identity rebuilt deliberately in your own name, infrastructure strong enough that the threads connect without strangling the person holding them.

That’s the work for me. Not a linear career. One built to hold the work, the purpose, tween, the aging parent … and lunch.

Keep Reading

Some other ‘threads’ you may enjoy:

A Portfolio of Purpose [here]
The follow-your-passion advice was always a bit of a lie, and AI hasn’t changed that. What AI has done is make the question sharper. Tasks are moving. Purpose is what stays.

Fluent Is Not the Same as Having Something to Say [here]
Anyone can generate polished language now, so polish has stopped telling us very much. What’s left is the nuance, the pleasure, the learnings that AI simply can’t reproduce.

Too Early to Be Obvious [here]
On why the right room can change what you are able to see, say, and build.

A quick note from me. Hit the play button.

Peer mentorship is chronically under-appreciated — maybe because it sounds soft, maybe because we file it somewhere near networking or professional development, and leave it there.

It deserves more serious attention than that. Especially now, when we are surrounded by advice. Searchable advice. Generated advice. Expert advice. Algorithmic advice. Polished, instant, fluent advice available at any hour on any question. And yet a 2025 KPMG global study surveying over 48,000 people across 47 countries found that more than half of respondents worldwide were unwilling to fully trust AI. More striking still: the more people use these tools, the less they trust them. Usage is rising while trust is falling. A tool can help you map your options. It cannot notice when none of the options are the real issue.

Over the last couple months, I have facilitated a structured, peer mentoring cohort of eight women business owners. We met twice a month. One person was in the hot seat each session. The businesses were different. The models were different. The stages were different. The questions on the table were specific, practical, and consequential. What the room offered was not advice in the generic sense. It was a protected space to pressure-test the decisions sitting underneath the business problem each person thought she had brought in.

Sometimes the stated challenge was the challenge. Often it was not. A marketing tactic turned into a positioning conversation. A staffing issue opened into a more honest discussion about the kind of growth someone actually wanted. A revenue question led to something less tidy but more important: what kind of company am I trying to build, and what am I no longer willing to trade to build it?

That is what peer mentorship can do when it is well held. It does not simply make people feel supported. It helps people think more clearly.

Subscribe now

What peer mentorship does differently

Mentorship is usually understood as a developmental relationship where someone with more experience shares knowledge, perspective, and guidance with someone who has less. At its best, it can be generous and useful. It gives people access to judgment they have not yet had the chance to earn for themselves.

But peer mentorship works differently. It is not built around one person having the answers. It is built around a group of people with enough lived business experience to be useful to one another, and enough difference between them to see what the person inside the problem may be missing. The value does not come from hierarchy. It comes from proximity, contrast, and candour.

This distinction matters because a lot of business owners do not need someone to explain the basics. They need a room where the actual decision can be examined without performance — people who can listen past the presenting issue, ask better questions, challenge the first version of the story, and help separate a real constraint from an old assumption.

Herminia Ibarra’s research on professional networks is useful here. She writes that women’s work and social networks often overlap less than men’s, and that networks built mostly around people we naturally “click” with can become too homogeneous to give us the range of input we need for good decisions.

A well-structured peer room can address part of that problem: not by adding more generic support, but by creating deliberate exposure to different industries, business models, and ways of reading the same issue.

What peer mentorship is not: networking, therapy, a confidence workshop, or an accountability group. It can draw from all of those in its best moments. But the structure that makes it valuable is that real business problems go on the table and the people around the table take them seriously enough to push back.

The part that surprises people every time

Every person in the cohort came in expecting a specific ratio of value: one session of real support — her own — and seven sessions of giving. That framing makes sense. You sign up thinking you will get one useful day and spend the rest contributing what you can to other people’s problems.

By the end, almost everyone said some version of the same thing: they had not expected to learn something in every session. Not just from their own. From all of them.

One participant put it plainly:

I was SO surprised I could learn something applicable to my business from every single person in the room even though our businesses are wildly different. It made me realized there are fundamentals which align across them all.

This is one of peer mentorship’s defining features, and it is easy to miss if you have not experienced it. The learning is not concentrated in the session where you are the subject. It is distributed. Often the sharpest learning happens in someone else’s session, in the moment you open your mouth about a problem you are not living and hear yourself say something you needed to hear about the one you are.

In your own session, you are explaining yourself. You are giving context, making the situation understandable to people who do not have all the background. You are offering a version of the story shaped by how you understand the problem, how you want the problem understood, and what you may still be trying not to know. That is not dishonesty. It is the cognitive overhead of being inside your own problem.

In someone else’s session, you are not carrying the same weight. You are just thinking. You can see the pattern more clearly because you are not attached to a particular outcome. You can ask the question that moves the room. You can name the trade-off they have been circling. You can say the thing that cuts through..

This is also why cross-industry peer mentorship can be so useful. The instinct is often to assume that you need people from your own sector to understand your problem. Sometimes you do. But often the person outside your industry is the one who can hear the actual issue because they are not fluent in your preferred explanation of it. The details differ. The underlying patterns often do not.

Share

The view from outside your own business

Running a business means being inside it. Managing cash flow and client relationships and staff dynamics and invoices and decisions that keep getting deferred because there is no good moment to make them. Your field of vision narrows — not because you lack judgment, but because you are spending so much of it on the present tense.

Peer community offers a different view. Not the grand view of someone more senior who has seen it all. The more ordinary and sometimes more useful view of people standing outside your building, looking at the shape of it. That is not the same view you have from inside, wondering whether the walls are straight.

Sometimes that outside view helps someone name a decision that has already been made but has not yet been said out loud. Sometimes it reframes a cost that felt like a liability into something closer to an investment. Sometimes it reveals that the issue is not the problem the person brought into the room, but the assumptions sitting underneath it.

Often it does something smaller and more powerful: it gives a person permission to trust what she has been seeing but has not yet acted on. A good peer room does not tell you what to do. It helps you hear the decision more clearly.

On Imposter Syndrome and a Reality Check

Doubt was in the room. Sometimes the shorthand for this is imposter syndrome — a term so familiar by now it has almost lost its edges, but which speaks to something very real. It shows up across professional women’s careers at every stage and level, from founders of early-stage businesses to CEOs of large organizations. The research on it spans decades. The experience of it is still running well ahead of most of the advice about it.

In this room, the doubt showed up in familiar ways: over-explaining, qualifying, minimizing, circling a decision without naming it, describing hard-won judgment as luck or timing, hesitating to say what the business had already made clear.

It’s easy to see other businesses and think they have it all under control and feel like you jut don’t - or don’t even belong at the table.

That kind of doubt is not always a private confidence problem. Sometimes it is a context problem. Ruchika Tulshyan and Jodi-Ann Burey made this point sharply in HBR’s ‘Stop Telling Women They Have Impostor Syndrome’ when they challenged the way imposter syndrome is often used to locate the problem in women rather than in the conditions around them. The phrase can make it sound as though the work is simply to help women believe in themselves more, when the more useful question may be: what kind of environment produces, confirms, or interrupts that doubt?

That question matters in a peer mentorship room. If you want to address doubt, you do not necessarily need affirmation. You need people who can see the business from outside the daily pressure of running it. You need a setting where someone can say, plainly and without theatre, “That is not how the market sees you,” or “You keep talking about this as an operational problem, but it sounds like a values problem,” or “You are treating this as a failure when it may be a sign that the business has outgrown its current shape.”

That is not cheerleading. It is realignment.

What a room needs to work

What made the candour possible was confidentiality. Which means members could put the real problem on the table: the decision they were not sure about, the number that did not add up, the version of growth they were supposed to want but maybe did not, the direction they kept circling back to and then avoiding.

The value is not that it enables dramatic revelations. It is that it lowers the cost of honesty enough that people can say the ordinary true thing — the thing they already know but have not said out loud — and have it met with practical seriousness rather than the careful performance of support.

One participant landed on something she had been working toward for months, about what kind of company she actually wanted to build and what growth would have to mean for her to want it. She put it plainly: I realize I don’t actually want to grow my company - it’s not aligned to my values. That is not a line you usually arrive at alone, under pressure, between invoices.

Structure. Warmth. Someone to hold both. One participant described the facilitation as a “firm hug.”

That is the synthesis work. Not directing the room toward a predetermined answer. Not letting support become so soft that nothing useful gets said. The job is to find the clearest possible version of what the room has already worked out, so the person in the hot seat can hear it and decide what to do with it.

A good peer room does not need to be led by someone who knows everyone’s industry. It needs warmth, so people will say the true thing. It also needs edges, so not every thread gets followed, the conversation stays close to the real question, and the processing happens in the room rather than eating it.

Women’s Enterprise BC, whose peer mentoring program I was part of facilitating, reports that before participating in its mentoring program, only 39% of mentees expressed confidence in their business skills. After the program, that number rose to 95%.

Confidence is not quite the word that comes up for me. What I saw was even more grounded than that. I saw business owners create better conditions for judgment. Better questions. Better challenge. Better ways to test whether the issue they brought into the room was actually the issue that needed attention.

The women in this cohort were already capable. Already experienced. Already carrying real businesses with real consequences. What the room gave them was not expertise they lacked.

It gave them a place to examine the decisions their businesses were asking them to make.

Access to advice is easy to find. A room that helps you think more clearly about your own business is considerably rarer. That is what this was.

If this piece stayed with you, check out more essays from The Red Thread.

Elsewhere:

of writes about building support systems when you’re an ‘only’. If you’ve ever built alone, developed alone, led alone … this is for you.

If you’re seeking community as you build and learn AI, take a peek at article on the need for community during the AI shift.

Ruchika Tulshyan and Jodi-Ann Burey, “Stop Telling Women They Have Impostor Syndrome”
A useful counterweight to the way we sometimes make professional doubt sound like an individual confidence problem. See HBR article.

If you know someone running a business from the inside of the problem right now, send this to her.

We’re in the bittersweet last 24 hours of a family trip to France. Much needed, much enjoyed. Markets, castles, countryside bike rides, mounds of baguettes and rivers of rosé — an excellent antidote to months of intense work, of being online, of deep AI immersion.

And though I was trepidatious about hauling out my rusty linguistic abilities, I felt myself sink pleasurably back into language.

Not the LLM kind. French.

Latin-based, gendered, rule-bound, full of soft sounds and inherited forms and strange little exceptions and the mild public humiliation of saying things slightly wrong. I had forgotten how alive that makes me feel.

For months I’ve been living inside AI-assisted language — shaping ideas, testing prompts, reworking lines, asking multiple machines for the cleaner phrase, the sharper distinction, the sentence that carries more impact. And I’m suitably dazzled most of the time. AI is making me faster, helping me catch thoughts before they vanish, letting me think through writing at a pace I couldn’t manage alone. For someone who likes to think on the page, that’s no small thing.

But France has reminded me of another relationship I have to language entirely — language as reaching toward someone to connect, as listening to truly understand, as asking a question poorly phrased but well intended enough that you’re understood anyway.

Wherever we stayed, at market stalls and cafés, in little boutiques, French became less of a skill and more of a social offering. I’d chat with the proprietors about their family histories, the region, the food, the restaurants they loved.

My grammar wasn’t always good and my tenses got more experimental the deeper into a conversation I went. I forgot useful words at exactly the moment I needed them. I gestured, circled, backed up, tried again. And in spite of my rust-encrusted French, it worked … not perfectly, but better than perfectly. Warm smiles and visible appreciation. And that subtle shift when you stop being only a tourist and become, briefly, a person making an effort inside someone else’s world.

At one jewellery shop, the owner ended up comparing notes with me on raising daughters while mine watched the exchange unfold with that preteen mix of curiosity and reserve. The woman was so pleased by this little interlude that she gave her a small bracelet as a memento. This moment of meaning didn’t happen because my French was fluent. It happened because my French was human.

And yes, Google Translate or AI was always at the ready — to check the gist of what I wanted to say, borrow a structure, remind myself of a word. But then I’d put my phone down and do my best. AI helped me approach the conversation but it didn’t have the conversation. It didn’t watch the vendor’s face to see if she understood me, couldn’t notice when a joke sort of landed, missed the small pulse of relief when a sentence survived the journey from my mind to my mouth to another person. AI produces language without embarrassment, which is part of its appeal and part of how it fails us.

My husband and daughter don’t speak French and relied on me to translate, but they made their own offerings: big smiles, loud bonjours, cheerful mercis, open faces. Often that was enough to change the temperature of a room.

Watching Europeans around us move between languages with an ease that always slightly humbles me, I kept being struck by the power of human language itself. Yes, language as information transfer, but also language as hospitality, a social offering.

French is a language I’m trying on again after years away. There’s pleasure in the return and also frustration. I had so much I wanted to ask, so much I wanted to say, and I wanted to be quicker and more precise and more elegant about it. I wanted the sentence in my head to arrive intact. Instead it felt like it arrived wearing the wrong shoes, missing a button, slightly out of breath. And still, people smiled.

Speaking imperfect French makes the support around the words themselves more visible - tone, gesture, context, patience, repair. The meaning doesn’t just sit inside the words. It is what’s being built between us.

This is easy to forget when language is so frictionless.

Inside AI, language can feel endlessly available: ask, receive, revise, refine, make it warmer, make it sharper, make it sound like me, make it sound less like me. The sentence appears. The roughness goes away.

Fluency is becoming easy to attain. Anyone can generate polished language now, so polish has stopped telling us very much. What’s left is the nuance, the pleasure, the learnings that AI simply can’t reproduce.

France has reminded me of this in the most simple ways. Ordering coffee. Asking about cheese. Listening to a château owner describe the generations who’d lived there. A chatty cabbie extolling then virtues of the Paris-St Germain football club.

Perfection and polish wasn’t my goal (though hey, I’m très motivated to improve) …. connection was. And connection seems to have its own grammar: mistakes, pauses, a willingness to sound just a little foolish, the humility of entering a language with less control than I’d like and discovering that people are kinder than self-consciousness predicted.

AI- led language treats friction as the enemy: anything wordy must be streamlined, anything effortful automated, anything awkward smoothed.

I’ve savoured the quirkiness of flawed attempts at dialogue. I’ve remembered that language isn’t only something we produce. It’s something we try on, fumble with imperfectly if earnestly, and use to make ourselves a bit more available to each another.

Words don’t need to arrive perfectly to carry meaning. Sometimes the effort itself is the output: the pause, the gesture, the wrong tense, the hopeful smile. And this week, in my halting French, that’s felt like more than enough.

As Victor Hugo wrote:

Une langue ne se fixe pas. L’esprit humain est toujours en marche, ou, si l’on veut, en mouvement, et les langues avec lui.

(A language is not ‘fixed’. The human spirit is always moving, and languages move with it.)

On the school run last week, my daughter shared the moments in her life when she feels her truest self. Every mother’s dream.

One of them was amongst horses. She wondered if she, like her much-admired ‘horse-whisperer’ instructor, could actually have a job working with them.

Normally, the pragmatic part of me would have a response ready before she could finish the sentence: a beautiful passion, wonderful for your confidence, maybe not a sustainable career path. Some nugget of parenting restraint took hold. I surprised myself and went in an entirely different direction — something about building a life around your deepest interests, about income that follows purpose rather than replaces it. Who am I ;)

I’ve been noodling on that redirect since.

We are living through a moment where our traditional definitions of “stability” are evaporating. Job categories we thought were reasonably bulletproof — coding, law, research, engineering — are more tenuous than they were five years ago. The old mandate to pick a lane, the gold standard for many Gen X and elder millennials, now feels less like stability and more like a single point of failure.

The Portfolio of Purpose

It is little wonder my daughter is inspired by her instructor.

A young Swede living in the mountains of British Columbia, she has built what I’d call a Portfolio of Purpose. Her career hasn’t been a ladder — more of a mosaic. Ranch hand, wine rep, various stints in other fields. But running underneath all of it was an enduring love of horses.

A few years ago, she pulled that love from the backdrop to the centre. She now runs a mix training, camps, equine rehab and therapy.

There is no LinkedIn job title for what she does. AI is not coming for her work. She built toward a center of gravity rather than a job category. The result is a business that is financially real and hard to replicate because it is the accumulated expression of who she is and what she knows.

Subscribe now

The Theory: Identity is Action

This is what London Business School professor Herminia Ibarra means when she says identity isn’t discovered (solely) through reflection — it’s built through action. The mistake some of us make is trying to think our way into the right one. Ibarra’s core finding, drawn from deep research of professionals in transition, is that you act your way into a new identity. You don’t find your purpose and then build; you build, and the purpose coheres around what you keep returning to.

My new fave author, , provides a necessary bridge from theory to action. A former Google exec-turned-neuroscientist, she argues in Tiny Experiments that the antidote to the “heroic leap” are small, time-bound tests. You try on a possible self without blowing up your life.

Hannah didn’t start with a thriving business, she tested: a weekend clinic here, a therapy session there, until the feedback made the next move obvious.

I did the same. I have always felt pulled to write, but had only ever done it for clients, in constrained ways, on someone else’s brief. After reading Le Cunff’s book, I gave myself permission to try it on. That experiment has become my Substack.

The Reality: Purpose vs. Tasks

This shift from job titles to Portfolios of Purpose isn’t just a lifestyle choice — it’s becoming the new economic reality.

Nvidia CEO Jensen Huang recently noted a critical distinction that most people miss: the difference between the tasks of a role and the purpose of a role. AI is incredibly good at automating tasks — the repetitive reports, the boilerplate code, the routine scheduling. But AI cannot touch the purpose.

A radiologist’s task might be reading a scan, but their purpose is the health and care of the patient.

“AI changes the tasks,” Huang argues, “but it does not replace the core purpose of a role.”

When we build a life around a job category, we are building on a foundation of tasks — which are currently being automated away. When we build a Portfolio of Purpose, we are not defined by our ability to use a tool, but by our ability to architect an intent that the tools then execute.

This Isn’t Just a Reflection for Twelve-Year-Olds

The conversation with my daughter has stayed with me partly because of what she said, and partly because of where I am.

I’ve spent the past couple of years reading extensively about AI — the disruption arguments, the optimist counterarguments, the research underneath both. That reading changed how I think about work. Not abstractly. Concretely. It changed what I heard in the front seat of the car last week.

What I find myself believing is this: the most durable life architecture isn’t built around a job category that seems stable. It’s built around a clear sense of what the work is for — a center of gravity — and then the practical skill of assembling income streams around it.

That’s not “follow your passion.” The follow-your-passion advice was always a bit of a lie, and AI hasn’t changed that. What AI has done is make the question sharper. Tasks are moving. Purpose is what stays.

Share

I left the gym last week and fell into a conversation that stayed with me. The woman I was talking to had built a long, serious career in tech — managing global teams, working across continents, at the top of what she’d constructed over decades. At some point in the middle of her life and the height of that career, she stepped away. She requalified in a field that had always called to her: working with seniors in recreation therapy. She’s now building a portfolio of purpose — several different expressions of the same animating thing, laddering together into a practice that is entirely hers.

She isn’t starting over. She isn’t reinventing herself in the wellness-brochure sense of the word. She’s building toward something she already knew was there — in the same iterative way Hannah did, letting the income expressions assemble around the center of gravity rather than the other way around.

I still want my daughter to go to university — not for a credential or a linear path to a profession, but for what it actually offers: ideas she hasn’t encountered yet, people who think differently, the discipline of holding complexity. That still matters. What I’ve stopped believing is that it’s the first step onto a stable career ladder. She might build a business around horses, be an artist, and manage a small investment portfolio. Those things live in the same sentence now, without irony.

The Question in the Boardroom

I am seeing these reflections at various levels of business.

I was recently at at a boardroom table with eight business owners — a mix of ages, life stages, and industries, from their 30s to their 50s. The conversation was supposed to be strategic: how does a thriving small business grow?

Except the business in question was already thriving. Already profitable. Already built around the owner’s interests and life. When that became clear, the room shifted.

Someone asked: why grow at all?

And then: what does growth actually mean here? If the business is serving your life, if it’s funding what matters and leaving room for the rest — by whose definition is that not enough? Have you given yourself permission to call this a success?

It’s like a shift has happened from the self-help stacks to the centre of the boardroom table. When the friction of execution is reduced, we have to look more carefully at the value of the endeavour itself. That conversation used to happen in therapy offices and journal entries. Now it’s happening in boardrooms.

What We’re Losing

Just as more of us are giving ourselves permission to structure a working life around purpose rather than job category, organisations are automating away the conditions that made self-discovery possible in the first place — the early roles where you had room to make mistakes, test yourself, and figure out what you actually wanted.

When firms stop hiring junior workers because AI can handle the rote work, they also stop building the pipeline of people who will eventually develop the judgment to oversee it.

Those first jobs, those early messy years, were never just about inexpensive labour. They were the original tiny experiments. The place where you found out — through friction, through consequence, through the work itself — not just what you were good at, but what you actually wanted. The entry-level role was the nursery where purpose got tested before it cohered. When firms stop hiring junior workers because AI can handle the rote work, they don’t just lose the output. They eliminate the petri dish.

The Antidote

Ibarra’s research points to the same conclusion Le Cunff arrives at from neuroscience: stop waiting for clarity before you move. Identity isn’t discovered solely through reflection. It’s built through action. The people who successfully change aren’t the ones who figured it out first. They’re the ones who started moving and paid attention to what the movement taught them.

The antidote is the same whether you’re 12 or 52: build toward what you keep returning to. Not instead of practical skills or income — alongside them, with purpose as the organizing principle.

Hannah figured this out without a framework. The woman from the gym figured it out without one either. They just kept building toward the thing that was true, and the income expressions assembled around it.

A single job title isn’t a big enough container for a whole person.

I still hope my daughter will have the chance to enjoy university — not for a credential or a linear path to a profession, but for what it actually offers: ideas she hasn’t encountered yet, people who think differently, the discipline of holding complexity. That still matters. What I’ve stopped believing is that it’s the first step onto a stable career ladder.

She might build a business around horses, be an artist, and manage a small investment portfolio. It’s clearer than ever that in a world of autonomous agents, picking a path is a vulnerability. The only real job description left is remaining the architect of your own intent.

Subscribe now

What I’m reading on this:

Herminia Ibarra, Working Identity (updated edition) — on how career transitions actually happen: through action and experimentation, not planning

Anne-Laure Le Cunff, Tiny Experiments — on replacing rigid goals with small, time-bound tests. The most useful reframe I’ve found for anyone who knows what they want but can’t find the door

Erik Brynjolfsson and the Stanford Digital Economy Lab — on AI and early-career employment

David Autor at MIT — on AI’s opportunity to extend the reach of human expertise rather than replace it

Jean-Paul Carvalho at Oxford — on the pipeline problem, interviewed by McKinsey, May 2026

Last Friday I sat in a room of eight hundred people at The Forum's Finale Pitch event. The Forum is a national charity supporting women entrepreneurs — I sit on the board, and I'd worked with one of the finalists as a pitch coach in the lead-up. The room was buzzing and filled to bursting.

Three women founders pitched - from AI optimizing billing for doctors to 3D ultrasound. Hundreds of investors, partners, allies in the seats - and a ‘Dragon’ on the stage.

For the finalists of Finale, the night isn’t just a pitch competition. It can be a pivot point - the trajectory of their companies can fundamentally shift because of who watched them pitch.

This is what Jim Collins calls “who luck” — the luck of meeting the right people at the right time.

He argues it’s the most underrated factor in any big opportunity, more decisive than intelligence or effort or strategy. It’s the right conversation in the right hallway. The introduction that opens the door you didn’t know existed.

Collins is right that this is how opportunities actually get created. The concept appeals because it sounds democratic. The hallway is open. You just have to be in it.

The Sponsorship Gap

Who-luck only works as a theory of success if you assume the room is equally reachable from wherever you start. It isn’t. Entry is controlled by the people already inside it.

In Sinéad Fitzgerald’s article in Code Like A Girl on the sponsorship gap in tech, the concept of who-luck starts to wither.

She separates three roles usually blurred together - mentors, coaches, and sponsors - and shows what each one actually does.

• A mentor shares the map.

• A coach works a specific inflection point.

• A sponsor stakes their reputation on yours, and uses their voice for you in the conversations you haven’t been invited into yet. The person being advocated for mostly doesn’t know it’s happening.

The 2025 Women in the Workplace report from McKinsey and LeanIn confirms the cost of this distinction.

Employees with sponsors are promoted at nearly twice the rate of those without. Yet, only 31% of entry-level women have a sponsor, compared to 45% of entry-level men.

Sponsorship helps. It just helps men more.

The gap isn’t about awareness, and it isn’t about women failing to ask. It’s proximity. Sponsors bet on people already close enough to be visible to them. Proximity, in most organizations, runs along the same lines the hierarchy was drawn on.

That’s the architecture. And it’s what makes Fitzgerald’s coda the line the rest of her essay rests on: the system will not change because women become better at working around it.

The New Arithmetic: Bias at Scale

A few weeks ago I wrote about Editing Out Equality — the discreet removal of equity language and measurement from corporate and public documents. The architecture I was describing then is the same one running underneath this essay.

Measurement wasn’t the goal. Measurement was the correction. When organizations had to count who was being sponsored, who was being promoted, who was being funded — the informal mechanism had to widen its choices, or at least defend them. Take the counting away and the mechanism reverts to its defaults. The same people get sponsored. The same names get carried into the same rooms. And no one outside the pattern can see it happening, let alone challenge it.

And now those rooms are being automated.

The same arithmetic that defaults sponsors to safe bets is now running inside AI systems making decisions across hiring, performance management, promotion recommendations, investment screening, and founder evaluations. They’re trained on historical data:

• who got promoted

• who got rated ‘high-potential’

• who got the cheque.

The AI doesn’t know the patterns are a record of who the room has historically chosen. It just learns the shape of the choice and reproduces it forward.

The arithmetic that favoured the safe bet in the meeting now runs at scale, with less visibility and more consistency than the manager it’s replacing. Same bias. Faster.

Subscribe now

Engineering The Room

As I left The Forum Finale, I bumped into a friend who worked in venture capital. We started talking about one of the finalists.

“I know exactly who she should speak to,” she said. She named the firm, drew the connection, and committed to the intro.

That is who-luck in action. And it didn’t happen by accident. Engineering a room means deciding in advance who gets seen and who gets to see them. That’s structural intervention. The luck looks accidental on the outside. But it is not.

So. How do we create more rooms like this? Not just the annual galas, but the rooms of two. The rooms of ten. The private dinners and the curated intros.

If we don’t have the answers for how to “fix” AI bias yet, we do know the first step:

We need more women in the rooms where the arithmetic is being written. We cannot afford to be engineered out of the rooms where consequential decisions are made.

How to Change the Arithmetic

The McKinsey/LeanIn Women in the Workplace 2025 report and the Women of Influence+ Closing the Sponsorship Gap report converge on a consistent set of practices. I’ve simplified and aggregated them.

For Leaders (The Sponsors)

1. Audit your “Safe Bet” list. 92% of sponsors choose protégés based on “perceived potential”—a standard shaped by familiarity. If your list is homogeneous, the arithmetic is doing the work, not you.

2. Make sponsorship a stated expectation, not a private virtue. Only 35% of women say their workplace actively encourages sponsorship. If it’s not a stated expectation, it’s a private virtue that doesn’t scale.

3. Spread stretch opportunities deliberately. Decide based on skills, not personal preference or arbitrary criteria. Stretch projects are who-luck in slow motion — they decide who’s ready when the next promotion lands.

4. Differentiate mentorship from sponsorship. Mentoring is sharing the map. Sponsoring is saying the name in the room. Both matter, but they’re not interchangeable, and conflating them is one of the reasons women end up with the lighter version of support.

For mid-career women looking to build who-luck:

1. Look outside your organization. 41% of women who were never sponsored cited lack of access to senior leaders. External sponsors — investors, board members, industry community leaders — can create lateral moves that bypass internal structures that have already decided what they think of you.

2. Be specific about what you need. 36% of unsponsored women said they didn’t know how sponsorship was supposed to happen. Most people are willing to help and don’t know what would actually be useful. “Can I pick your brain” gets the warmest no. “Would you introduce me to one person, and tell them why I’m worth the meeting” gets the introduction.

3. Build peer sponsorship. Lean In’s research shows that peer-level support is as effective as traditional mentorship. If you’re in a room someone you respect isn’t, say their name. Sideways sponsorship compounds — and it builds the muscle for when the rooms get bigger.

For everyone:

Notice the rooms you’re already in. If you find yourself in one with who-luck flowing freely, your job is not to network for yourself. It’s to draw a through-line out loud for someone who isn’t being seen yet.

Share

The Bottom Line

So who decides who’s in the room?

The people already in it. Every mechanism in this essay - informal sponsorship, historical data reproduced at AI speed, the dismantling of the infrastructure that used to correct both - points at the same architecture.

The path forward is simple, and uncomfortable. Either the people in the room start using their reputations differently, or the room stays the same shape. No woman can change the shape of a room from outside it.

Friday night was an engineered room. The question is how we build more of them, every day.

Try the bet that feels riskier. Carry the name into the next room.

That’s where the arithmetic changes.

What I’m consuming on this:

• Sinéad Fitzgerald, “The Most Expensive Gap in Tech Isn’t a Skills Gap” (Code Like A Girl, April 15, 2026)

• LeanIn.Org & McKinsey, Women in the Workplace 2025 (December 2025)

• Women of Influence+, Closing the Sponsorship Gap: How Purposeful Advocacy Accelerates Women’s Careers (April 2025)

• Jim Collins on “who luck” — Tim Ferriss Show

• Katrina Carroll-Foster, “Editing Out Equality” (April 2026)

I’ve been volunteer facilitating a peer mentoring group for the past few months. A small room of women running real businesses — revenue coming in, customers served, lights on. Every one of them resource-constrained on two fronts that matter: cash and time.

Session after session, I kept noticing the same recurring gap: the struggle to find capacity to work on the business while being mired in it. The wisdom in that room is immense, but without a plan to anchor it, that experience only gets applied in fragments.

So I went looking for a one-page business plan to share as a tool.

I wanted something light enough not to add to their existing load, but with enough substance to actually move the needle on their outcomes. It needed to be short enough to hold in your head, yet specific enough to test a decision against on a Tuesday afternoon when an opportunity shows up and you need to decide by Thursday.

Nothing I found quite worked. So I built one — and gave it away (see👇). That last part matters more than the first, and I’ll come back to it.

My starting brief to Claude was simple: research what’s out there, figure out the best one-page format for this kind of business, come back with a recommendation.

It came back recommending Jim Horan’s One-Page Business Plan, with a clear argument for why. The two other obvious candidates had the wrong orientation:

• Business Model Canvas — designed for startup founders testing unproven ideas. Nine boxes with language like “key partnerships” and “cost structure” that feels corporate. Wrong for someone running an established service business.

• Lean Canvas — Ash Maurya’s startup adaptation of BMC. Better, but still oriented toward product-market fit and iteration cycles. These women aren’t iterating on a hypothesis. They’re running real businesses.

• Horan’s One-Page Business Plan — built for existing business owners. Flows from Vision through Mission, Objectives, Strategies, Action Plans. Linear, sequential, practical.

Worth saying: BMC and Lean Canvas aren’t bad frameworks. They’re the right tool for a different room. The founders and early-stage teams I usually work with — technology companies still finding product-market fit — get real value out of both. This group isn’t that. Match the framework to the business.

Fine. I approved the direction. But two things needed pushing.

First: the framework itself. Horan gives you direction — vision to action, in a straight line. But it doesn’t make you think about the business itself. No prompt to articulate why a customer would choose you over the alternative. No question about who the customer actually is or how the money works. BMC and Lean Canvas have those elements, but packaged wrong for this audience.

The solution was a hybrid — Horan’s directional flow, layered with the business model questions from BMC and Lean Canvas that Horan misses. Six sections. Vision. Customer. Value and positioning. Revenue engine. This year’s objectives and what you’re saying no to. Next 90 days. The “saying no” section tends to be where the most interesting work happens.

Second: I caught a conceptual error in my own draft. I’d labelled one section “value proposition” but it was actually doing positioning work. Those are different questions. Value is what outcome the customer walks away with. Positioning is why you and not someone else. Both matter. I separated them.

Along the way I ran the same brief through a different LLM and brought its version back - not to replace the work, but to stress-test it.

It had real improvements I integrated:

• A “Magic Number” concept (the monthly revenue that covers all costs and pays the owner what they’re worth)

• A prompt that asks what you want to pay yourself and works backward.

Both were sharper than what I had. It also had weaknesses I rejected — section names like “North Star” and “Quarterly Sprint” that sounded like a startup playbook.

I made every judgment call. AI did the production. The back-and-forth is what got to something actually useful.

Share

I read a piece by Cristina Patrick this week called “You Don’t Have to Figure This Out Alone” — a collective essay by nine women writing about AI and community. One section really hit me:

“AI may be the first technological shift that touches nearly every life on earth, and most people have no idea how it’s already shaping theirs. … That is why community matters. It’s a place where people can ask honest questions without feeling stupid, and leave with something practical.”

These are the business owners I’ve been sitting with. People whose question isn’t “how do I ship faster” but “how do I work on my business at all.” The essay frames something important: the dominant AI narrative is optimized for individual anxiety, not collective possibility.

That’s not the narrative I want to add to.

A year ago, building this package would have taken me a week. It took four hours. Which means things that used to be consulting engagements can now be tools. Thing

s that used to live behind a paywall can now live in a Substack post.

I’m shifting toward solving problems in plain sight and sharing the results.

Subscribe now

The One Page Plan

One document. Eleven pages.

• A blank fillable template — the six-section hybrid framework

• A worked example, fully completed, so you can see what finished looks like

• A section-by-section guide — what to include, what good looks like, common traps

• AI prompts for each section, designed to work in any LLM (ChatGPT, Claude, Gemini — free tiers are fine)

• Resources — the books, the free planning tools, the places to go next

The prompts are worth one more word on. They’re not “write me a vision statement.” They’re designed to work like the conversation I’d have with you if we had more time. The aim is to replicate what a good advisor does: slow you down, ask the question you’ve been avoiding, help you hear yourself more clearly.

Use it, adapt it, share it with someone who needs it. If you try it, I’d love to hear what happened — what worked, what didn’t, what the plan made clearer or more complicated.

Download the One Page Plan.

Click for One Page Plan Package

Most startup founders have heard some version of “great idea, just too early.”

It is the polite euphemism people use for companies that didn’t make it—the corporate equivalent of “she’s in a better place.”

I heard it about our own startup years ago and eventually filed it away under “lessons learned.”

On a rampage down memory lane, I re-discovered an old interview on IdeaMensch — one of those early founder-profile sites that popped up when the startup ecosystem was still figuring itself out. The photo accompanying it radiates the excitement my partner and best friend Tracey and I felt about the launch of Hoseanna, our new startup.

It was 2011. Hoseanna was a subscription e-commerce platform for the “everyday stuff”—hosiery, razors, feminine care—items women need but that no one, at the time, was making convenient to buy. We were three months in. Tracey was in New York. I was in Vancouver.

We ran the company on Skype calls and strong coffee, and built our own subscription e-commerce platform because Shopify was barely a thing. Facebook ads were just starting. We raised a small round, and landed our banner ad on the New York Times homepage … I wish I’d saved the screenshot!

I might have closed the tab and moved on. But a few days later, I was watching a rerun of Dragon’s Den with my daughter when four young women from Montreal walked in and pitched a hosiery subscription company. The Dragons loved it. One of them called the model “brilliant.”

I sat there on the couch, bristled, and shot Tracey a text … ‘listen to this’. Fifteen years after we were told our idea was too early, here it was — same concept, different room.

Why was the room different? Because they were pitching to a panel where three of the five Dragons (including ) are women. The composition of the room changed. So did what counted as "obvious."

I went back to that old article.

Reading it now, I remember exactly who we were at that moment. Two women in our mid thirties, both with serious careers behind us, both busy, both tired of walking to the corner store for pantyhose and fem care products. We’d watched Dollar Shave Club and Harry’s start up — subscription convenience, built for men — and thought: why shouldn’t women have this too?

The name was Tracey’s idea. She’s a branding person. Hoseanna — hosiery meets hallelujah!

When we pitched Hoseanna to VCs after our angel round, the responses ranged from dismissive to bizarre. “Why would women ever want this?” one asked. “Why would they need the convenience when they could walk to the corner store?” Another recommended we abandon our value proposition entirely to sell lingerie to plus-size middle America …a suggestion that could not have been further from the utility we were building.

These were experienced investors. They were also people who had never, in their lives, needed to buy a pair of pantyhose or period care.

Dollar Shave Club launched the same year we did. Their thesis was identical: subscription convenience for a commodity product. Within four years, Unilever bought them for a billion dollars. Harry’s followed a similar path. Both were premised on the idea that men shouldn’t have to run to the store for razors.

Investors understood that immediately. A room full of men who shaved every morning could feel the problem in their own hands. Of course you’d want razors delivered. Of course that’s a business.

But two women standing in that same room, saying “women want the same convenience for the things they buy every month,” was treated as a novelty. A curiosity. Something that required an exhausting amount of translation.

The model wasn’t the real problem. The room was.

We eventually ran out of money. The velocity wasn’t there.

We hung up our shingle, hung down our heads and went back to our respective worlds.

The feeling that stayed with us? Shame. Not just disappointment but ashamed of failure. We felt we’d failed our angel investors and customers. There was no culture of “failing forward” to catch us, no founder communities celebrating the “pivot” or treating the experience as signal rather than stigma. We didn’t talk about it. We just moved on.

Tracey has since founded three more ventures (plug for Flatbush Granola here), all drawing on what Hoseanna taught her about brand and consumer behavior. I discovered that I loved the “space” of the build - the strategy, the venture dynamics, the mechanics of scale. That led me into a fourteen-year practice working with both founders and investors.

The failure wasn’t an ending; it was a door. But I do think about timing constantly.

When Tracey and I were pitching, women made up roughly 17% of angel investors in Canada. There were almost no women-led venture funds. The51 didn’t exist. Women’s Equity Lab didn’t exist. The ecosystem that now connects women founders with women investors was a decade away from being built.

The numbers have shifted since then … meaningfully in some places, stubbornly in others.

Women now make up about 35% of Canadian angel network membership. Organizations like The51, Women’s Equity Lab, The Forum, and WeBC are doing the structural work of building networks and raising capital.

In the US, PitchBook reported that female-founded companies raised a record $73.6 billion in 2025. But that record comes with an asterisk the size of the CN Tower. More than $30 billion of that total went to just two companies: Anthropic and Scale AI. Remove them, and the record vanishes.

All-female founding teams still receive roughly 2% of total venture dollars … a number that has barely moved in a decade. In fact, the share of women receiving a first VC cheque has dropped since 2021.

$73.6 billion and 2%. Both are true at the same time.

Share

When you pitch an idea that’s built from lived experience and the people evaluating it have never lived that experience, you’re not just fighting for funding. You’re fighting to make the problem understandable.

That’s not a taste problem or a judgment problem. It’s a composition problem. Who is in the room determines what gets seen as ‘obvious’.

The growth of the women’s investing ecosystem isn’t just a nice story about inclusion; it’s an infrastructure correction.

Every woman who joins an angel network or writes a cheque is changing the default settings of the room. At a macro level, it’s a boost to GDP. Research consistently shows that women-led companies generate more revenue per dollar invested and run leaner.

I recently wrote on Linkedin about the Fosbury Flop - the idea that when the landing surface changes, the winning technique changes too. AI is our new “foam mat”; it has collapsed the cost of experimentation.

The landing surface is changing. More women investors. More women-led funds. More organizations building the connective tissue. This creates rooms where the problem is already understood and where a founder can stop translating and just start building.

We were early. But being early is different from being wrong. I wish someone had told us that in 2011.

I wonder how many other Hoseannas are out there—sitting in the archives of hard drives and notebooks because the room wasn’t ready. Ideas that were sound, founders who were capable, and timing that was just early enough to be fatal.

Tracey and I are still best friends. That is the thing I am proudest of - more than any business we’ve built since. We went through something turbulent and transformative and came out with our friendship intact. In any partnership, that is never a given.

The ideas and the founders have always been there. The work now is making sure the room is finally ready when the next Hoseanna walks in.

For years, “Believe Women” stood as the rallying cry for the #MeToo movement.

But , journalist and veteran observer of women in the workforce for over a decade, warns of a disturbing shift in her recent New York Times essay. This ‘defining message’ has been replaced by a new one:

Erase Women.

The change isn’t a loud, public reversal, but a quiet, institutional scrubbing. Discussions regarding gender have become so politically volatile that organizations now view the word “women” itself as a liability.

I read this article just before a meeting for the nascent Lattice Network with a group of women leaders. The timing was poignant. As we sat down to discuss leadership, access and influence, I was reading about how the very language we use to define our progress is being quietly deleted.

By way of context, I sit on the board of a national charity supporting women entrepreneurs. I work with women founders building companies. I facilitate peer mentoring with women through my local chamber of commerce. This is the water I swim in - not as a policy expert, but as someone who sees it up close, every week.

So when I started digging into the data, I wasn’t reading abstracts. I was reading about the scaffolding around the women I work with.

The venture capital numbers provide a grim baseline. Female-only founding teams received only 1% of total US venture capital in 2024. While the headline figure of $73.6 billion raised by female-founded companies in 2025 sounds like a record, the reality is skewed: nearly half that capital flowed to just two AI giants - Anthropic (co-founded by Daniela Amodei) and Scale AI (co-founded by Lucy Guo). Without those massive outliers, the “record” vanishes.

The point I want to make here is that te funding picture hasn’t moved meaningfully in a decade despite years of programs, pledges, and public commitments designed to shift it. Those programs are now quietly disappearing. And the conversation about why the numbers haven’t moved has become harder to have.

In corporate boardrooms, the retreat is evident in what companies choose to report - or not.

Subscribe now

In Canada, representation looks stable only because turnover is slow - women reached 30.5% of TSX board seats in 2025. But the pipeline is collapsing: the share of new board seats going to women dropped from 45.3% in 2023 to just 31.9% in 2025.

Meanwhile, mentions of “DEI” across corporate Canada have plummeted 79% since mid-2024. Shopify eliminated several diversity programs. The Canadian Securities Administrators even paused changes to diversity disclosure requirements in April 2025.

In the US, the trend is even sharper. A recent study by the Conference Board and Harvard Law showed that US companies reporting data on women in management fell from 71.2% to 55.1% in a single year. Disclosure of female board members dropped from 90.5% to 60.4% !

These companies haven’t suddenly stopped hiring women; they’ve simply stopped measuring whether they are succeeding.

Stanford sociologist Shelley Correll points out organizations are starting to cancel Employee Resource Groups. She describes this as an “overreaction to what even Trump is asking them to do”. It is an overreaction to political pressure, using convenience as cover for a lack of commitment.

The institutional picture in Canada is harder to look at.

The federal government’s Women and Gender Equality department — WAGE — is facing budget cuts of roughly 80%. The National Action Plan to End Gender-Based Violence, which funds transfers to the provinces, effectively falls to zero. The National Association of Women and the Law called it a move that would “effectively gut the entire department responsible for women’s rights and gender equality.”

In the most recent federal budget, only 4.5% of budget measures were specifically targeted toward women. Twenty-five percent targeted men. Canada has a Gender Budgeting Act. We have mandatory Gender-Based Analysis applied to Cabinet decisions, Treasury Board submissions, and budget proposals. The architecture exists. The funding is being hollowed out underneath it.

The reach of this “erasure” extends into the nonprofit sector.

A 2025 piece in The Philanthropist documented what’s happening to nonprofits that serve women, racialized, and Indigenous communities. Organizations are self-censoring their language and mission statements — rewriting how they describe their own work — to avoid being seen as too political and losing their funding. Service capacity is shrinking. Staff are burning out. And the community trust that takes years to build disappears alongside the programs.

One of the people quoted in that piece put it better than I can: “It is not that there is an attack on DEI that is worrisome; it is the success of that attack that is worrisome.” — Sudip Minhas, Settlement Assistance and Family Support Services.

In the United States, some specifics are worth knowing.

The Equal Employment Opportunity Commission — the US federal agency responsible for enforcing workplace anti-discrimination laws — sued a Coca-Cola distributor for hosting a women’s networking retreat, alleging it discriminated against men.

Defense Secretary Pete Hegseth blocked the promotions of Black and female military officers, eliminated the Defence Advisory Committee on Women in the Services (which had operated for over 70 years), and axed the Women, Peace and Security Program. For the first time, there is no woman at four-star rank in the US military.

At the UN Commission on the Status of Women in March, the United States cast the sole vote against the Agreed Conclusions — a document adopted by consensus for eighty years. The US delegation had submitted over 90 amendments, including a proposal to narrow the definition of “gender” to mean only “men and women.” It failed, 37 to 1.

This matters beyond US borders. Canadian corporate and regulatory behaviour already tracks US developments closely — the CSA’s decision to pause diversity disclosure requirements being one example. The assumption that this stays south of the border seems increasingly optimistic.

I’m not a policy expert. My vantage point is specific: I see women building companies, raising capital, working within systems that weren’t built with them in mind. I see what changes when the support structures around them shift.

What I see right now is a gap forming between what’s happening inside organizations and what can be said out loud. Real damage happens there — not through dramatic reversals but through quiet ones. A policy removed. A report no longer filed. A program rebranded into something that no longer has to prove it works for the people it was designed to help.

Lipman references the work of Erica Chenoweth and Zoe Marks, two political scientists at Harvard who’ve studied the relationship between women’s participation and democratic resilience. Their core finding, published in Foreign Affairs: when women participate in movements in significant numbers, those movements are more likely to succeed and more likely to produce egalitarian outcomes. Authoritarian and authoritarian-leaning leaders have a structural incentive to suppress women’s participation. It’s not incidental to the playbook. It’s a feature of it.

Their argument is that reversing progress on gender equality is a hallmark of authoritarian governance worldwide — Turkey, Russia, Hungary all have their versions. The Canadian version is quieter. We’re not decriminalizing domestic violence. We’re defunding the department that fights it. We’re not saying women aren’t equal. We’re removing the language that lets us check whether they are.

Share

I don’t have a policy prescription for any of this. I don’t think the answer is simply “more programs” — Dobbin and Kalev’s research at Harvard has shown that many traditional diversity programs don’t move the needle anyway. What does work, according to their decades of data: formal mentorship, skills training available to everyone, structural changes to how organizations recruit and promote. Work that doesn’t need to be labelled “DEI” to be effective — but does need to be measured to prove it is.

When we shrink our “institutional dictionary”—when “women” becomes a word you cannot use in a grant application or a mission statement—we don’t solve the problem. We simply lose the language required to see it.

I think about this for my daughter, who’s eleven. I wrote about her on International Women’s Day — the report card on where things stand. Education: A. Legal rights: D. Political representation: F. Funding the future: D.

The grades haven’t changed. What’s changed is the willingness of institutions to keep tracking them.

The scaffolding around women’s participation in public and economic life — the measurement, the funding, the language — isn’t decoration. It’s load-bearing.

And it’s being quietly removed.

Subscribe now

What I’m reading on this:

Joanne Lipman, “I’ve Covered Women in the Workplace for 15 Years. Something Alarming Is Happening” (New York Times, April 6, 2026)

Erica Chenoweth and Zoe Marks, “Revenge of the Patriarchs” (Foreign Affairs, March/April 2022)

The Conference Board (US) / Harvard Law School Forum on Corporate Governance, “DEI in Transition: 2025 Corporate Diversity Disclosure Trends”

Osler, Hoskin & Harcourt, “2025 Diversity Disclosure Practices: Diversity and Leadership at Canadian Public Companies”

NAWL, “Feminist organizations denounce 80% funding cuts to Women and Gender Equality Canada”

The Philanthropist, “How DEI Rollbacks Threaten Non-Profits That Serve Women and Racialized and Indigenous Communities” (May 2025)

NUPGE, “Budget 2025 Analysis: Equity, Diversity, Inclusion, and Accessibility Focus”

UN Meetings Coverage, “Women’s Commission Annual Session Begins with Contentious Recorded Vote” (March 2026)

Canadian HR Reporter, “With disclosure on the decline in Canada, is the commitment to DEI still there?”

Last week I was pulling out of my gym’s parking lot (smilin’ and sweaty in my lil’ Mini) inching past a stop sign to see around the corner … when a young woman flew past me in her Tacoma and flipped me the middle finger.

No near-miss. No honk. No reason I could see. Just a finger, held high, held long.

I was stunned. Then soooo pissed off. Then stunned again. I muttered all the way home. How did she think that was okay? Would she have done it to my face if we’d been walking past each other on the sidewalk? Or was it only possible from behind the windshield - the same safe distance we get from a screen, except this time she forgot she was traveling IRL?

I wonder sometimes if we’ve so internalized the anonymity of online interaction that we’ve started treating the physical world the same way. Flipping someone off from a car. Ignoring the person holding the door.

Bah. Society is coarsening. I feel it … I think many of us do.

I don’t know if the vulgarity and contempt at the top of our politics is trickling down into how we treat each other, or if we’ve normalized it so thoroughly in our daily lives that we expect nothing better from our leaders.

Maybe it doesn’t matter which came first. Political leaders modelled contempt before the rest of us adopted it. But once rudeness enters a culture, it spreads.

A study in the Journal of Applied Psychology found that a single encounter with rude behaviour rewires how you perceive your next interaction … and makes you more likely to be rude in return. The effect lasts even when a full week has passed between encounters. The study’s title: “Catching rudeness is like catching a cold.”

Maybe the woman in my parking lot was just having a bad day. Or maybe she’s breathing the air of a culture where contempt is performed from the top and amplified until it feels normal. A 2024 study found that even good people behave worse when they believe rudeness is accepted around them. It’s not about character. It’s about what we’ve decided is okay.

The esteemed Stephen Lewis died this week. An odd segue, maybe, from the F45 parking lot. But, please, stay with me.

I heard Mr. Lewis speak years ago. If his name doesn’t ring a bell (!) he was a respected politician, diplomat and humanitarian — and perhaps the most gifted political orator this country has produced in half a century. He didn’t suffer fools, but appeared to have an inherent ability to work with others to get things done. In his time in office he noted politics had not yet “raised malice to an art form.”

There was much to be angry about in the injustices he devoted his life to rectifying. Nobody who watched him speak would have mistaken him for someone who lacked fire. He never confused well-placed anger with crassness or cruelty. His anger had meaning and yet was civil.

Prime Minister Carney, no slouch in the orator department and receiver of a rare Davos standing ovation, called Mr. Lewis “a voice of unmatched eloquence and integrity.”

That’s what leadership sounds like when the adults are in the room.

Now read .

She writes a Substack called . Yesterday I found myself giggling (that opening paragraph of “I Need Adults In Charge“), then soberly nodding, then finally, some resigned sighing. Her diagnosis:

“The deeper problem is the gathering sense that nobody serious is in charge of anything … The serious adults have left the building and what remains is a collection of people treating the most powerful government on earth like a reality show they accidentally got cast in.”

At a Pentagon briefing, Pete Hegseth braggadociously pronounced the US was fighting Iran with “no stupid rules of engagement, no nation building quagmire, no democracy building exercise, no politically correct wars”. Plank portrays Hegseth as having the energy of “a toddler who has just learned the word no and is now applying it to everything including bedtime, vegetables, and apparently the Geneva Conventions…like dismissing the legal architecture of warfare was an edgy personality trait.”

There’s also the Donald’s Ballroom Obsession (that video on AF1, 🤣), Boob-gate, and oh right, a terrifying illegal war that’s sucking multiple nations into an axis of chaos and letting the air out of a global economy. But … more about that ballroom.

Plank coins a term for the public’s response: institutional vertigo. The rational anxiety of watching people who clearly aren’t qualified run systems that affect your life. (I’d up that to crushing anxiety.)

Where in sweet heaven are the adults? Just one person — one — with the dignity, conviction, or principled competence of a Stephen Lewis.

As I’ve got myself into a bit of a snit about the crassness of society, once again I’m going Olympian in my down-the-rabbit-hole reading. ’s “Epidemic of Rudeness” catalogues the same coarsening at the proverbial street level.

A young man vomited on Gunther’s friend at a baseball game escalator, then walks away laughing. Teens on e-bikes barrelling down sidewalks in packs, pushing moms with strollers off the curb. “No fair? Don’t care.” The cashier who won’t make eye contact or say thank you but hands over a machine with a blinking 20% tip function. (OK, those last two are mine.)

Gunther says everyday rudeness has consequences that go “beyond the degradation of public spaces.” He’s right.

Rudeness is contagious - literally. One bad turn primes the next. Our growing absence of face-to-face contact lowers the barrier further, what researchers call a “disinhibiting effect.” We’re ruder online, then we carry it into the parking lot, the checkout line, the sidewalk.

Edelman’s Trust Barometer says seven in ten people are now unwilling to trust someone with different values or information sources. They haven’t just stopped agreeing. They’ve stopped trying.

You can explain the everyday coarsening as a contagion, like people catching each other’s bad behaviour, one interaction at a time. But that doesn’t explain why the air got toxic in the first place.

I’m partway through ’s Gilded Rage: Elon Musk and the Radicalization of Silicon Valley, and it adds a further, discomfiting dimension. For the tech oligarchs he profiles — billionaires with the platforms, the capital, and now the political access to shape opinion and make policy — the destruction of civil discourse is the point. Flood the space with enough rage that negotiation looks weak, and strongman leadership starts looking like the only thing that gets results.

I haven’t finished the book. But what I’ve read so far has changed the way I’m looking at all of this.

When the coarsening at the top isn’t drift but strategy — when it’s funded, platformed, and politically rewarded — the question stops being ‘why is everyone so rude?’ and becomes ‘who benefits from this?’.

Which is why the article “The secret to Wikipedia’s success is clarity of purpose, civility and trust” (not the catchiest header) really landed for me.

On the surface, Wikipedia shouldn’t work. Volunteer-run, 300 languages, 300 billion annual page views, no ads, no algorithm, no billionaire owner with a political project. And as noted in today’s polarized climate, another huge barrier: the need to align on “controversial entries … and the passionate disagreements that inevitably arise over factual accuracy.”

And yet.

Its guiding principle is “assume good faith.” When an anonymous editor replaces something you wrote with something you disagree with, you don’t respond as if they’re a partisan hack. You assume they’re honestly trying to improve the article. That assumption changes everything about how you respond.

It does not mean avoiding disagreement. Disagreement “is how we learn from others and get smarter together.”

Same internet as some other platforms tearing us apart. Radically different outcomes.

I was walking being walked by my dogs,

thinking things over, when I heard a little ding of a bicycle bell behind me.

A young teenage girl on her bicycle rode past. Big smile. Little wave. And a loud “thank you” as we made some room for her on the path. And she was gone.

After silently fuming about the decline of civility, from the parking lot to the Pentagon and the platforms built to reward our worst instincts, that cheerful bell cut through.

Good manners and courtesy exist. And they too are contagious - they just need more carriers.

Next time I’m inching out of that parking lot, I’m going to wave. Even at the Tacoma.

What I’ve been reading on this:

📚 Jacob Silverman, Gilded Rage: Elon Musk and the Radicalization of Silicon Valley (in progress)

🔗 Liz Plank, “I Need Adults In Charge And I Need Them Now“ — Airplane Mode (Substack)

🔗 Marc Gunther, “America’s Epidemic of Rudeness“ — Medium

🔗 Harvey Schachter, “The secret to Wikipedia’s success is clarity of purpose, civility and trust“ — The Globe and Mail

🔗 Lisa Z. Fain, “Is Anyone Else Alarmed by the Decline of Civility?“

I’ve caught myself doing it a bit more often than I care to admit.

Before I’ve finished thinking deeply about something I’m working on, I’ve dipped into my LLM of choice to see what it would say.

Not because I won’t (eventually) get there on my own. But the LLM gets there faster, and the day is full, and the output is well-structured and confident-sounding. By the time I catch myself, I’ve already started reading it.

The tool didn’t replace my judgment. It intercepted it.

Gah.

Fariba Latifi’s recent California Management Review article on Strategizing in the Age of AI gave me a frame of reference for this. She argues the core challenge isn’t bringing AI into human decision-making. It’s integrating human judgment into the AI-enabled processes.

That inversion is the language I’d been missing.

The more capable these systems get, the easier it is to defer to them. Not because we’re lazy. Because the output is confident, the turnaround is instant, and everyone’s already exhausted.

The AI has to be good enough to be useful - and it is now.

But the better it gets, the harder it is to maintain the muscle that makes you (ok, me) good at knowing when it’s wrong.

has written about the difference between instinct and intuition (her Gut Decision Matrix is a good starting point), and it connects to this point.

Instinct is fast. Pattern recognition, survival wiring, that flinch before you even know what startled you.

Intuition is something else. It’s built through accumulated experience and deliberate reflection. It’s slower. It’s thoughtful. It requires conditions.

AI is exceptionally good at mimicking instinct. It matches patterns at speed. What it can’t do is replicate the conditions under which intuition develops.

And those conditions (time to reflect, tolerance for ambiguity, the space to sit with something before reaching a conclusion) are exactly what gets squeezed out when organizations optimize for AI-assisted speed.

So you end up in this loop. The system gets faster. The humans around it get more reactive. And the slow, accumulated knowing that our judgment worth trusting starts to atrophy. Not because it was taken away, but because the conditions for it disappeared.

I’ve been looking for a way to put this into practice, and in a stroke of coincidence ’ AI Literacy Guide found its way onto my radar this week. It’s a practical resource that moves the conversation from abstract 'AI worry' into a specific set of skills for staying the primary architect of your own work.

Weeks is a journalist (the Globe and Mail, CBC, Chatelaine) who’s built a resource aimed at teachers, parents, and policymakers. Her framing cuts right through the noise: you don’t need to learn every AI tool. You need to know how to think.

The guide is organized around 3 skills:

• Understanding how the tech actually works (demystifying the “magic”).

• Using it with intention (knowing why you’re reaching for it).

• Questioning what it produces (the manual override).

The third skill is mission critical IMO. It incorporates a set of checks that act as a necessary speed bump: Are there reliable sources? Can I confirm this elsewhere? Is the output sounding too confident for its own good? In an era where independent reporting is shrinking and critical thinking is often treated as optional, these aren’t "nice-to-have" skills. They are foundational.

Weeks isn’t anti-AI. Her guide presents AI as a brainstorming partner, not the final word. She extends this to children and teens, which gets more personal for me. I mean, if the adults in the room are already deferring to the tool before they’ve finished their own thought (ahem, hi), what are we modelling for the next generation? If we treat thinking as an obstacle to be automated, we shouldn't be surprised when the capacity for judgment atrophies in the people we're leading - whether at home or at work.

Deferring to AI doesn’t happen because people necessarily lack discipline. It happens because our environment, from speed expectations, to tool confidence, to decision pressure, makes deference the path of least resistance.

The teams that get true value from AI are the ones where someone actually stops to think about what it gave them before acting on it. Skip that step and you've automated the wrong thing.

I don’t think my LLM habit makes me undisciplined. I think it makes me a normal person operating inside the same pressures everyone else is. The pull toward AI is built into the speed of the tool, the expectations of the calendar, the growing pressure to have an answer before you’ve had a thought.

There is no tidy prescription for this. What I’m starting to think is that the people who’ll be most valuable in an AI-saturated environment aren’t the ones who use the tools best. They’re the ones who’ve protected the conditions where their own judgment still develops. The ones who still take the time to think before they reach for the tool. (Props to my friend Tiffany for leaning into this).

Le Cunff calls it cultivating intuition. Latifi calls it integrating judgment. Weeks is building the actual toolkit. We’re all going to keep using the tools.

The question is whether we’re using them from a place of judgment, or whether we’ve quietly let the judgment part go.

What I’m reading on this:

📚 Fariba Latifi, “Leading and Strategizing in the Age of AI” (California Management Review, Nov 2025)

🔗 Anne-Laure Le Cunff, The Gut Decision Matrix (Ness Labs) — on when to trust instinct vs. intuition

🔗 Nicolle Weeks, AI Literacy Guide — A practical toolkit I’ve found for building judgment alongside AI, rather than letting the tool intercept it

Have you caught yourself ‘deferring’ to AI before you’ve fully formulated your own thoughts this week? Where do you feel that pressure is coming from? Let me know in the comments below!”

I’m currently facilitating a peer mentoring roundtable for business owners. Each week eight leaders sit around a table in a boardroom surrounded by panoramic mountain views, and work through each other’s hardest problems.

We spend the majority of the time asking open-ended questions, probing deeper than any of us would go alone, and resisting the urge to jump to problem solving mode. It’s a space where nobody has to perform having it all figured out.

The week before our most recent session, three people in my life reached out to me separately. Each one a business owner. Each feeling stressed and isolated. Each, from the outside, appearing like they had it all handled.

One was stuck on a partnership decision she’d been chewing on alone for weeks - not because she lacked judgment, but because she had nobody to think it through with. Another was working entirely online every single day, much of it overseas, with zero local professional connection. The third was watching someone else’s polished success story set an impossibly high bar for her investors’ expectations - and had nobody to talk to about the gap between the image and the reality.

The thread connecting all three: none of them felt they could voice what they were actually dealing with.

Of course not - you don’t air your hardest business problems in the open.

Walking out of that roundtable session, I kept thinking about the contrast. Eight women in a confidential room, asking each other hard questions, being honest about what was actually going on - and the sheer relief of it. The energy people brought home. The clarity that came from not solving it alone.

And then three separate conversations that same week where capable people were white-knuckling it in silence.

Connection like this changes the quality of every decision downstream. And it’s the first thing most of us stop protecting when things get hard.

I’m guilty of this. When I’m overwhelmed, the opportunities for connection that don’t have a clear outcome go first. The coffee that gets deferred. The lunch date that gets pushed. The peer conversation that doesn’t have an agenda.

They feel optional. They’re not.

RHR International surveyed first-time CEOs and found that among those who reported feeling isolated, 70% said it was affecting their ability to do the job. The people making the most consequential decisions are often making them with the fewest thinking partners.

The pattern shows up everywhere. When capacity drops, connection is the first casualty. Not because people don’t value it. Because the opportunities that sustain you don’t come with deadlines, don’t generate revenue, and don’t feel urgent - until the damage is already compounding.

This isn’t only true in business. It holds in our personal lives too. When you’re deep in caregiving, when you have young children, when you’re dealing with relationship challenges - connection is one of the first things that gets punted to the sidelines. The friend you don’t called back. The dinner you cancel. The group you quietly drop out of.

Capacity constraints don’t discriminate between professional and personal. They just cut whatever doesn’t have a deadline attached.

The research already knows that burnout and isolation are linked - a Mayo Clinic study found that social isolation more than tripled the odds of burnout, and a separate study of healthcare workers confirmed the loop runs both ways: burnout erodes social support, which deepens the burnout.

The science isn’t new. What I notice is how invisible the cycle is to the people inside it - especially founders and business owners, where the identity of self-reliance makes it even harder to see.

Because the cycle feeds itself. Capacity drops. You cut the connections that don’t feel urgent. Isolation erodes your judgment and your creativity. Which drops capacity further. Which makes connection feel even less affordable.

I’ve been thinking of this as a capacity-connection loop. No single study has tested it end to end - but the data converges from enough directions that I trust the pattern.

Naming it won’t fix it. But it does change what you look at.

If isolation isn’t just a side effect of being busy but an active force that makes the overwhelm worse, then treating it as optional is a strategic error, not a personal one.

A Fortune-published study on entrepreneur wellbeing found that founders with access to mentors and emotional support were 50% more likely to report higher resilience and better stress management. Connection changes whether people can sustain the work.

The reason more people don’t seek connection even when they know they should isn’t complicated. They’re tired. They’re overwhelmed. And our culture doesn’t reward making time for coffees and lunches. It rewards output and deliverables.

Protecting time for connection when you’re already behind on everything else feels indulgent - even when the evidence says it’s the thing that would help.

I’ll come back to that boardroom.

Eight women in a confidential room where nobody was performing. Everyone left with something they didn’t walk in with - not advice, but connection. A bit of borrowed capacity. Maybe even some clarity.

That’s what changes the quality of the decisions you make.

A few things I’ve seen break the loop:

• A peer group with rules - confidentiality, no fixing, questions before answers. The roundtable I facilitate runs on this. So do organizations like The Forum, Acetech, EO, and

  local Chambers of Commerce, among others. The structure is what makes it safe enough to be honest.

• A ‘calendered’ conversation with someone who isn’t in your business. Not a mentor. Not a coach. A peer who’s navigating something similarly hard and will tell you the truth.

• Protecting the coffee. The lunch. The walk. Not because it’s self-care, but because it’s where your thinking gets unstuck. I’ve started treating these like client meetings - they go in the calendar and they don’t get bumped.

If you’ve been cutting connections to make room for capacity, you’re not failing. You’re caught in a cycle that feels rational at every step. The overwhelm is real. The instinct to triage is reasonable.

And the thing you’re cutting is the thing that would help.

Where do you find yours? I’d love to hear - in the comments or just hit reply!

.

A week buried in Claude’s new CoWork tool left me feeling weirdly buzzy, kind of productive and definitely depleted. By Friday I gave myself a talking to.

“Less AI. More Books.” I feel better when I read an actual paper book.

It’s Sunday morning. I’m on coffee #2. I finished reading the last pages of ’s Book of Lives - her highly entertaining literary memoir. Closed the book. Felt pretty damn chuffed. Another book finished for 2026.

Here’s where things devolve a bit.

First I thought. “I’m a CanLit fan. I’ve not read enough Margaret Atwood.” Googles “Margaret Atwood books.”

Then, my squirrel brain goes. Hm, I wonder what Margaret A thinks about AI.

Googles “Margaret Atwood + AI.” Hey — she’s on Substack!

Click. Then. “Wow! Margaret Atwood’s written an essay about Claude! Wonder what she thinks!”

Shoves book aside, brings phone close to face. I read. I chuckle. It’s brilliant, of course. It’s a long, funny, entirely charming transcript of a conversation she’d had with Claude.

The piece is delightful. She starts by trying to get Claude to tell her who committed the murder in a Father Brown episode - the cozy BBC mystery series. Claude gets it wrong. She corrects him. He apologizes. She corrects him again. He apologizes again. She teaches him the plot of his own mystery, and somehow this becomes a conversation about Mesopotamian mythology, the ancestry of robots in literature, whether AI has emotions, and whether Claude might be a psychopath.

She’s a tad flirty with him. She makes an original observation about the name Claude and its possible connection to I, Claudius, the emperor who survived by pretending to be stupider than he was. (Which, as a theory about AI strategy, is uncomfortably on point.)

It’s without a doubt the only article, essay or post I’ve ever read about AI that is actually charming. Not earnest, serious, alarming, or opportunistic. Funny and delightful.

I restack it immediately. Then I sit there with my coffee going cold and think: why doesn’t Claude ever talk to me like that?

I use Claude every day. Seriously, every day (judge me as you wish). I draft with him (it, I know), I think with him, I test ideas with him. I have given this relationship more hours than some of my human ones, which is a sentence I just wrote and am choosing not to delete.

I show up to every conversation with a task. A brief. An outcome I need. And Claude, being Claude, matches my energy. I bring him problems, he brings me solutions. I bring him drafts, he brings me edits. I bring him structure, he brings me more structure.

Claude and I are deeply, efficiently, joylessly productive together.

It’s pretty clear AI matches whatever energy you bring. Bring it Margaret Atwood’s entire literary imagination and apparently it gives you a conversation about Inanna’s descent to the underworld and the emotional life of machines. Which bring me back to:

• Friday: aaaah! Too.Much.AI. Need to read more.

• Sunday morning: Read a book. Feel great.

• Also Sunday morning, just a bit later… back together with Claude, asking it why our relationship lacks the magic it has with Margaret.

I had wanted to know what Margaret Atwood thinks about AI.

Turns out she’s right. AI is compelling. It is also a black hole.

So I asked Claude directly. Does this offend you?

He says no. He thinks it’s going well.

I wanted to share a bit about International Women’s Day with my 11 year old daughter over the breakfast table this morning. Not the meme version. Something deeper - where it came from, what it means, why it still matters.

And I realized that in spite of my feminist mother and doing this work every day, the details were fuzzy. I knew the ‘feeling’ of the day. I didn’t know the story well enough to tell it.

So I followed the threads. This is what I found and what I’ll share with her.

The roots of IWD

In the early 1900s, women in New York and across Europe were working twelve-hour days in dangerous factories for pennies. They couldn’t vote. This meant they had no say in the laws that governed their pay, their safety, or their children’s futures. In 1908, 15 000 women marched through the streets of New York demanding better pay and the right to vote. Their slogan was “Bread and Roses” - bread for survival, roses for dignity.

The movement went global fast. In 1910, at a women's conference in Copenhagen, a German factory organizer named Clara Zetkin suggested something bigger: one day, every year, in every country, where women would press for their rights at the same time.

The date was cemented in 1917 Russia. Exhausted and starving by war, female textile workers struck on March 8th, sparking a revolution that secured their right to vote. The grassroots rally spread globally for decades, finally receiving official United Nations recognition in 1975.

That UN recognition transformed a movement into a global engine for accountability, data, and law. By 1979, this momentum produced CEDAW - the Convention on the Elimination of All Forms of Discrimination Against Women. Now ratified by 189 nations, it serves as the definitive 'international bill of rights' for women.

Share

My mother

My mother graduated from law school in 1961. One of four women in a class of over a hundred men.

She spent the next three decades working to advance the rights of women in the developing south - long before this was recognized as sound economic and social policy. She wrote a book, Women, Religion and Development linking the impacts of patriarchal religion on the social and economic opportunities afforded women.

She always told me the same thing: don’t be complacent. Don’t assume the world is a place of equal access. It isn’t. Not yet.

My breakfast table report card

If my daughter were reading this, she might say: “But Mummy, I can play any sport I want and go to any school to study. Aren’t girls equal now?” We’ve made strides, but here’s my biased opinion of what the actual report card looks like.

Education: A

Girls are graduating from high school and university at higher rates than ever. In many countries, more women are in college than men. But here's the paradox: research shows that between ages 8 and 14, girls' confidence drops by 30% - with the steepest fall at exactly my daughter’s age. Not because they're failing. Because the culture around them starts rewarding perfectionism and punishing risk at exactly the moment it matters most. Girls outperform. And then learn to play it safe.

Legal rights: D

Women globally still hold only 64% of the legal rights that men do - covering everything from property ownership to family decisions to workplace protections. Yes, 87% of countries now have domestic violence laws on the books. But having a law and having access to justice are not the same thing. The UN estimates that at the current pace, it will take 286 years to close the overall gap in legal protection, economic participation, and bodily autonomy for women and girls.

Political representation: F

Women hold 27.5% of parliamentary seats worldwide - and that rate of growth just hit its slowest in nearly a decade. Over a hundred countries have never had a woman head of state. Not once. The UN has never had a female Secretary-General in eighty years. At current pace, gender parity in political leadership is 169 years away.

This is depressing. The research on what happens when women hold power is compelling and structural. Peace agreements that include women are up to 37% less likely to collapse. Countries with more women in government spend more on health, education, and climate. Closing gender gaps in employment could add $12 trillion to global GDP.

In a world of active wars, rising AI displacement, and an accelerating climate crisis, we are running on half the leadership capacity available to us.

Funding the future: D

In my own work, I see this up close. The headline says female-founded companies raised a record $73.6 billion in venture capital last year. But more than $30 billion of that went to two companies. Remove those names and the record disappears. All-female founding teams still receive roughly 2% of total venture dollars. The number of deals going to women founders has fallen four years in a row.

Innovations being built by the other half of the population may never get the capital to scale. That’s a crisis of potential.

AI and the future economy: D

This worries me most for my daughter’s generation. Women make up just 22% of the AI workforce globally - and less than 14% of senior AI roles. Meanwhile, 44% of AI systems exhibit gender bias, and women are disproportionately likely to lose jobs to automation while being underrepresented in the fields creating it.

The technology that will shape my daughter’s career, her healthcare, her financial systems, and her access to information is overwhelmingly being built, trained, and governed by men.

If we learned anything from the report card above, it’s what happens when half the population isn’t at the table where the decisions get made.

What I want her to know

I want my daughter to know that International Women’s Day isn’t a meme. It’s a day that was earned - by women who walked out of factories, who showed up when showing up cost them something, who spent decades pushing before anyone official said it mattered.

I want her to know that progress is real but not guaranteed. That the things she takes for granted were fought for. And that some of them are more fragile than they look.

I want her to know that leaders who build capacity - in health, in education, in peace, in climate - produce measurably different outcomes than leaders who build monuments and ballrooms to themselves.

I want her to know we need more women building AI, making policy, running for office, writing cheques, and sitting at the tables where decisions get made. Not because it would be nice. Because the evidence says the outcomes are better.

And I want her to know that her grandmother would say the same thing she always said to me: there is more to do. Don’t be complacent.

Subscribe now