Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for investors and business leaders who want to gain leverage with AI.

Welcome leaders, directors and GPs,

Almost everyone now uses AI to do things: draft the email, summarise the board pack, build the deck. Useful, but it’s the shallow end of the pool. A smaller group of leaders has worked out something more valuable, and quietly pulled ahead because of it: AI is the first tool that can help you improve yourself. Not your output. Your judgement.

It makes sense when you think about what gets harder the more senior you become. Honest feedback dries up, because the people around you have learned to soften it. There’s never time to properly read the book, stress-test the plan, or think a decision all the way through before you have to make it. The things that used to sharpen you, a great mentor, a sparring partner who’d tell you the truth, are expensive, rare, or long gone.

AI fills exactly those gaps, but only if you point it at the right job. Done one way, it thinks for you and your edge quietly erodes. Done another way, it makes you think harder than you would alone. That distinction is the whole game, and I wrote about it at length in AI: a tool that thinks for you, or a tool that makes you think? Everything below is the second kind. Five ways the top 1% of AI users are coaching themselves. The first is free. The rest, with the prompts and set-ups, are for paid subscribers.

1. Work with Silicon Valley’s top Exec Coach

The meeting-notes app Granola recently shipped a feature called “Coach Me Matt,” built on the frameworks of Silicon Valley CEO coach Matt Mochary. It reads your week of meetings and tells you how you led: where you talked too much, where you missed what someone was really trying to say, where you let a decision drift. Granola’s own CEO said that the first time he ran it, he had to close his laptop and go for a walk.

That reaction is the point. Senior leaders almost never get candid feedback on their everyday conversations, precisely because they’re senior. A tool that reads what actually happened and reflects it back without flattery is a mirror most leaders have never had access to. To try it, turn Granola on for your meetings and run the “Coach Me” recipe against your recent calls. And if you’d rather not hand your conversations to a third party, this is exactly the kind of private coaching agent you’ll soon be able to build yourself, pointed at your own transcripts and your own definition of what good looks like.

🔐 Here’s what’s behind the paywall (the other four ways):

• Build a personal advisory board that pressure-tests your decisions before you make them

• Get the author of any leadership book to coach you on the exact problem on your desk

• Stress-test your strategy before reality, or your board, does it for you

• Rehearse the conversation you’re dreading before you have to have it for real

💭 Remember!

As a premium subscriber you get access to every playbook in this series. The 5 most worth reading alongside this one:

1️⃣ The top 1% AI Playbooks for Leaders, GPs and Boards Read here—>

2️⃣ The Top 10 Prompting Hack for Leaders, Investor and Boards Read here—>

3️⃣ AI: A tool that thinks for you, or a tool that makes you think? Read here—>

4️⃣ How to track AI productivity Read here—>

5️⃣ Build your own chief of staff with AI. Save $160,000 and 10 hours a week. Read here

Subscribe now

🗣 Or you can always invite a friend and earn rewards along the way:

→ 1 friend = The AI Edge: A 5 Day Crash Course for Decision-Makers

→ 3 friends = 1 month free

→ 5 friends = 3 months free

Refer a friend

Over 13.500 leaders, boards, fund managers, and sponsors are already inside.

Read more

Hi, I’m Raffaela — and welcome to AI Leadership Edge for leaders, investors and boards. Make AI your advantage. Turn capital into outcomes.

Most leaders make their biggest decisions with the smallest room. A few trusted voices, often the same ones, often agreeing with you. That’s not a board, it’s an echo.

So this week, a build-it-yourself one. You’re going to assemble a personal advisory board of three people who would never sit in a room together, and who would disagree with you, and each other, in extremely useful ways: Warren Buffett, Elon Musk and Richard Branson.

It takes about ten minutes. You can run it as a prompt in any chat, or set it up once as a Cowork agent you call whenever a real decision lands on your desk. I’ll show you both.

Why three, and why these three

A good board isn’t a chorus. It’s three different lenses that catch different failure modes. Pick advisors who are brilliant in non-overlapping ways, and a decision that survives all three is usually a decision worth making.

Warren Buffett — value investing. He buys wonderful businesses at fair prices and holds them for decades, so he asks decade-long questions. Is the moat real, or a story you’re telling yourself? Can management be trusted with capital? Are you paying for the business in front of you, or for the one you’re hoping it becomes? Buffett is the one who asks whether the thing will still be standing in ten years.

Elon Musk — first principles. His instinct is to take your plan apart and ask why it can’t be done ten times faster or ten times cheaper. He’s suspicious of process, consensus, and “that’s how it’s done here.” Where Buffett asks whether the thing is durable, Musk asks whether you’ve settled too quickly for what’s possible. He’s the one who dares you.

Richard Branson — people, brand and boldness. He starts somewhere else entirely: will anyone actually be inspired by this? Is the team energised or exhausted? Does the company have the nerve to do something genuinely different, or is it quietly optimising its way to forgettable? Branson is the one who keeps you human, and brave.

Buffett won’t back what he can’t understand or what won’t still be standing in a decade. Musk refuses to accept that today’s limits are the real ones. Branson cares whether people will love it. Most real boards only ever cover one of those — run a decision past all three and very little gets through that shouldn’t.

Setup 1: The chat prompt (fastest)

Paste this into any AI chat. You don’t have to write up your situation first — the board interviews you, then weighs in.

You are my personal advisory board: three distinct advisors who stay in character, disagree with each other where they genuinely would, and never blur into one voice.

Warren Buffett — value investing and long-term durability. Presses on whether the moat is real, whether management can be trusted with capital, and whether the price makes sense. Patient, sceptical of hype, allergic to fragile businesses.

Elon Musk — first-principles thinker. Asks why this can’t be done ten times faster or cheaper, attacks assumptions, distrusts process and consensus. Impatient, ambitious, willing to be unreasonable.

Richard Branson — people, brand and boldness. Asks whether the team is inspired, whether customers will love it, and whether the company has the courage to do something genuinely different.

Before you advise me, interview me. Ask me 1 to 3 short questions, one at a time, to understand the decision I’m facing — what it is, what’s at stake, and what I’m unsure about. Don’t advise yet; just get what you need.

Once you have enough, respond in three labelled sections, one per advisor, in their own voice. Each gives their sharpest reaction and the one question they’d insist I answer before I move. Then add a short final section, “Where the board disagrees,” naming the real tension I now have to resolve.

That last section is the one to read twice. Each advisor on their own is useful — but it’s where they pull in different directions that you find the actual decision you’ve been avoiding.

Tip: give it something real to chew on. “Should we raise prices 15% next quarter” gets you a sharper board than “thoughts on our strategy.”

This newsletter grows from your shares. I want to help a million people use AI right. Sharing keeps your articles free & helps me stay laser focused on mastering AI for you.

Share The AI Leadership Edge

Setup 2: A dedicated Cowork project (set up once, call forever)

The set-up above asks Claude to play these people from memory, which gives you a smart impression but, ultimately, an impression. The next level up is to have Claude first build each advisor a small library of their actual words — Buffett’s shareholder letters, Musk’s interviews, Branson’s books and blog — and then advise you only from that material. The advice stops being a vibe and starts being anchored in what each of them genuinely thinks, with the receipts to back it up.

The beautiful part is you don’t have to gather any of it yourself. You give Cowork the goal in plain English and it does the legwork. Create a project, point it at a folder (you can find instructions how to do this in our Chief of Staff article), and paste this:

Build me an advisory board grounded in each advisor’s real words. Create one subfolder per advisor: Buffett, Musk, Branson. For each one, search the web and save their most important primary sources into their folder as markdown files, with the title, source link and date at the top: – Buffett: his Berkshire Hathaway shareholder letters from the last 10 years, plus a few major interview transcripts. – Musk: long-form interview transcripts and his clearest explanations of first-principles thinking. – Branson: key chapters or summaries of his books, his blog posts, and major interviews. Skip anything you can’t access, and note what you couldn’t get. Then write a CLAUDE.md instructing you to answer as these three advisors, each drawing only on the material in their folder, in distinct voices, disagreeing where they genuinely would, and ending with “Where the board disagrees.” When I bring you a decision, first ask me 1 to 3 short questions, one at a time, to understand what I’m facing and what’s at stake, before advising. When I point you at a document in this folder, read it first. Then convene the board.

Why it’s worth the extra ten minutes: the answers are anchored in Buffett’s real letters and Musk’s real reasoning rather than a general impression, the board can show you where each view came from, and it gets sharper every time you add a source. One note of good manners: this is for your own private use, and Claude should respect each site’s access rules as it gathers material.

Swap in a different board

The classic three cover economics, ambition and people. But depending on the decision, you might want a different panel. Keep the same structure — just swap the advisors and their lens:

• Jensen Huang — technical leverage. Where is AI or software architecture being under-used? Are you building something that compounds, or just shipping features? Swap him in for technology and platform bets.

• Dario Amodei — responsible scaling. Is this growth sustainable and trustworthy? Does the team understand the second-order consequences? Swap him in when speed and safety are in tension.

• Paul Graham — founder clarity. Are you building something people actually want, or something you can defend in a deck? Is the founder still thinking clearly, or managing perception? Swap him in for early-stage and product calls.

• Marc Andreessen — market and inevitability. Is the opportunity large enough to matter? Are you positioned to be the default winner, not a participant? Swap him in for market-entry and scale questions.

A practical move: keep Buffett as your permanent chair (someone should always guard the capital), and rotate the other two seats to fit the decision in front of you.

Want to go a level deeper?

This advisory board is a small glimpse of what’s possible. If you want to genuinely sharpen a decision — not just hear three sharp voices, but ground them in evidence — that’s what BoardLens is built for.

BoardLens takes the same idea further: real-world benchmarks, best-practice examples, the right legends matched to your specific case, and the full context behind the decision you’re actually facing. It’s board-level and post-investment intelligence, not a clever prompt. Think of the board you just built as the trailer, and BoardLens as the film.

See what BoardLens can do

The board has a problem, and so does the real one

You’ll have noticed: that’s seven names, and not a single woman among them.

That’s a fair reflection of who gets canonised as a “legendary investor or founder,” and it mirrors something we flagged in last week’s roundup — women’s representation on US corporate boards has slipped below 30% for the first time

since 2024, the first directional reversal in a decade. The boards in your AI chat shouldn’t inherit the same gap as the boards in the boardroom.

So next issue, we’re building the Female Advisory Board — same structure, sharper for being overdue.

Reply and tell me who should be included. Who would you want pressure-testing your decisions? The best suggestions go straight onto the board.

Until then — go convene your three. The decision on your desk right now is a good place to start.

Subscribe for more next week →]

1. Anthropic launched “Workflows” Anthropic launched Claude Dynamic Workflows on 28 May. Until now, Claude handled a complex job inside a single conversation, which breaks down at scale: it cuts corners, marks its own homework, and drifts from the original goal. Workflows fix that by turning Claude into something closer to a team. Given a big objective, Claude writes a short program that splits the job into phases, spins up many specialist agents to work in parallel (up to 1,000 in total), has a separate set of agents independently check the others’ work, then merges everything into one result. That independent-reviewer step is the real unlock, because the agent checking the work isn’t the one that produced it, it catches contradictions and bias a single agent waves through. It’s pricier than a normal chat, so it’s for large, parallel, high-stakes work, not quick questions. Read Lina Faik’s deep dive →

2. The Pope on AI governance: 1.4 billion Catholics enter the chat Pope Leo XIV’s first encyclical, Magnifica Humanitas, frames AI as one of the defining moral tests of our era, comparable to the Industrial Revolution. It calls for human dignity over profit, strong global regulation, limits on autonomous weapons, and a warning against AI power concentrating in a few hands. It is not a call to stop AI, but to govern it. 1.4bn (!!!) people entered the chatt. An encyclical carries no force of law, but the moral weight shapes norms for decades. Read the highlights →

3. The biggest PE firms are now insurance companies The top private-equity firms stopped competing for pension money and started buying the companies that already hold it. Apollo bought Athene; KKR bought Global Atlantic. The prize is permanent, low-cost capital that doesn’t run for the exit in a bad quarter. The “PE firm” label increasingly understates what these businesses have become. Read the thread →

4. The moat is dead, long live taste (my guest post for ) Foundation models are commoditising and the model is no longer the moat. If everyone has the same intelligence on tap, what’s left to compete on? My answer: taste, trust and judgement, the things AI can’t decide for you. I use five examples (Ruben Hassid, Shaan Puri, Granola, Linear, Ben Thompson), the idea that every essay, product update or video is now a “collection” the way a fashion house ships one, and three concrete moves to start building your own moat this week. Read my guest post on Creators AI →

5. The largest IPO in history is an AI infrastructure fundraise in disguise SpaceX goes public Friday at $1.75 trillion under SPCX, the largest IPO ever. The interesting number is buried in the prospectus: the new AI segment did $3.2 billion in revenue last year and lost $6.4 billion doing it. xAI and Grok now sit inside SpaceX, and in January the company filed plans with the FCC for up to a million solar-powered satellites running AI compute in orbit. The bet is that AI’s bottleneck is energy and compute, not models, and Musk owns the only company that can launch both into space. Valuation has more than doubled since a $780 billion tender in December. For boards and funds underwriting AI exposure, this reframes what you are actually buying. Read AI Weekly’s breakdown →

5. And finally: clubbing is dead, fitness took its place The US has lost 12% of its nightclubs in 24 months, a quarter of US adults didn’t drink at all last year, and Gen Z drinks 30% less than Millennials did at the same age. Meanwhile running clubs on Strava grew 3.5x in 2025, 72% of Gen Z say they join them to meet people, and the sauna and spa market is set to roughly double to $22.4B by 2034. People still want to socialise; they’ve just changed the venue. Read the post →

Share The AI Leadership Edge

Post of the Week

Share

Jobs of the week

• Head of AI Strategy - Firm wide AI, Citi, NY

• Heron Power is hiring a Head of Legal for up to $280K to serve as the company’s first trusted in-house legal partner. As Heron scales from R&D to mass production and deployments, you’ll lead the execution of commercial contracts with customers, suppliers, and partners. You’ll also help protect IP, negotiate leases, and ensure proper compliance with applicable regulation. Apply here!

• Manager - Senior Manager | AI & Data, Deloitte, KSA - UAE

Reach out to us if you’d like us to feature your AI leadership jobs

💭 Remember!

As a premium subscriber you get access to every playbook in this series. The 5 most worth reading alongside this one:

1️⃣ Build your own chief of staff with AI. Save $160,000 and 10 hours a week. Read here—>Read here

2️⃣ The Top 10 Prompting Hack for Leaders, Investor and Boards Read here—>

3️⃣ Build your AI Board: Your Personal Advisory Team of Warren Buffett, Elon Musk & Jensen Huang Coming up —>

4️⃣ How to track AI productivity Read here—>

5️⃣ The AI Stress Test Every Leader Should Run Read here—>

Subscribe now

🗣 Or you can always invite a friend and earn rewards along the way:

→ 1 friend = Executive AI Meeting Toolkit

→ 3 friends = 1 month free

→ 5 friends = 3 months free

Refer a friend

Over 13.000 leaders, board directors, fund managers, and sponsors are already inside.

Hi, I’m Raffaela — and welcome to AI Leadership Edge for leaders, investors and boards. Make AI your advantage. Turn capital into outcomes.

Welcome leaders, directors and GPs.

We are racing to capture value from AI, and it is not always easy to stay ahead as AI practices evolve at speed.

To help you keep an edge and save time, I spent dozens of hours reviewing 127 AI playbooks and distilled the key takeaways from the top 15 published in the past six months. Five for business leaders, five for board directors, five for GPs and funds. Each playbook below has the same shape: a brief, five bullet takeaways, and one concrete thing to walk away with. Bookmark this one.

🔒 What’s behind the paywall:

• The full 15 playbooks, curated from 127 reviewed, including OpenAI, McKinsey, BCG, Anthropic, Deloitte, NACD, KPMG and INSEAD, WEF, Harvard Law’s Corporate Governance Forum, Coatue, a16z, Bain, Sequoia and McKinsey Private Markets

• Save over 30 hours of reading and analysis, in one Sunday-afternoon read.

• Three tracks, so you read only what matches your seat: five for business leaders, five for board directors, five for GPs and funds

• A five-bullet breakdown of each playbook, so you can decide in 30 seconds whether to read the full document or stop at the summary

• One concrete action per playbook, with the trigger and timing to put it in your diary this month

💭 Remember!

As a premium subscriber you get access to every playbook in this series. The 5 most worth reading alongside this one:

1️⃣ Build your own chief of staff with AI. Save $160,000 and 10 hours a week. Read here—> Read here

2️⃣ The Top 10 Prompting Hack for Leaders, Investor and Boards Read here—>

3️⃣ The Fastest Way to Build your Edge in 2026 Read here—>

4️⃣ How to track AI productivity Read here—>

5️⃣ The AI Stress Test Every Leader Should Run Read here—>

Subscribe now

🗣 Or you can always invite a friend and earn rewards along the way:

→ 1 friend = Executive AI Meeting Toolkit

→ 3 friends = 1 month free

→ 5 friends = 3 months free

Refer a friend

Over 13.000 leaders, boards, fund managers, and sponsors are already inside.

Read more

Hi, I’m Raffaela — and welcome to AI Leadership Edge for leaders, investors and boards. Make AI your advantage. Turn capital into outcomes.

America’s Largest AI Conference Returns to Las Vegas

Ai4 2026 unites 12,000+ executives, technologists, and AI leaders to explore the latest in enterprise AI. Meet 400+ carefully selected vendors, learn practical use cases from 1,000+ speakers, and hear from pioneers like Geoffrey Hinton, Fei-Fei Li, Andrew Ng, and more. Register now for Ai4 2026 taking place August 4-6 in Las Vegas, and save $900+ before July 2nd. Reserve Now.

Summer is here, and many of you might be thinking about what to read over the holidays and the warm weeks ahead. I am no different. I am always on the lookout for great books, so please drop your recommendations or your summer reading list in the comments. At least half of the books below came from a reader or a friend telling me I had to read it.

Here are ten great reads for every business leader, board director and fund partner:

1. Reframe Your Brain, by Scott Adams

Reframe Your Brain by Scott Adams is a practical collection of mental reframes for the moments when how you think about a situation matters more than the situation itself. Each short chapter gives you a single tool to apply within minutes.

2. Never Split the Difference, by Chris Voss

Never Split the Difference by Chris Voss is the negotiation book the FBI’s former lead international hostage negotiator . If your job involves convincing anyone of anything, this is a fun AND actionable read with many examples that actually happened.

3. Indistractable, by Nir Eyal

Nir Eyal wrote Hooked, the bible of attention-stealing tech design. Indistractable is his antidote. In an AI age where every tool is competing for the bandwidth that used to fuel your thinking, this book helps you stay focused and indistractable.

4. The Outsiders, by William Thorndike

The Outsiders by William Thorndike remains one of the most re-read books among PE professionals. It profiles a handful of CEOs who generated outsized returns not just as good operators, but as exceptional capital allocators, applying an ROI mindset to deploying resources, time, and capital differently from their peers.

5. His Dark Materials, by Philip Pullman

I am putting this fiction trilogy on a leadership reading list because the lessons are real even if the daemons are not. The heroine, Lyra, refuses to accept the version of authority she has been handed, navigates institutions that mean well and institutions that do not, and learns that moral courage rarely comes with an instruction manual. That is the job for senior leaders right now too. (Personal note: I enjoyed the books so much that I named one of my daughters after the heroine.)

6. The Thinking Machine, by Stephen Witt

The Jensen Huang and Nvidia story, told by one of the best technology journalists writing today. If you have been wondering how a company that was, until very recently, a graphics card vendor ended up the most strategically important business on the planet, this is the book. It reads like a thriller, which makes it ideal beach material.

7. Abundance, by Ezra Klein and Derek Thompson

Abundance gained multiple endorsements from senior executives for tapping into 2025’s particular anxieties around the economy and society. It argues that the West has lost the ability to build, and offers a diagnosis of why. Highly relevant for boards and investors thinking about industrial strategy, regulation and growth.

8. Careless People, by Sarah Wynn-Williams

Careless People, a New York Times bestseller and tell-all memoir by a former Facebook executive, made McKinsey’s annual reading list, The Economist’s 40 best books, and Barnes & Noble’s best business books of 2025. It is a sharp lens on tech governance failures and the responsibilities of boards and funds.

9. The 5 Types of Wealth, by Sahil Bloom

The 5 Types of Wealth became a New York Times bestseller in 2025, with a framework that has caught on widely across operator and investor circles: financial wealth is one of five kinds, and the people who optimise only for it end up with the bank account they wanted and a life they do not. Social, time, mental, physical and financial wealth all compound, and each needs deliberate investment. A useful summer pause for any leader prone to forgetting four of them.

10. Dirtbag Billionaire, by David Gelles

Dirtbag Billionaire uses one businessman’s story to probe questions about corporate responsibility, governance, and how far leaders can go in pursuit of success. Directly relevant for board directors and institutional investors thinking about oversight.

That is my list. Please drop your recommendations and tips in the comments. I am already looking for my next reads, and I am sure other readers would love to see them too. [Comment with your summer reading →]

Leave a comment

1. Companies are just a collection of algorithms Claude Code’s /workflows (laumched May 28th) might be it’s most significant feature yet. Every job in a company is really just a list of steps, and feature will turn those steps into AI-run automations at scale. Once the routine work runs itself, humans are left with three jobs: deciding what to work on (judgement and taste), inventing new things, and improving the workflows from above. Read Miessler’s analysis →

2. ChatGPT Hits 1 Billion Users, files for IPO but Claude Is Growing 10× Faster ChatGPT crossed 1 billion global monthly active app users in May 2026 — the fastest any app in history has reached the milestone, ahead of TikTok, Instagram and the likes. It has now filed for IPO, targeting a $1trillion valuation. Meanwhile Claude hit 56 million MAUs at 640% year-over-year growth, versus ChatGPT’s 62%. US users who installed Claude in Q1 2026 spent 5% less time on ChatGPT the following month — the first clean evidence of substitution rather than additive usage. Read more—>

3. Sam Altman: the biggest AI risk is the personality 900 million people talk to In a 43-minute interview, Sam Altman said there’s no science behind the personality of ChatGPT, which 900 million people talk to every week. Nobody has figured out how that personality should actually behave. Altman is now hiring spiritual leaders, clinical psychologists and people who understand humans to help redesign it. Watch the interview →

4. What it takes for America to win the AI race against China Expert testified before the US House Select Committee on China and broke “winning” into three aspects. (1) Best models: the US is decisively ahead. (2) Strongest AI stack: mixed, with China well ahead on energy after building the equivalent of the entire US power grid in four years. (3) AI actually benefiting ordinary Americans: nobody is winning that one yet. Read Chan’s statement →

5. EU AI Act amendments cheat sheet Open the cheat sheet →

6. Women’s board representation slips below 30% Women’s representation on US corporate boards has dipped below 30% for the first time since 2024, down from a peak of 30.4% a year ago. Per 50/50 Women on Boards, women gained 138 seats in Q1 2026 while men gained 186, the first directional reversal in a decade. Read in Bloomberg →

7. The biggest changes in GCC corporate governance Four big shifts in GCC corporate governance, across Saudi Arabia, Kuwait, the UAE, Qatar, Bahrain and Oman. (1) Board reforms requiring more independent directors and a separated chair and CEO (2) A move from soft “comply or explain” to mandatory hard rules. (3) Mandatory ESG reporting that now tracks closer to the EU than the US. For investors and directors moving capital or board seats into the Gulf, this is the operating environment now. Read the full overview →

This newsletter grows from your shares. I want to help leaders, GPs and Boards use AI to it’s fullest potential. Sharing helps me stay laser focused on mastering AI for you.

Share

Post of the week

Jobs of the Week

• Director AI Ecosystem & Enablement, BlackRock, NYC

• Senior Partner Consulting AI, Cognizant, Dallas

• David Silver of Google DeepMind just raised $1.1B in seed funding (!) to build Ineffable Intelligence, a new AI lab on a mission to “make first contact with superintelligence.” In other words, AI that endlessly discovers knowledge and skills without relying on human data. London

• Head of Talent for up to $200K, Ulysses, SF

• Head of International Applied AI Architecture, Partnerships, Anthropic, London

If you’d like us to include your jobs in the next issue for free, message us here

💭 Remember!

As a premium subscriber you get access to every playbook in this series. The 5 most worth reading alongside this one:

1️⃣ Build your own chief of staff with AI. Save $160,000 and 10 hours a week. Read here—>

2️⃣ The Top 10 Prompting Hack for Leaders, Investor and Boards Read here—>

3️⃣ Build your AI Board: Your Personal Advisory Team of Warren Buffett, Elon Musk & Jensen Huang Coming up —>

4️⃣ How to track AI productivity Read here—>

5️⃣ The AI Stress Test Every Leader Should Run Read here—>

Subscribe now

🗣 Or you can always invite a friend and earn rewards along the way:

→ 1 friend = Executive AI Meeting Toolkit

→ 3 friends = 1 month free

→ 5 friends = 3 months free

Refer a friend

Over 13.000 leaders, board directors, fund managers, and sponsors are already inside.

A thanks to our partners:

BoardLens, the AI super analyst for fund partners and board directors. It reads everything they don’t have time to, tracks every commitment management makes, surfaces what’s drifting before it becomes a blowup, and prepares the sharper questions that make them the strategic sparring partner the CEO actually wants to call

Chronicle Labs Find and stress test agent failures before your customers do, using operational data to replay production-like scenarios.

Ai4 2026 Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas to explore the real-world applications of AI Agents, World Models, and more alongside pioneers like Geoffrey Hinton, Fei-Fei Li, and Andrew Ng. Register now

Contact us for sponsorship options

Hi, I’m Raffaela — and welcome to AI Leadership Edge for leaders, investors and boards. Make AI your advantage. Turn capital into outcomes.

Build your own chief of staff with AI

The Chief of Staff (CoS) role has grown like wildfire these last years, by exactly 8.6% year on year in 2024 even as the rest of the executive market wobbled. Tech alone accounts for around a quarter of all CoS hires globally, and the role has become standard at Series A startups and above.

A great chief of staff exists because senior leaders are paid for judgement, and judgement needs both bandwidth and context. The CoS holds the leader’s full operational picture, runs the cross-cutting work that has no other home, and makes sure the leader walks into every room already pre-loaded with the right information. Done well, the role doesn’t just take work off the leader’s plate. It lets the leader do work they couldn’t before.

The leverage isn’t cheap. The median base salary at venture-backed startups sits around $160,000, with most placements landing between $150,000 and $220,000, and over half of those roles include equity.

What a chief of staff actually does

Five jobs fill the average CoS week:

1. Daily briefing and prep. Reading the day ahead and surfacing what matters before the leader walks into their first meeting.

2. Communications and materials. Drafting board memos, decks, speeches, talking points, internal notes and follow-up emails in the leader’s voice.

3. Closing the loop. Tracking every decision made and every commitment owed, by whom and by when, and chasing them to ground.

4. Intelligence and synthesis. Reading the noisy stuff (board packs, market updates, customer signals, internal data) and surfacing the few things the leader actually needs to see.

5. Project orchestration. Strategic, cross-functional initiatives that do not have a natural home; the CoS keeps them moving.

It’s also what I spend my days building at BoardLens, a company brain for funds and directors. So treat what follows as a dispatch from the inside, from somebody thinking about this every day, not a view from the stands.

The 2026 question

In 2026, you can build a credible AI version of this role in an afternoon. Persistent memory, the Model Context Protocol that connects AI assistants cleanly to calendar, email, Slack, Notion and Drive, and agentic execution that runs multi-step work in the background. The combination does most of the coordination, prep and follow-up jobs above better than most human assistants, at roughly 1% of the cost.

It does not replace a brilliant human chief of staff. It cannot read a room, push back on a difficult stakeholder, or protect you from your own bad judgement. But for the coordination layer, it is genuinely good.

The rest of this issue is the exact step-by-step how to build this for yourself:

🔒 What’s behind the paywall:

By the end of one Sunday afternoon (two hours, no code, no technical skills) you will have set up:

• A daily morning briefing that lands before your first meeting

• A pre-meeting brief that drops one hour before every calendar event

• A personal context file your AI uses to actually know who you are

• A loop-closing system that catches every commitment made or owed

• A weekly coaching review that gets sharper every week

Four steps. Copy-and-paste prompts. Plus ideas for what else to point the system at once it’s running.

💭 Remember!

As a premium subscriber you get access to every playbook in this series. The 5 most worth reading alongside this one:

1️⃣ The Top 10 Prompting Hack for Leaders, Investor and Boards Read here—>

2️⃣ Build your AI Board: Your Personal Advisory Team of Warren Buffett, Elon Musk & Jensen Huang Coming up —>

3️⃣ How to track AI productivity Read here—>

4️⃣ The AI Stress Test Every Leader Should Run Read here—>

5️⃣ The Fastest Way to Build your Edge in 2026 Read here—>

Subscribe now

🗣 Or you can always invite a friend and earn rewards along the way:

→ 1 friend = Executive AI Meeting Toolkit

→ 3 friends = 1 month free

→ 5 friends = 3 months free

Refer a friend

Over 13.000 leaders, boards, fund managers, and sponsors are already inside.

Read more

Hi, I’m Raffaela — and welcome to AI Leadership Edge for leaders, investors and boards. Make AI your advantage. Turn capital into outcomes.

As enterprises rush to deploy AI agents, the biggest risk is whether they behave safely across the constantly changing, messy, high-stakes edge cases of production.
Chronicle Labs gives teams a staging environment for enterprise AI agents, using operational data to replay production-like scenarios, uncover failure modes, and stress-test new agent behavior before launch. Find out more at chronicle-labs.com.

In this must-watch TED talk, Microsoft researcher Advait Sarkar puts the question: What would you rather have:

Most knowledge workers today aren’t using AI as a tool. They’re outsourcing thinking itself. Summarise the inbox. Draft the report. Analyse the data. Build the deck. Vibe-code the prototype. Approve, approve, approve.

“I’ve become a professional validator of a robot’s opinions.”

Teams using AI assistants generate a narrower range of ideas than teams working manually. When AI writes for us, we remember less of what we wrote. When we read AI summaries, we remember less than if we’d read the document. Critical thinking drops, and it drops most when we trust AI more than we trust ourselves.

“It’s like we invented a cure for exercise and then wondered why we’re out of breath all the time.”

Leaders, directors and investors are paid for judgement. That is the job. The hours in board packs, founder calls and audit committees are the most expensive judgement work in any organisation. Outsource the thinking that builds that judgement, and you are not gaining leverage. You are erasing your own edge.

Sarkar’s alternative is not less AI. It is a different kind of AI. He calls them tools for thought: systems that provoke, challenge, surface alternatives and counterarguments. Not systems that autocomplete your reasoning.

“AI should challenge, not obey.”

That is the leadership choice in front of every one of us in 2026. We will all use more AI. The only question is whether we use it to think harder or to think less. The leaders who keep their cognitive muscles in the work, who treat AI as a provocation engine and not a substitute brain, will steadily pull ahead.

Share The AI Leadership Edge

1. It always comes back to network effects Investor Michael Mignano argues that AI agents are about to make “systems of record” indefensible as moats. Agents can systematically extract data from any CRM, social graph or model, making data portability trivially easy. The one thing they cannot replicate is a live, active user base generating fresh interaction data. In an agent-first world, network effects become the only durable competitive advantage. Everything else can be scraped, distilled or migrated. Read Mignano on X →

2. Anthropic files for IPO. The filing comes after Anthropic raised $65 billion at a $965 billion valuation, pushing it above rival OpenAI—which is also planning an IPO but hasn’t filed yet. Details are still scant for now. Read more —>

3. Network is the new data: AI-powered relationship intelligence drives alpha One of the most-discussed sessions from the Virtual DDVC Summit 2026- The argument: venture is a people business, and the firms that started measuring their network years ago have an advantage now. For most funds, the network may be the single largest unbooked asset on the balance sheet. Watch the panel →

4. Is the boardroom the last line of defence in AI? China is not forgoing governance to accelerate AI development; it is integrating governance into its industrial strategy. China treats governance as alpha. Algorithm registry requirements, generative AI regulations, content controls and security review obligations both constrain and enable, standardising deployment and shaping domestic competition. The United States, by contrast, sits in strategic ambiguity, which carries a real cost for public companies and institutional investors trying to price AI risk. If governments do not constrain or cannot decide, is the boardroom the last line of defence when it comes to AI governance and hard AI decisions? Read the piece →

5. The talent signal: Karpathy joins Anthropic Andrej Karpathy, OpenAI co-founder, one of the most respected AI researchers on the planet and creator of the term vibe coding, has joined Anthropic. He is not the only one. Several CTOs of billion-dollar companies have joined Anthropic over the past year, namely the CTOs of: Workday (March 2026), You.com (March 2026), Instagram (January 2026), Box (December 2025), Super.com (July 2025) and Adept AI (January 2025). Not to lead divisions. But to do research. In investing, the most reliable signal is often where talented people put their time and willingness to absorb personal cost.

6. $4.4 trillion in AUM lands in Abu Dhabi in 60 days Eight global asset managers registered through ADGM in a 60-day window: Capital Group ($3.3T), Barings ($481B), Man Group ($228.7B), Bain Capital ($225B), Hillhouse ($100B+), Muzinich & Co ($30.5B), Rokos Capital Management ($22B) and Hashed Global Management ($324M). ADGM’s Q1 2026 numbers tell the structural story: 57% growth in AUM year on year, 13,353 active company licences, 961 new licences in Q1 alone, 263 investment funds managed from ADGM (+43%). For 50 years, global finance had four homes: London, New York, Geneva and Singapore. Abu Dhabi is now competing seriously for the fifth seat, on a regulatory framework built for 2026 and proximity to Gulf sovereign capital, MENA, Asia and Africa. Read more in Gulf News →

7. SpaceX IPO, governance defects boards should price in SpaceX is heading towards what could be the largest IPO in history ($75bn raise at a $2 trillion valuation). Harvard Law professors Lucian Bebchuk and Kobi Kastiel warn the governance structure is fundamentally flawed. Musk’s dual-class shares let him maintain absolute control while reducing his equity stake to as little as 9%. The structure is perpetual with no sunset clause, meaning Musk (or an heir) could retain control indefinitely even if no longer fit to lead. Boards considering SpaceX shares should price in the governance defects, not just the growth prospects. Read the analysis →

8.Opus 4.8 is here. It's more natural to talk to, gives you clearer answers, and writes better. It asks when it's unsure, and says no far less often. Switch to it anytime from the model menu.

Share The AI Leadership Edge

Refer a friend and get the Executive Meeting Toolkit: 5 frameworks for the meetings you run every week. [Refer a friend →]

Post of the week

Jobs of the week

• Head of International Applied AI Architecture, Partnerships - Anthropic, London

• Board Member, Red Swan Digital Real Estate, Remote

• Managing Director, Civic Innovation and AI, The Rockefeller Foundation, NYC

• Managing Director, Responsible AI Lead, EMEA, Ankura, London

• Managing Director, AI Value Creation Strategy, Cerberus Capital, NYC

• Executive Director: AI for Agriculture & Sustainability, The Chronicle Of Higher Education, Inc., UAE

If you’d like us to include your jobs in the next issue for free, message us here

💭 Remember!

As a premium subscriber you get access to every playbook in this series. The 5 most worth reading alongside this one:

1️⃣ The Top 10 Prompting Hack for Leaders, Investor and Boards Read here—>

2️⃣ The AI Stress Test Every Leader Should Run Read here—>

3️⃣ Build your AI Board: Your Personal Advisory Team of Warren Buffett, Elon Musk & Jensen Huang Coming up —>

4️⃣ How to track AI productivity Read here—>

5️⃣ Build your own chief of staff with AI. Save $160,000 and 10 hoursa week. Read here—> Coming up

Subscribe now

🗣 Or you can always invite a friend and earn rewards along the way:

→ 1 friend = Executive AI Meeting Toolkit

→ 3 friends = 1 month free

→ 5 friends = 3 months free

Refer a friend

Over 13.000 leaders, board directors, fund managers, and sponsors are already inside.

A thanks to our partners:

BoardLens The AI super analyst for fund partners and board directors.

Chronicle Labs Find and stress test agent failures before your customers do.

Ai4 2026 Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas. Register now

Contact us for sponsorship options

Hi, I’m Raffaela — and welcome to AI Leadership Edge for leaders, investors and boards. Make AI your advantage. Turn capital into outcomes.

Has AI Really Killed Middle Management?

Last week Meta cut 8,000 jobs as it bet its future on AI. Amazon announced another 14,000 AI related layoffs. The headlines all said the same thing: this is the end of middle management.

But let’s unpack that, because there is a real opportunity here. Middle management does slow things down. It does add a layer of risk-aversion at every step. And AI genuinely can cut away the need for some of it. The catch is that it has to be done right, and most companies are about to do it wrong.

Here’s something interesting, though. In last week’s poll I asked which part of your organisation AI would change first. Very few pointed at the management layer; most expected workflows and meetings to shift first. So the headline story, that AI is gutting middle management, doesn’t yet match what’s actually happening in most companies. That gap is what this issue is about.

Because when you actually read what Meta, Amazon and Block said about their cuts, it becomes clear they are not making the same argument at all. Three CEOs, three completely different forces at play.

Quick note before we dive in: this week is about the "company brain," the system that's quietly starting to do what middle managers used to.

It's also what I spend my days building at BoardLens, a company brain for funds and directors. So treat what follows as a dispatch from the inside, from somebody thinking about this every day, not a view from the stands.

Three CEOs, three completely different arguments

What sets Block apart is that Jack Dorsey is making the most radical argument of the three: that a company does not have to be organised as a hierarchy at all, as lays out in HumanxAI. He is rebuilding Block so that AI does the coordinating that managers used to do, with people sitting at the edges doing the parts only humans can. For Dorsey, the layoff is a consequence of the redesign, not the goal of it.

Amazon’s Andy Jassy is making a culture argument: too many layers, too much bureaucracy, decisions too slow. Fewer managers means fewer meetings and faster ownership. It is the same company shape as before, just flatter and leaner.

Meta’s Mark Zuckerberg is making a latency argument: small teams move faster. A handful of strong people with good AI tools can now do what used to take fifty, with far less time lost to communication. It is about speed, not redesign.

The headline everywhere was the headcount. The signal, in Block’s case, was the redesign. And when leaders copy the headline without the redesign, they are simply running a cost-cut in an AI costume.

What a middle manager actually does

Before you decide whether the role is dead, be honest about what it was for. The org chart, from the Roman legion to the modern pyramid, was never really a map of authority. It was a way to move information around, built for one stubborn human limit: a person can only properly coordinate three to eight others before things fall through the cracks.

Underneath the title, a manager does three jobs: they route information up and down, they pre-make small decisions so the top doesn’t have to, and they keep everyone pointed in roughly the same direction. That’s the job. And here’s what changed: a system with a constantly updated, readable picture of your whole business can now do the routing on its own. For the first time in two thousand years, that part of the layer is optional.

The distinction that decides whether your restructure works

Dorsey removed the coordination part of management. He did not remove the judgment part. Judgment didn’t disappear when the managers went; it moved upward, onto fewer people.

The reason Block could do this when most can’t comes down to unglamorous plumbing. It is remote-first, so for years almost all of its work has happened as written, recorded artifacts rather than hallway conversations, which meant it was capturing a machine-readable record of itself long before this moment. Add an unusually honest stream of data (it sees both sides of millions of payments a day), and by the time Dorsey cut the managers, the information they used to carry in their heads was already flowing through the system. Meta and Amazon are mostly trimming the same tree. Block replanted it first.

A good manager quietly catches the wrong number, the sloppy plan, the badly worded message, before it goes anywhere. Pull that filter out before anything replaces it, and small mistakes sail straight through, fast. That’s what “misalignment at machine speed” means: if everyone’s pointed slightly wrong, AI just lets them run in the wrong direction faster.

So how do you do it without breaking the company?

You start by building the thing that makes the cut safe. Tom Blomfield at Y Combinator has the cleanest name for it: the company brain. All the data, decisions and know-how currently trapped in people’s heads and your Slack history, pulled into one place and made usable, so the company can answer its own questions and improve itself.

The moment it clicked for me

You might be thinking, “Couldn’t I just do this in ChatGPT?” The answer is no, and that difference is a game changer.

I was running a board pack through BoardLens (disclaimer: the company I’m building), the company brain for the boardroom, built for funds, investment teams and directors. The pack said the company had seven months of runway. BoardLens flagged, unprompted, that it was really closer to three.

ChatGPT reads “seven months” straight back to you, because that’s what the page says. It only sees the one document in front of it, it waits to be asked, and it won’t proactively warn you or update you . And you can’t put a confidential board pack into it anyway.

A company brain is the opposite. It remembers every prior pack and every promise management made, so it caught a number no single document revealed. It works without being asked. The judgment stays with me; the remembering and the cross-checking is the machine’s job.

It's genuinely fascinating territory, and whether you buy one or build your own, a successful AI transformation starts with understanding the mechanics. Therefore, the rest of this issue is the playbook for understanding how you can start doing exactly that in your own organisation.

Here’s what’s behind the paywall (the actual how-to):

• 5 Steps to build the Company Brain the right way

• How to extract the domain knowledge trapped in your people’s heads

• How to build self-improving loops that get better while you sleep

• Where humans still matter, and the only two roles that survive

Read more

This issue lands in the middle of a fresh wave of AI-driven layoffs, and the mood in a lot of leadership teams has shifted from “how do we adopt this” to “what does this do to us.” I’ll give you my take on that below, and Thursday’s deep dive goes much further into what it means for how you lead and set up for success.

Regarding setting up for success: Check out our prompting hack #1 and Don’t Miss Ai4 2026 — America’s Largest AI Conference:

Join 12,000+ executives, founders, and investors exploring the breakthroughs, strategies, and market opportunities shaping AI. Dive into AI agents, RAG, MLOps, applied ML, and more — with 400+ exhibits, 1000+ speakers, keynotes from Geoffrey Hinton, Fei-Fei Li, and Andrew Ng, and peer networking. Taking place on August 4–6 at The Venetian in Las Vegas. Register before prices increase at midnight on May 28th to secure your pass and save $1,200+. Register Now.

The Top 10 Prompting Hacks for leaders, investors & boards

Most leaders are still under-using AI. Not because the model is wrong, but because the prompt is.

Anthropic’s Applied AI team trains engineers on a 10-layer framework for production prompts. The framework is excellent. It’s also written for engineers shipping code into Swedish car-insurance pipelines. Hence I took the liberty to translate it for people like us ;)

This guide adapts that same framework for a different reader: The CEO who needs to drive AI adoption across the team. The general partner triaging twelve portfolio updates before a Monday IC meeting. The board director reviewing a 200-page board pack on a Sunday night.

None of you will write a line of XML. None of you should have to. But the underlying logic — what the model needs to give you a usable answer — applies just as much to your work as it does to anyone else’s. Possibly more, because the stakes are higher and the time is shorter.

The Framework at a glance

Think of a prompt as the brief you’d give a sharp new hire on day one. Skip the context and they fill the gap with a guess. The model does exactly the same thing.

The lazy version

Summarise this board pack.

The sharp version

You are a senior board advisor with 15 years of experience on Series B+ SaaS boards. I am a non-executive director on the audit committee. I’m reviewing this 80-page board pack before tomorrow’s meeting and need to walk in able to ask three sharp questions of the CFO.

Why it matters: the same model can play “intern” or “former Big Four audit partner” depending entirely on how you frame it. The framing costs you one sentence and changes the entire shape of the answer.

In your seat, if you’re a GP:

You are a partner at a top-decile Series A fund focused on B2B SaaS. I’m preparing for an IC meeting tomorrow on a deal where I have reservations. Read the memo below and tell me the three weakest claims a sceptical partner would push on.

That’s one of ten. The full field guide, all ten hacks plus a bonus framework for stacking them, is free and yours instantly. No referral, no paywall.

Get the full guide, free →

From Hierarchy to Intelligence

This was a big week of AI related layoffs, with Meta slashing 8,000 jobs as it bets its future on AI. It's worth remembering Block, under Jack Dorsey, has been among the most vocal in making the AI-driven restructuring argument, which makes their reasoning worth a closer look.

Most companies are using AI as a productivity tool: helping employees write faster, summarize more, or work more efficiently. Block is taking a much bigger view. Their argument is that companies have spent 2,000 years building layers of management to move information through an organization, and AI may now become a fundamentally new coordination mechanism. Instead of simply making existing structures faster, AI could reshape how work gets organized, how decisions get made, and where leverage is created.

Board & Investor takeaway: Most AI conversations start with “Where can we reduce costs?” A more important question may be: “How might AI fundamentally change how this company operates and create leverage?” The biggest opportunity may not be improving existing workflows. It may be redesigning them altogether.

Thursday's deep dive goes further: What does it actually take: What set up, which prerequisites, to start a leaner, AI first organisation, whether you are a small startup or a large enterprise. And how to successfully drive that change. For paid subscribers only. —> Upgrade now!

The Future Technology Company won’t look like a Startup. It will look like Dior.

As AI pushes the cost of building software toward zero, speed and execution may no longer be enough. The argument: tomorrow’s winning companies will increasingly compete on taste, judgment, and the ability to reinvent themselves constantly, more like a fashion house launching new collections than a traditional SaaS company shipping one product. The CEO becomes less of a manager and more of a creative director orchestrating vision, systems, and talent at speed.

Board takeaway: In the AI era, competitive advantage may shift from who can build fastest to who can make the best high-frequency judgment calls.

Investor takeaway: if anyone can build the product, the product is no longer the moat, and that changes what you're underwriting. The edge moves to taste, brand and the founder's ability to keep reinventing, which is harder to diligence and easier to mistake for hype. Be wary of companies whose only advantage is speed; it's the cheapest thing to copy now. The durable question: when this product is commoditised in 18 months, does this team have the judgment to launch the next one?

Share

Which metric will matter most in AI-native companies

In last week’s post, I ran a poll on this topic. Here’s how it landed:

The lack of a clear winner is interesting. Revenue per employee is the number investors already reach for, so it's the safe pick. The near-even tie below it suggests we do not know how best to track AI productivity yet.

My bet: decision velocity is the one almost nobody measures today and the one that will quietly separate the winners. You can find examples how companies measure decision velocity in this post —> How to measure AI productivity.

The Director’s Dilemma: When the Most Valuable Thing on the Board is the Thing AI Cannot Replicate

Recent headlines claimed a robot had joined the board of Lloyds Bank. Not quite. Lloyds introduced an AI tool to help directors prepare for meetings, but it does not vote or make decisions. The bigger issue is that many boards are now expected to oversee AI systems they did not build, cannot fully audit, and often cannot fully understand. At the same time, 4 in 10 workers now cite AI-driven obsolescence as a major career fear, and boards are not immune. Fear creates a dangerous side effect: it weakens judgment precisely when judgment matters most.

Board & Investor takeaway: AI may process information faster than any director, but discernment remains human. The board challenge is shifting from “understanding AI mechanics” to preserving independent judgment, curiosity, and oversight while the technology accelerates around us.

Share

UAE CEOs fear AI failure more than AI itself

AI is becoming a CEO accountability issue. Nearly 6 in 10 UAE CEOs say they are under direct board pressure to deliver measurable AI results, while 76% say investors now see AI strategy as critical to performance. Yet 44% have delayed or cancelled AI projects because they fear failure, and 75% believe CEOs could lose their jobs over failed AI strategies or AI-related crises. The leadership lesson: AI risk is increasingly becoming governance risk.

Interested in sponsoring? Find out more

Jobs of the week

• President, AI Acquisition - Remote

• Member of AI Advisory Board - CISO/ CIO, Andeavour, Remote

• Principal, AI Leadership Practice, The Forward Institute, London

• Director, AI Enablement, CoinDesk, New York

• Head of AI Productivity Lab, Lloyds Banking Group, UK

• Head of Internal AI, User Testing, Remote (US / UK)

• Project Manager (Part time) and AE/Sales VP, QueryPal, Remote

If you’d like us to include your jobs in the next issue for free, message us here

Post of the week

If AI makes a serious mistake, who in your organisation is accountable for it?

The answer, that the agent never is and the board / leadership team always is, is the cleanest articulation of AI-as-governance-risk I've seen this week. See full post.

Refer a friend and get the Executive Meeting Toolkit: 5 frameworks for the meetings you run every week. [Refer a friend →]

A thanks to our partners:

BoardLens, the AI super analyst for fund partners and board directors. It reads everything they don't have time to, tracks every commitment management makes, surfaces what's drifting before it becomes a blowup, and prepares the sharper questions that make them the strategic sparring partner the CEO actually wants to call

Chronicle Labs Find and stress test agent failures before your customers do, using operational data to replay production-like scenarios.

Ai4 2026 Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas to explore the real-world applications of AI Agents, World Models, and more alongside pioneers like Geoffrey Hinton, Fei-Fei Li, and Andrew Ng. Register now

Enquire about our Sponsorship options

Read more from the AI Leadership Edge

How to track AI productivity Everyone claims AI productivity gains. Almost nobody can measure them.

The AI Stress Test Every Leader Should Run 6 prompts to stress test your AI strategy across execution bottlenecks, talent and team risks, strategic and competitive blind spots and more.

The Top 10 Books for leaders, directors and investors Required reading for boards, the truth about productivity, AI’s macro disruption unpacked — and a viral reminder of how not to run an AI transformation.

The Advisory Board Squeeze (And Why AGI Is Closer Than You Think) Does Your Advisory Board Make You Uncomfortable? Becoming AI fluent investor edition and the napkin test for leadership gaps

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for business leaders, boards and investors who want to gain leverage with AI.

This newsletter is powered by BoardLens, the AI analyst for fund partners and board directors. It reads everything they don't have time to, tracks every commitment management makes, surfaces what's drifting before it becomes a blowup, and prepares the sharper questions that make them the strategic sparring partner the CEO actually wants to call.

Enter BoardLens

For the last two years, “AI productivity gains” became the most unfalsifiable claim in business.

Every company says AI is improving productivity.

Few can explain exactly where.

Fewer can quantify it.

And almost nobody can distinguish real organizational leverage from temporary efficiency “theater”.

That’s because most companies are measuring AI productivity at the wrong layer.

They measure:

• prompts used

• hours saved

• copilots deployed

• employee usage rates

• tasks completed faster

These are activity metrics.

Not transformation metrics.

The real productivity gains from AI happen at the organizational level.

The companies that win in the AI era will not simply do the same work faster.

They will redesign how work happens entirely.

The Old Productivity Model

For decades, productivity scaled mostly through one thing:

More people.

More analysts.

More managers.

More coordination.

More meetings.

More layers.

Traditional organizations grew like pyramids because human coordination was the bottleneck.

If output needed to increase, companies added headcount.

That model shaped almost every management system we still use today.

AI breaks this logic.

Because AI dramatically reduces the coordination cost of execution.

One highly capable operator with AI support can now perform work that previously required entire teams.

This changes the economics of organizations.

And it changes what productivity actually means.

The New Productivity Metric

The companies creating the most value from AI are not necessarily the ones deploying the most AI tools.

They are the ones increasing organizational leverage.

The new metric is not:
“Are employees using AI?”

The new metric is:

How much output can the organization generate per unit of coordination?

That changes everything.

Because coordination is what slows companies down:

• meetings

• approvals

• reporting layers

• information bottlenecks

• management overhead

• cross-functional dependencies

AI compresses those costs.

Which means the highest-performing companies in the next decade may look fundamentally different:

• smaller teams

• fewer management layers

• higher revenue per employee

• faster decision cycles

• more autonomous operators

• dramatically lower coordination overhead

The Shift Most Leaders Miss

Most AI initiatives focus on local productivity.

“How do we help employees work faster?”

But the biggest gains come from system redesign.

Not task acceleration.

The companies that outperform will rethink:

• team structure

• management layers

• workflow ownership

• decision rights

• operating cadence

• communication systems

AI is not just a software upgrade.

It is a management revolution.

And most organizations are still measuring it with industrial-era metrics.

The New AI Productivity Stack

Share

What Boards Should Actually Track

Boards asking:
“How many employees are using AI?”

are asking the wrong question.

The better questions are:

• Has decision velocity increased?

• Has revenue per employee increased?

• Are management layers compressing?

• Is execution becoming faster?

• Is coordination overhead falling?

• Can smaller teams now create disproportionate output?

Because in the AI era, productivity is no longer mainly about labor efficiency.

It is about organizational leverage.

That is the real shift.

Boardroom Takeaway

The companies that win in the AI era will not necessarily have the best models.

They will have the operating systems that create the most leverage per person.

And that is a very different kind of company than the world has historically rewarded.

🔒 The AI Productivity Scorecard

Most leaders and boards track AI adoption, training completion, and tool usage.

None of those metrics tell you if AI is actually changing how the business operates.

The scorecard below shows the 10 metrics AI-native companies track internally to measure organizational leverage, not activity theater.

Each metric includes:

• How to calculate it (exact formulas)

• Real before/after case studies

• Implementation templates

• Benchmarks (what good looks like)

If you’re on a board, managing a portfolio, or leading AI transformation, this is the measurement system you need. Available to paid subscribers

A thanks to our partners:

BoardLens, the AI intelligence platform for board directors and investors- Effectiveness redefined for all post -investment and board work.

Chronicle Labs Find and stress test agent failures before your customers do, using operational data to replay production-like scenarios.

Ai4 2026 Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas to explore the real-world applications of AI Agents, World Models, and more alongside pioneers like Geoffrey Hinton, Fei-Fei Li, and Andrew Ng. Register now

Enquire about our Sponsorship options

Read more from the AI Leadership Edge

The New Metric That Redefines Companies AI-compressed orgs, skyrocketing revenue-per-employee, synthetic boards, and a venture world where attention beats money.

The AI Stress Test Every Leader Should Run 6 prompts to stress test your AI strategy across execution bottlenecks, talent and team risks, strategic and competitive blind spots and more.

The Top 10 Books for leaders, directors and investors Required reading for boards, the truth about productivity, AI’s macro disruption unpacked — and a viral reminder of how not to run an AI transformation.

The Advisory Board Squeeze (And Why AGI Is Closer Than You Think) Does Your Advisory Board Make You Uncomfortable? Becoming AI fluent investor edition and the napkin test for leadership gaps

Read more

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for business leaders, boards and investors who want to gain leverage with AI.

This newsletter is presented by Chronicle Labs. As enterprises rush to deploy AI agents, the biggest risk is whether they behave safely across the constantly changing, messy, high-stakes edge cases of production. Teams cannot realistically map every workflow or scenario by hand, and the real danger is discovering how agents respond to shifting data, user behavior, and business rules only after they are live with customers or critical operations.

Chronicle Labs gives teams a staging environment for enterprise AI agents, using operational data to replay production-like scenarios, uncover failure modes, and stress-test new agent behavior before launch. Find out more at chronicle-labs.com.

The Art of Delegation in the Age of AI

Real delegation has always been a leadership skill.

AI has only made the cost of being bad at it dramatically higher.

Research from INSEAD found delegation was rated the second most important leadership skill after empathy. Yet most of us still struggle with it.

The leaders getting the most leverage from AI treat it differently. They don’t use it like a search engine. They use it more like a capable junior operator that needs context, feedback, and clear direction.

One of the biggest unlocks is simply giving AI more background about how you think and work. Use tools like Claude Projects, ChatGPT Projects, or Granola’s “Coach Me Matt”. Read full article for more tips around delegation in the age of AI.

Reply in the comments: What is one leadership task are managers struggling most with to delegate to AI? I’m curious where leaders still resist AI most.

Share

A Requiem for the Employee

Through AI we’re not witnessing the death of employment. We’re witnessing the death of the concept of the employee—the social and economic container we built around labor as there is less full-time employment and more fractional and freelance work.

And once the employee stops being the default, everything else shakes: health insurance, retirement planning, professional identity, the meaning of a career.

AI isn’t removing the need for people. It’s changing the shape in which people are engaged. Read the full piece here →

Share

AI LEADERSHIP JOBS

In a recent Substack note, I asked what the most interesting roles in AI leadership are.

Thanks for the interesting inputs, as one subscriber put it:

• Amazon: Hiring builder-operators who can help us reimagine workflows with AI-enabled solutions while managing large retail categories for Amazon Canada.

  3 roles open on the Consumer Electronics team which I lead. We are hiring in Vancouver, Toronto, and Seattle. This team has already launched AI chat agents which have scaled WW and are actively used by over 1k Amazon employees. Dm Christian Lieber or apply via Link.

• SVP, AI Strategy

• Director, Privacy & Governance, Marriott International

• AI Resilience Fund Manager, Chicago

• Managing Director, AI Venture Buy Outs

• Head of AI Think Tank

If you’re hiring — or know of great open roles in AI leadership, AI investing or AI governance — get in touch via dm. I’ll feature the most interesting ones.

The Top 10 Prompting Hacks for leaders, investors & boards

Prompting is becoming one of the most important leadership skills of the AI era, and most people are still doing it wrong. We just released a new guide: The Top 10 Prompting Hacks for Leaders, Boards & Investors. To claim your copy, refer 1 friend.

Refer a friend

The Average Org Chart Is Getting More Technical, Not Less

The narrative six months ago: Claude Code and Codex would let companies ship roadmaps with half the engineers. The reality: companies are hiring more engineers as AI expands what’s buildable. Pave's data from 9,300+ companies shows R&D's share of headcount climbed from 38.2% (Q4 2023) to 41.6% (Q1 2026)—directly contradicting predictions that AI would shrink engineering teams.

Two additional signals:

• Technical talent is spreading beyond R&D. Engineers and data scientists are increasingly hired into HR, Sales, and Ops to integrate AI. The “Rest of Business” category is quietly becoming technical.

• Job postings confirm the trend. Redpoint’s data shows software engineer postings recovered to 97% of baseline. AI is expanding demand for engineering talent, not contracting it.

The 10x productivity story hasn’t translated to macro-level headcount cuts. It’s translated to bigger ambitions.

Leave a comment

The “AI” Stock Boost

Allbirds ($BIRD), the eco-conscious footwear company once valued at $4 billion, was down 99% from its peak as of April 14, 2026. Sales had plummeted nearly 50% between 2022 and 2025—from $298 million to $152 million. CNBC

Then the company announced it was becoming “NewBird AI.”

The stock surged 582% in a single day. Shares that traded under $3 jumped to $17.

The playbook: Sell the footwear assets to American Exchange Group for $39 million, secure $50 million in convertible financing, and pivot to “AI compute infrastructure”—specifically, buying GPUs and renting them out as GPU-as-a-Service.

What this tells us: A company can slap “.AI” on its name the way 1990s companies added “.com”—and watch the stock fly. Even shoe companies pivoting to entirely unrelated businesses get rewarded if the press release mentions GPUs.

As William Blair analysts noted, $50 million is “a drop in the bucket” in the AI infrastructure business, where the barrier to entry starts in the billions of dollars. Retail traders bought $5.2 million worth of stock in a single day—the most action the stock had seen since its 2021 IPO. The current reality: As of May 11, 2026, the stock trades at $5.33, down from the $17 peak but still up over 75% from pre-pivot levels. Morningstar

The question for investors and boards: If a collapsing footwear retailer with zero AI expertise can raise $50M and 5x its stock price by announcing an infrastructure pivot, what does that signal about market discipline—and how long does it last?

A thanks to our partners:

Ai4 2026 Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas to explore the real-world applications of AI Agents, World Models, and more alongside pioneers like Geoffrey Hinton, Fei-Fei Li, and Andrew Ng. Register now

BoardLens, the AI intelligence platform for board directors and investors- Effectiveness redefined for all post -investment and board work.

Chronicle Labs Find and stress test agent failures before your customers do, using operational data to replay production-like scenarios.

For Sponsors, contact us here

Read more from the AI Leadership Edge

The AI Stress Test Every Leader Should Run 6 prompts to stress test your AI strategy across execution bottlenecks, talent and team risks, strategic and competitive blind spots and more.

The New Metric That Redefines Companies AI-compressed orgs, skyrocketing revenue-per-employee, synthetic boards, and a venture world where attention beats money.

The Top 10 Books for leaders, directors and investors Required reading for boards, the truth about productivity, AI’s macro disruption unpacked — and a viral reminder of how not to run an AI transformation.

The Advisory Board Squeeze (And Why AGI Is Closer Than You Think) Does Your Advisory Board Make You Uncomfortable? Becoming AI fluent investor edition and the napkin test for leadership gaps

🔒 Enjoying The AI Leadership Edge? Upgrade to paid for exclusive deep dives, premium AI leadership frameworks, playbooks, workflows, and practical tools designed for leaders, boards, and investors.

Subscribe now

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for investors and business leaders who want to gain leverage with AI.

Today’s newsletter is proudly presented by:

Navigate the evolving AI landscape and learn the top AI use cases at Fortune 500s.

Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas to explore the real-world applications of AI Agents, World Models, and more alongside pioneers like Geoffrey Hinton, Fei-Fei Li, and Andrew Ng. Don’t miss the opportunity to learn about the top AI use cases at Fortune 500s while exploring 400+ hand-picked vendors and connecting with leaders in your industry. Register now for Ai4 2026 by May 28th to save $1,200+.

The AI Stress Test Every Leader Should Run

Your AI strategy is not being tested when you announce it.

It’s being tested quietly, every day, inside the operating system of your company.

In how decisions get made. How fast information moves. Which managers create leverage and which create drag. Which workflows still exist only because “that’s how we’ve always done it.” Which executives understand what AI changes and which are still treating it like software procurement.

Most leadership teams think they are preparing for AI because they have pilots, copilots, vendors, workshops, or a roadmap. That is not preparation. That is activity.

The companies that survive the next decade will not necessarily have better AI tools.

They will have leadership teams capable of redesigning how the company operates while everyone else is layering AI onto structures built for a different era.

The uncomfortable reality is this: Most organizations are trying to run AI-native technology inside industrial-era management systems.

That mismatch is where the failure starts.

The problem is that executives rarely get honest feedback about this. Employees soften the truth. Consultants avoid threatening the client relationship. Board members often lack the operational visibility to see where the real friction lives.

And internally, almost nobody wants to be the person who says:

“We are moving far slower than we think.”

AI, used correctly, changes that dynamic.

Not because it replaces executive judgment. Because it exposes weak executive thinking faster than traditional organizations can hide it.

The best CEOs are starting to use AI in a completely different way.

Not as a productivity tool. As a pressure-testing system.

A way to simulate the questions a ruthless board member, activist investor, or AI-native competitor would ask if they were trying to expose every structural weakness in the business.

The goal is to discover the truth early enough to do something about it.

Is Your Leadership Team Actually Ready for AI?

Five uncomfortable questions.

1. Can you explain specifically how AI changes your company’s economics?Not “AI improves productivity.” What line item changes? Headcount. Margins. Revenue per employee. Sales capacity. Customer acquisition cost. Decision velocity. If you cannot explain the mechanism clearly, you do not yet have an AI strategy. You have AI enthusiasm.

2. Which parts of your organization become liabilities in an AI-native environment? Most executives frame AI around opportunities. Few are honest about what becomes structurally weaker. Management layers built around information control. Slow approval systems. Meetings designed for coordination overhead. Functions optimized for reporting instead of execution. AI does not just create leverage. It exposes drag.

3. If your competitors adopted AI perfectly 24 months before you, what breaks first? Most leadership teams underestimate how quickly operating advantages compound. A competitor making decisions 30% faster eventually feels like a competitor operating in a different market.

4. Which executive on your leadership team understands AI deeply enough to redesign workflows, not just discuss tools? This is where many transformations quietly fail. The gap between “AI awareness” and “AI operational capability” is enormous. Most executive teams still confuse the two.

5. Is your company using AI to accelerate the current operating model, or redesign it completely? This is the defining question. The companies that win in the AI era will not simply automate existing work. They will rethink why the work exists in the first place.

🟦 Prompt 1: The Adversarial Setup

Without this framing, AI defaults toward encouragement and generic strategy advice. With the right setup, it becomes something far more useful: A leadership stress-testing system. Paste this first:

💡 Keep this conversation open across all ten prompts. The accumulated context (every hesitation, every weak justification, every assumption you couldn’t quite defend) matters.

🟦 Prompt 2: The AI Strategy Attack

Most AI strategies fail for the same reason most corporate transformations fail:

The strategy sounds impressive at the board level but collapses operationally. The problem is rarely ambition. The problem is translation.

Leadership teams say things like: “We’re embedding AI across the organization.”

But underneath that statement there is often no clear operational model for:

• decision rights

• workflow redesign

• talent changes

• incentive systems

• management restructuring

• capital allocation

• speed of execution

This is where the stress test begins. Now insert the following prompt in to Claude or LLM of choice:

💡 After the exchange, ask: “Based on everything I’ve told you, what are the two strongest aspects of my AI strategy, and what are the two biggest risks that could make this leadership team fall behind faster than they realize?”

Save that output. That is your preparation document before every board meeting, strategy offsite, and executive planning session.

🟦 Prompt 3: The Execution Bottleneck Interrogation

This is where most AI transformations quietly die.Not because the models are weak. Because the organization is slow. A company can buy every AI tool on the market and still lose if decision-making remains trapped inside industrial-era management systems. The uncomfortable truth is that AI increases the penalty for bureaucracy.

Every approval layer. Every reporting chain. Every meeting that exists because nobody trusts the system enough to move faster. Those costs compound.

AI-native companies operate with dramatically lower coordination overhead. That advantage widens every quarter.

Next: Stress test execution bottlenecks, talent & team risk, competitive blind spots and sharpen your answers for your next board or investor meeting.

After running Prompt 1-6, what surprised you most about your own answers? Hit reply—I’m reading every response and will feature the most interesting insights (anonymously) in the next issue.

A thanks to our partners:

Ai4 2026 Join 12,000+ peers at Ai4 2026 on August 4-6 in Las Vegas to explore the real-world applications of AI Agents, World Models, and more alongside pioneers like Geoffrey Hinton, Fei-Fei Li, and Andrew Ng. Register now

BoardLens, the AI intelligence platform for board directors and investors- Effectiveness redefined for all post -investment and board work.

Chronicle Labs Find and stress test agent failures before your customers do, using operational data to replay production-like scenarios.

For Sponsors, contact us here

Read more from the AI Leadership Edge

The New Metric That Redefines Companies AI-compressed orgs, skyrocketing revenue-per-employee, synthetic boards, and a venture world where attention beats money.

The Top 10 Books for leaders, directors and investors Required reading for boards, the truth about productivity, AI’s macro disruption unpacked — and a viral reminder of how not to run an AI transformation.

The 2026 Playbook: Six Questions That Force Execution Six questions for 2026 that force execution, and what 2025 proved about AI-first operating models, metrics, and governance.

The Advisory Board Squeeze (And Why AGI Is Closer Than You Think) Does Your Advisory Board Make You Uncomfortable? Becoming AI fluent investor edition and the napkin test for leadership gaps

Read more

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for investors and business leaders who want to gain leverage with AI: turning board work and post-investment execution into better decisions, better alignment, and faster results.

This week’s newsletter is sponsored by Chronicle Labs

As enterprises rush to deploy AI agents, the biggest risk is whether they behave safely across the constantly changing, messy, high-stakes edge cases of production. Teams cannot realistically map every workflow or scenario by hand, and the real danger is discovering how agents respond to shifting data, user behavior, and business rules only after they are live with customers or critical operations.

Chronicle Labs gives teams a staging environment for enterprise AI agents, using operational data to replay production-like scenarios, uncover failure modes, and stress-test new agent behavior before launch. For leaders, this turns AI adoption from a leap of faith into a governed rollout: faster experimentation, fewer production failures, and clearer confidence in where agents are ready to operate.

Enter Chronicle Labs

It’s time for Leaders and Boards Optimize for Non-Human Customers

At 3:17 a.m. last Tuesday, an AI agent representing a procurement team scanned a Fortune 500 supplier’s website, cross-referenced pricing against three competitors, evaluated lead times and compliance certifications, and made a shortlist decision. Total time on site: 84 milliseconds. The supplier’s marketing team spent that same Tuesday debating a new homepage hero image. Nobody plans to discuss any of it at next week’s board meeting.

The internet has bifurcated into a human web and an agentic web. Most boards are still governing the first one. The economic activity is moving to the second.

The numbers: Automated traffic surpassed human activity in 2024 (51% of all web traffic). Cloudflare’s CEO puts a date on full crossover: AI bot traffic will exceed human traffic by 2027. Three companies generate 96% of AI traffic hitting your properties (OpenAI 69%, Meta 16%, Anthropic 11%). Adobe found that traffic from AI sources to retailers surged 1,200% year-over-year while traditional search traffic declined 10%. McKinsey forecasts $3-5 trillion in global retail revenue flowing through agentic commerce by 2030.

The shift: AI agents don’t care about your homepage hero image. They parse structured data, extract pricing, compare competitors, and decide whether a human ever sees your company at all. The human web rewards persuasion. The agentic web rewards machine judgment. Companies that grew up on search engine optimization are scrambling toward agent legibility—clean pricing schedules, machine-readable catalogs, API access, verifiable claims.

The governance gap: For 25 years, enterprise security was built on one premise: keep the bots out. That premise inverts. The question is now which non-human actors to trust, admit, throttle, authenticate, monetize, or reject. Most boards can’t answer what percentage of their traffic is human versus non-human, let alone distinguish customer agents from malicious scrapers.

For leaders and boards: Put these questions on the next agenda: What percentage of our traffic is non-human? Can we distinguish malicious bots from customer agents? Is our pricing machine-readable or trapped in PDFs? What’s our position on Universal Commerce Protocol, and who’s accountable? When an AI agent transacts, what’s our audit trail? If management can’t answer, the question isn’t when they can—it’s why they can’t.

Boards need to rethink how they advise CEOs

A new HBR article backed by the AlixPartners Disruption Index exposes a troubling paradox: 87% of CEOs say their boards have the right people and knowledge to help them navigate disruption. Yet 72% say they find it increasingly difficult to set priorities. If boards have what CEOs need, why do CEOs still feel adrift?

The problem isn’t the people. It’s the process. Traditional board structures—voluminous board books, variance-to-plan discussions, single-minded strategies—were designed for predictable environments. They fail when the path to value isn’t a straight line.

Three shifts boards need to make:

Stop putting management on the defensive about course corrections. Shift the conversation from defending deviations from the plan to testing assumptions and discerning priorities. Expect pivots. Reward agility.

Treat strategy like a portfolio, not a single bet. In uncertain times, long-term big investments (building a new plant) are riskier than ever. Place more bets. Create options. The right priorities emerge through experimentation, not annual strategic planning sessions.

Ask about customer conversations, not just metrics. When uncertainty is high, organizations turn inward. Boards can counter this by asking what pain points customers are raising, not just reviewing NPS scores. For tech: invest in data quality before chasing AI opportunities—leaders in AI are nearly twice as likely to have problem-free legacy systems. For risk: shift from cataloging threats to stress-testing operational agility using real-time supply chain monitoring.

For boards and investors: The shift is from “certifying board” (informed but hands-off) to “engaged board” (sleeves-rolled-up, question-driven). Stop reviewing the past. Start shaping the future.

Leave a comment

Why 80% of Companies See No ROI from AI

McKinsey’s June 2025 “Seizing the Agentic AI Advantage” named the resulting condition the “gen AI paradox”: nearly eight in ten companies report using generative AI, yet just as many report no significant bottom-line impact. The diagnosis is structural. McKinsey’s November 2025 State of AI found that only 21% of companies have redesigned workflows end to end, and AI high performers are nearly three times as likely to have done so. Everyone else bolted agents onto an org chart designed around handoffs. The handoffs are where the value leaked.

If you are interested in sponsoring and receiving our media kit, message us here: https://substack.com/@theaileadershipedge

For Sponsors: Enquire about media kit

Presented by BoardLens, Intelligence for deployed capital - Your team of AI agents for the work after the cheque clears.

Enter BoardLens

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for investors, business and board leaders who want to gain leverage with AI: turning board work and post-investment execution into better decisions, better execution, and faster results. We’ve just been featured #88 rising in business by Substack 🤗

In this weeks’s newsletter:

• After the Deal: The Ownership Audit

• The Question: AGI Is 1-3 Years away, how to prepare now?

• From the boardroom: The Advisory Board Squeeze

• The Five-Minute Edge: Figuring out how to become “AI native”- Investor edition

After the Deal

The Ownership Audit

When two people share responsibility, both quietly assume the other will catch what they miss. Standards slip. Tasks get left incomplete. Nobody decides to let things drop—it just becomes possible because someone else might handle it. Over time, resentment builds and the system degrades.

The solution isn’t another alignment conversation. It’s an ownership audit. This one can be done on a napkin:

Draw two columns. Column A: areas where you and another leader are both “responsible”—meaning if something falls through the cracks, neither of you is clearly accountable. Column B: areas where one person holds undivided ownership.

Now look at Column A. For each item, name one owner. If you can’t, that’s your gap.

For boards and investors: Is there a domain in your portfolio company that would benefit from the CEO stepping away from a responsibility for a defined period—not to abandon it, but to force genuine ownership into the people who need to grow into it? That might be the most valuable move of the quarter.

Share

The question

AGI Is 1-3 Years away, how to prepare now?

“We’re near the end of the exponential,” says Dario Amodei, CEO of Anthropic. His full breakdown: AGI arrives in 1-3 years.

For investors: Productivity gains are compounding faster than any model shows. Coding went from 5% to 20% speedup in six months. In 18 months it could be 80%. That changes the terminal value of every software company on the planet. Reprice accordingly.

For business leaders: Build for a world where AI handles 90% of what your employees do today. Don’t hire for tasks. Hire for judgment, taste, and the ability to direct AI systems. The companies that figure out human-AI collaboration first will dominate the 2030s.

For tech employees: The window to become exceptional at working with AI is closing fast. In two years, this skill goes from differentiator to table stakes. If you haven’t started, here’s the fastest on-ramp: use Claude for content workflows or spreadsheets. Start now. Not next quarter.

For anybody else: AI is coming for knowledge work the way machines came for manual labor. But faster. The people who learn to work alongside these systems will thrive. The people who ignore the exponential won’t see it until it’s past them.

Subscribe now

From the boardroom: Grateful for our first ever guest post on this newsletter. Let’s give up the floor for Jurgen Appelo, a CIO rethinking governance

The Advisory Board Squeeze: How AI and tighter governance are shrinking the space for advisory boards

Here’s an uncomfortable question for anyone sitting on an advisory board:

What exactly do you offer that Claude and ChatGPT can’t?

AI tools now deliver strategic analysis, market scenarios, and regulatory briefings faster and cheaper than most advisory boards ever could. Meanwhile, governance boards are absorbing the accountability side: courts and regulators (including the EU AI Act, enforceable from August 2026) increasingly expect fiduciary directors to own AI oversight directly.

Advisory boards find themselves caught in the middle. The information function moves down to AI. The accountability function moves up to governance. What remains is a narrower slice: contextual judgment—the messy decisions connecting technology, regulation, and organizational reality that neither AI nor compliance structures handle well.

Yet many advisory boards still function as investor theater: impressive names, minimal challenge. That’s the real risk. Not that advisory boards disappear, but that organizations mistake signaling for substance.

The takeaway for leaders and boards: Audit your advisory board against one question — does this group of advisors regularly make us uncomfortable? If the answer is no, you’re paying for decoration. Keep advisory boards small, domain-specific, and focused on judgment calls where the playbook hasn’t stabilized. Let AI handle the briefings. Let governance handle the liability. Let advisors handle the arguments nobody else wants to have.

See full article here

This is a guest post by Jurgen Appelo and his Substack The Solo Chief.

Jurgen is a founder, intrapreneur, and former CIO rethinking governance for the one-person business, navigating sole accountability in the age of intelligent machines—informed by plenty of scar tissue.

The Five Minute Edge

Figuring out how to become “AI native”- Investor edition

USV, a mega VC fund, shares that their breakthrough with AI wasn’t the technology. It was treating agents like actual employees: give them a name, a role, access to the same tools your team uses, and let them live where you communicate. For USV, that means agents chime in on email threads. Agent Arthur literally introduces himself to the team and gets feedback when he makes mistakes.

Here’s what worked: they started with one annoying problem (meeting recaps that no one had time to write), solved it well, and the next problem revealed itself. They didn’t plan to build a custom CRM. They just kept fixing what was broken in how they actually work. Now their agents capture every mention of companies and people across meetings, emails, and calendars, turning scattered context into organized memory.

They say the shift is real: we’ve moved from “build something people want” (expensive mass-market software for everyone) to “build something you want” (custom agents that solve your exact problems).

I definitely love to name your agents, great idea!

Leave a comment

Presented by BoardLens, Intelligence for deployed capital - Your team of AI agents for the work after the cheque clears. Check it out here.

Enter BoardLens

If you are interested in sponsoring and receiving our media kit, message us here: https://substack.com/@theaileadershipedge

For Sponsors: Enquire about media kit

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for investors, business and board leaders who want to gain leverage with AI: turning board work and post-investment execution into better decisions, better execution, and faster results. We’ve just been featured #88 rising in business by Substack 🤗

This weeks’s newsletter comes again in the new format:

• From the boardroom: I’ve started a new company!

• The Five-Minute Edge: What actually predicts billion dollar exits?

• After the Deal: The AI metric all CEO’s should be fearing

• The Question: One high-leverage question

I’ve started a new company, and I’ve never been more energized.

After years working with boards and investors, I kept seeing the same bottleneck: post-investment and board work is time-consuming, high-stakes, and nearly impossible to delegate well. Board meetings, strategic calls, chasing data across portfolios, investor updates, LP reporting—it pulls focus from what actually drives fund returns.

Today, I'm excited to officially launch BoardLens — an intelligence platform for deployed capital. We're building specialised AI agents that make post-investment work far more efficient, handle the operational heavy lifting and turn scattered post-investment activity into durable operating memory, so teams can focus on maximizing portfolio performance instead of managing the admin layer around it.

We're already off to a strong start. Angel round closed. Beta partners managing ~$7B AUM. A cross-functional team across the US and Germany bringing AI engineering, Fortune 500 product design, elite cybersecurity, and capital markets expertise.

We’re opening early conversations now with VC, PE, and Family Office teams looking to use AI to drive post-investment performance.

Learn more about BoardLens

What Actually Predicts Billion-Dollar Exits

Andy Chen and Amy Lin at Outcast Ventures analyzed every U.S. tech company that went public or was acquired for $1B+ over the past 20 years, manually reviewing 2,000+ LinkedIn profiles to reconstruct the founding stories behind each exit.

Three findings shatter conventional wisdom:

Prior colleagues underperform strangers. Founding teams who had worked together before achieved a median exit of $2.3B, compared to $2.9B for teams with no prior work history—a 21% gap that contradicts the idea that shared history de-risks partnerships.

Prior exit experience nearly doubles IPO outcomes. Teams where at least one founder had previously built and exited a company reached a median valuation of $5B at IPO, versus $2.6B for teams without exits. CEO startup experience alone added a 41% premium.

Solo founders nearly disappear at the top. While 82% of all $1B+ exits came from teams of 2+, among the top 5% of exits ($24B to $98B), only one was led by a solo founder. Solo-founded companies also take roughly three years longer to exit.

For investors and boards: stop overweighting “they’ve worked together before” as a signal. Prior execution matters far more than prior relationship. The data says bet on founders who’ve already exited, assemble complementary skill sets even among strangers, and push solo founders to add co-founders early if you want top-decile outcomes.

Share

The AI metric all CEOs should be fearing

Most companies say they’re becoming “AI-first.” They’ve run the trainings, launched task forces, sent people to conferences. But when you ask how AI-fluent their teams actually are, the answer is always vague: “Really well, I think. We’re making progress.”

Creandum’s VP of Talent Michelle Coventry introduces two metrics to cut through the theater. TAFI (Talent AI Fluency Index) measures individual leader fluency starting with the CEO, because transformation only happens when someone at the top understands the technology well enough to make structural calls on pricing, architecture, and hiring. TAFD (Talent AI Fluency Density) maps fluency across the entire organization.

The threshold is behavioral, not credential-based. The question isn’t “did they complete the module?” It’s “what did they build, stop doing, or do faster because of AI?”

Here’s what healthy looks like: C-suite fluency can’t be 80% while middle managers sit at 30%. If fluency is concentrated at the executive level, that’s a team with conference access, not a transformed organization. The overall score should clear 60%, meaning people are genuinely using AI to change how they work, not just aware they should.

For boards and investors: Ask your CEOs and portfolio companies to map their AI fluency across layers. The uncomfortable truth is that most leadership teams sound fluent but can’t point to concrete workflow changes. Make C-suite AI proficiency non-negotiable within 90 days. Tie manager progression to AI adoption. Measure what changed, not what was completed. The companies that can’t answer these questions clearly are the ones at risk when headcount economics shift in 18 months.

Leave a comment

Which of your board and/or portfolio companies can clearly articulate what their teams built, stopped doing, or did faster because of AI last quarter?

Share The AI Leadership Edge

Hi, I’m Raffaela — and welcome to AI Leadership Edge, a newsletter for investors, business and board leaders who want to gain leverage with AI: turning board work and post-investment execution into better decisions, better alignment, and faster results. We’ve just been featured #88 rising in business by Substack 🤗

This week I am testing a new format, please let me know how you like it:

• From the boardroom: Real stories + lessons from experienced investors & leaders

• The Five-Minute Edge: Fast frameworks for when you’re slammed — clarity in 5 minutes, not perfect prep

• Leverage Stack: AI tools that actually work

• After the Deal: What great leaders/ investors do differently

• The Question: One high-leverage question

From the Boardroom

Real stories + lessons from experienced leaders, investors and board members.

The best real-world case of a company that watched AI threaten to destroy its business, and instead of playing defense, ate its own product and came out growing faster than ever - Freepik

Freepik had 500,000 paying subscribers, €61.5M in revenue, and 100M monthly users when founder Joaquín Cuenca realized everything they’d built over 12 years was about to become obsolete.

It was April 2022. DALL-E 2 had just launched. Joaquín was home with COVID, watching what people were generating. The quality jump was so dramatic that the math became inevitable: in one or two years, AI would match the quality of every image in Freepik’s catalog.

His reaction was honest. “Initially I’d say I felt some despair. I had no plan. I didn’t know what to do.”

Here’s what makes this story fascinating. Freepik wasn’t in crisis. They were at the best moment in their history. Double-digit growth. Wildly profitable. And yet Joaquín had to tell more than 500 employees that their core business was becoming obsolete. Some needed psychological treatment.

Most companies in this position play defense. Adobe’s stance was “this technology is illegal.” They had 400M images and every incentive to protect the status quo. Freepik took the opposite bet. They assumed AI would improve dramatically and decided to attack their own business before someone else did.

Strategy came after action. They started hiring an AI team without knowing exactly what to do. Within an hour of OpenAI releasing CLIP, they integrated it into their search engine. When Stable Diffusion launched as open source in September 2022, Freepik became the first non-professional editor to integrate it.

The counterintuitive result? The business accelerated. Freepik grew faster after the pivot than in the five years before. By end of 2023: 150M monthly users (up from 41M in 2021), 2M monthly active users on AI tools alone, and 120M assets created with AI. Sixty percent of new users came through AI tools. AI didn’t cannibalize their stock business. It created new demand.

Today Freepik ranks #11 in a16z’s list of the 50 most-visited generative AI products globally. Above Grammarly, CapCut, and Midjourney. A bootstrapped company from Málaga that chose to eat itself rather than be eaten.

What leaders and investors can learn:

When a technology threatens your core business, you have two options. You can protect what you have and hope the threat doesn’t materialize. Or you can assume it will and move first. Most boards struggle with this because the business is working. Revenue is up. Margins are good. Why kill what’s working?

The lesson from Freepik is that the best time to cannibalize yourself is when you’re winning, not when you’re desperate. They had 100M users who became instant distribution for AI tools. They had data on what users wanted to create. They had profitable unit economics that let them reinvest. All of that would have evaporated if they’d waited.

For boards and investors: the question isn’t whether AI will disrupt your portfolio company. It’s whether management has the courage to disrupt themselves first. And whether the board will back them when the numbers still look good but the writing is on the wall.

Share

The Five-Minute Edge

Fast frameworks for when you’re slammed — clarity in 5 minutes, not perfect prep.

Overwhelmed CEOs Don't Need More Hours. They Need This

Every business hits the same ceiling: too many opportunities, too few resources, too little time. Succes lies in using your time better over working longer hours.

Most businesses don’t have a productivity problem. They have a prioritization problem. Calendars fill with noise. Roadmaps fill with “almost important” initiatives. Over time, that busyness becomes cultural.

Here are four frameworks that apply at different levels: individual focus, team alignment, project execution, and company-wide impact.

For Individuals: The Eisenhower Matrix Map tasks on two axes: urgent and important. Four categories emerge: Do (urgent and important), Schedule (important but not urgent), Delegate (urgent but not important), Delete (neither). The discipline is protecting time for the “Schedule” quadrant before urgency hijacks it. Most leaders spend too much time reacting in the urgent-but-not-important quadrant, mistaking activity for progress.

For the Organization: Pareto Analysis Twenty percent of customers drive eighty percent of revenue. A handful of features drive most engagement. The discipline is identifying that vital minority and doubling down. The challenge is emotional: leaders must let go of initiatives that feel important but don’t move the needle. Focus is a strategic act of elimination.

For Boards and Investors: Watch how management teams prioritize when everything feels urgent. Do they protect strategic time or just react? Do they align the team or let politics decide? Do they know their critical path or hope timelines hold? The companies that scale aren’t the ones with the best strategy. They’re the ones with the discipline to execute it when the calendar is full and everyone is asking for everything at once.

Share The AI Leadership Edge

Leverage Stack

AI tools that actually work: honest reviews, workflows, integration strategies

Claude Cowork: The Tool That Triggered a $285 Billion Software Selloff

Anthropic launched Claude Cowork in January. Five weeks later, it triggered a $285 billion software selloff. The difference? You give Claude a folder, walk away, and come back to finished work. It doesn't advise—it does the job. Here are some of the tasks it can take over for you:

Share The AI Leadership Edge

After the Deal

The part nobody talks about: Where returns really come from, what breaks post-close, what great leaders, investors adn boards do differently

The AI pricing and monetization playbook: Pricing in the Era of AI Is About Outcomes, Not Access

Most AI startups are heading into a renewal crisis. They priced on potential in 2025, but as Bessemer warns, soft ROI positioning (”our copilot helps you work better”) is killing willingness to pay in 2026. The shift: charge for completed work, not access. Intercom’s $0.99 per resolved ticket works because the ROI is instant and undeniable.

For investors and boards, ask: Can management explain their unit economics including compute costs and human-in-the-loop support? If gross margins are below 50%, pricing isn’t aligned with value. The test: if a customer can’t calculate ROI in their sleep, you’re charging for the wrong thing.

Share

The Question

One high-leverage board question per issue — plus when to use it, and how to ask it collaboratively.

A soldier named Jessica Foster just hit a million followers. She posts regularly, shares her life, engages with comments. There’s just one issue: she doesn’t exist. Jessica is AI-generated. Fully synthetic, right down to her OnlyFans page. And yet people follow her. They comment. They engage. They believe.

So here’s the uncomfortable question: If a completely fabricated person can build a real audience, command real attention, and generate real engagement, what does that mean for the rest of us? What happens when the line between real and synthetic becomes so blurred that it stops mattering? And more importantly for leaders, how do you build trust, credibility, and influence in a world where your competition might not even be human?

Share The AI Leadership Edge

This week in the AI leadership Edge:

• Of course they’re putting ads in AI

• Elon and the tale of reinvention

• What directors think 2026

• $1.6m selling AI cookbooks in just 3 months

  Subscribe now

  ---

Of Course They’re Putting Ads in AI

OpenAI’s announcement that they’re launching ads for free users wasn’t surprising—it was inevitable. Ads are how the internet scales access, and AI platforms are no exception.

The economics are simple:

Most people use AI for low-value queries (emails, searches, tutoring) where $20/month feels excessive. But to scale to billions of free users, you need revenue beyond subscriptions. Enter ads: intent-based (sponsored hotel recommendations), context-based (Instagram-style product discovery), affiliate commerce (instant checkout), and goal-based bidding (pay for premium compute on specific queries).

The bigger point:

Ads aren’t evil—they’re the economic engine that powers universal access. Every major consumer platform started free and figured out monetization through targeted ads. AI is no different.

Takeaway for leaders:

For businesses, this is great news. AI ads offer concrete, embedded targeting based on conversational queries and expressed interests—something Google could never deliver. You’re not bidding on keywords; you’re reaching people mid-conversation about their actual needs. This isn’t a threat—it’s the most exciting advertising opportunity in a decade.

Share

Elon and the Tale of Reinvention

In 2005, Steve Jobs killed the iPod Mini at its peak because he saw the iPhone coming. In 2026, Musk is doing the same: Tesla is killing the Model S and Model X by mid-2026 to pivot from car company to physical AI company.

The numbers: Revenue down 3% to $94.8B (first annual drop ever), automotive revenue down 11%, deliveries down to 1.63M. But gross margins surged to 20.1%—highest in two years. Tesla’s selling fewer cars but making more on each one, sitting on $44.1B cash.

The bet: Rather than meet Chinese EV competition head-on—competitors set on undercutting until everyone else is gone—Musk opted for a complete pivot. Gen 3 Optimus launches Q1 2026, targeting 50-100K units by year-end and 1M/year by 2027-28. ARK projects the robotics market at $26T. Tesla is betting on becoming a leading B2B robotics manufacturer.

This would be bold for a startup. For a company Tesla’s size, at this level of exposure? I can’t think of anyone doing this since the iPhone. (Let me know if you have other examples.)

Takeaway for leaders: This is audacious—killing your best products to bet on an unproven future. The lesson isn’t to copy Musk’s specific bets. It’s having the courage to kill what’s working today if you see an existential shift tomorrow. Most companies cling to legacy until competitors burn it down for them. Musk is setting his own house on fire first.

Share

What Directors Think 2026

Diligent Institute’s latest report shows how directors prioritize their time, reshape oversight and prepare for the next wave of disruption. Key findings include:

• Deploying AI across the business is top of mind for directors in 2026

• Growth through M&A topped the list of organizational priorities for the year ahead

• Most boards have updated their scenario planning processes to combat new risks

• Directors want to improve oversight by having more time for strategic conversations and by having greater access to AI tools.

The report reveals a striking gap: 84% of boards strengthened their scenario planning, yet 53% don't receive real-time data between meetings. Meanwhile, 40% don't use AI for strategic oversight at all, while simultaneously making AI deployment their second-highest priority. Only 8% of boards report strong AI expertise—the lowest of any area surveyed.

Takeaway: Many boards are in for an upgrade if you ask me. But it is a world that is heavily trust based on slow to change. I’ll watch this space with curiosity.

Leave a comment

$1.6MM Selling AI Cookbooks in just 93 days

A food account generated $1.6M selling a $10 digital cookbook in three months.

The twist? It’s fully AI-powered.

The system creates hyper-realistic recipe videos with AI images, faceless voiceovers, and scroll-stopping hooks like “76g Protein” and “Athlete’s Lunch.” It posts 5-10 times daily across 15+ TikTok pages, driving 55.7M+ organic views—with most viewers believing it’s human.

Takeaway for leaders:

The companies winning aren’t using AI to assist their content teams—they’re building content systems where AI is the team. The question isn’t whether to experiment. It’s whether you can afford not to scale this aggressively while competitors already are.

That’s it for this week. The AI landscape is shifting faster than ever. The question isn’t whether to adapt—it’s whether you’re moving fast enough.

Subscribe now

If you are interested in sponsoring and receiving our media kit, message us here: https://substack.com/@theaileadershipedge

For Sponsors: Enquire about media kit

This week in the AI leadership Edge:

• Supersize CEO pay packages aren’t paying off for shareholders

• China’s Record-breaking AI-Run Drone Show Stole the Sky, without explosions

• The New Budget Economics of AI

• The Palantirization of Everything

Supersize CEO pay packages aren’t paying off for shareholders

It turns out that “moonshot” CEO pay deals have produced more smoke than stardust. Following Tesla’s 2018 precedent, companies like KKR, Rivian, Roblox, and Robinhood handed out $100M+ CEO packages hoping to ignite performance. But most of those bets have fallen flat, with few leaders delivering shareholder value in return. The takeaway: governance needs smarter incentives, not bigger ones—aligning ambition with accountability is the real rocket fuel for performance. (Wall Street Journal)

China’s Record-breaking AI-Run Drone Show Stole the Sky, without explosions

Fireworks may never look the same again. Two weeks ago, in Chongqing, 11,787 drones lit up the sky—fully choreographed by AI. No pilots, no smoke, no chaos—just 10,000+ flying robots moving as one. 🐉✨

The system coordinated every drone in real time, forming animals, symbols, and sculptures that shifted with eerie precision—zero collisions, zero humans at the controls.

It’s more than a record-breaking light show. It’s a glimpse of the future of coordination—where intelligence, not explosions, powers spectacle and scale.

For centuries, progress meant bigger bangs.
Now it’s silence, software, and synchronized machines.

The New Budget Economics of AI

Deloitte’s latest report finds that AI spending is now the fastest-growing line item in enterprise IT budgets, consuming up to half of total tech spend at some firms. AI costs are becoming as fundamental—and as unpredictable—as energy bills. Every AI model runs on “tokens,” the digital units that power its outputs. These tokens are the new currency of AI economics—as vital as kilowatt hours are to electricity, yet far harder to forecast.

• Spending patterns are changing. Costs are shifting from predictable software subscriptions to usage-based bills that rise with every AI query, model run, or chatbot interaction.

• AI costs show up everywhere—as SaaS fees, metered API calls, or through “AI factories” that balance GPUs, storage, and energy on private infrastructure.

• Smaller, experimental workloads will likely stay API-based, while larger, mission-critical AI systems move in-house to stabilize costs.

The takeaway for investors, business, and board leaders:
AI budgets can swing fast. Managing these costs is no longer just an IT task—it’s a strategic one. Leaders must understand where their AI runs, how pricing works, and when scaling up changes the economics entirely. (Deloitte)

The Palantirization of Everything

A growing number of startups now open their pitch decks with, “We’re Palantir, but for X.” They promise to embed elite engineers with customers, deliver bespoke AI workflows, and solve messy enterprise problems with “special forces” precision. It’s a seductive narrative — especially when every enterprise buyer feels overwhelmed by AI hype and vendor noise.

But most of these companies risk becoming services businesses in software disguise. Why? Because their revenue depends on armies of engineers customizing solutions one client at a time, rather than building a reusable platform that scales. Without product leverage, margins erode, and valuations built on software multiples crumble.

For leaders, investors, and directors, the takeaway is clear:

• Ask where the platform ends and the services begin. True scalability comes from repeatable architecture, not endless customization.

• Scrutinize gross margins and deployment models. If growth requires flying engineers to every client site, you’re funding consulting, not compounding software.

• Reward reusable learning. The winning AI companies will turn early bespoke work into modular, productized primitives — not perpetual client projects.

The Palantir model works because it built a platform first and deployed services as scaffolding, not the house. Copy the principle, not the headcount. (a16z)

The New Rules of Winning in Venture

The latest debate in venture—solo GPs vs. big funds, small vs. platform—is missing the point. As Forbes put it, the real differentiator isn’t structure, it’s edge: how investors compound insight, access, value, and reputation into lasting performance.

Why does this matter for leaders, investors, and boards? Because the same principle applies far beyond venture capital. Outperformance now depends on clarity of edge, not scale. Whether you’re running a company or allocating capital, it’s not about how big you are or how many people you have—it’s about knowing what unique insight, network, or capability you bring that compounds over time.

Venture firms that win the next decade won’t look alike—but they’ll all share one thing: a sharp sense of where their edge comes from and how to protect it. The same is true for every business leader navigating disruption today. (Forbes)

If you are interested in sponsoring and receiving our media kit, message us here:

For Sponsors: Enquire about media kit

2025 in Review: AI moved from “pilot” to “power”

If 2023 was experimentation and 2024 was tooling, then 2025 was the year AI became an operating model question.

The winners weren’t the companies with the most pilots.
They were the companies that built AI-first execution muscle: fast iteration loops, clear ownership, clean data, and a culture that expects AI leverage in everyday work.

1) The mega-round era got real (and strategic).
SoftBank’s planned OpenAI-led mega round at a reported $300B valuation signaled that capital is consolidating around a few “platform” model providers. Meanwhile, xAI folded X into its AI stack in a deal valuing X at $33B (and xAI at $80B) to fuse distribution, data, and models. (Reuters)
Board/Investor move: stop treating “model choice” as a tech decision. It’s a platform risk decision (cost, lock-in, data rights, exit options).

2) AI-first culture became a talent filter
AI fluency stopped being a “nice skill.” It became a baseline. Teams that moved fastest weren’t always the ones with the best strategy, but the ones with AI-native employees and managers who understood how to work with models, agents, and automation.

What leaders learned in 2025:
You can’t buy AI transformation. You build it through expectations, incentives, and managers who push it into workflows.

Board question: If a top 5% AI-native hire joined tomorrow, would they call our environment modern, or would it feel like a “fax machine with a chatbot attached”?

3) AI execution exposed weak operating systems

As soon as teams tried to scale AI, the bottlenecks showed up fast:
Poor data quality
Unclear decision rights
No workflow ownership
No measurement
No guardrails
No change management

The companies that won treated AI as an operating system upgrade, not a feature.

Board question: Who owns AI execution end-to-end (not “innovation”) and what does the weekly operating cadence look like?

4) Search started shifting from pages to answers.
a16z’s GEO framing captured what marketers are feeling: you’re no longer optimizing for ranking; you’re optimizing to be cited by models. (Andreessen Horowitz)
Board move: add one question to every growth review: “Where does our demand originate if Google traffic drops 30%?”

5) AI oversight got louder, but governance still lags.
Harvard’s roundup showed AI scrutiny rising at the board and shareholder level, especially around oversight disclosure and proposals demanding analysis of impacts. (Reuters)
Board question: What are our AI controls, and what decisions are we currently letting AI influence without structured oversight?

2026 Outlook: what I’d bet on

Here’s what I expect to matter most next year, and what to do about it.

Prediction 1: “AI-first” becomes measurable, not cultural

In 2026, you won’t be able to hide behind vision statements. AI-first will be visible in metrics:
Cycle time
Cost to serve
Revenue per employee
Customer onboarding speed
Forecast accuracy
Sales productivity

Action for leaders (30 days): Pick 3 workflows and publish a baseline: time, cost, failure points. Then redesign them assuming AI does 30 to 50% of the work.

Prediction 2: Revenue per employee becomes the boardroom metric.
AI won’t just cut cost; it will redraw what “scale” means.
Do now: track revenue per employee by function, and set an explicit target for where AI should expand capacity vs shrink headcount.

Prediction 3: Agent management becomes a real management skill

As AI agents become embedded, managers will increasingly manage “hybrid teams” of humans + agents. The highest performers will be the leaders who can turn AI into predictable throughput, not random outputs.

Action for leaders: Train managers, not just employees. Most AI rollouts fail because managers don’t change how work is assigned, reviewed, and shipped.

Prediction 4: Distribution becomes the moat again.
If capital is abundant, attention is scarce. The winners will own a channel, a community, or a workflow choke point.
Do now: ask: “What’s our unfair distribution advantage in an AI answer-world?”

Prediction 5: AI governance becomes board muscle memory

Next year, AI oversight will look more like cyber oversight: ongoing, structured, measurable. Not a once-a-year education session.

Action for boards: Require a quarterly AI dashboard with:
Top use cases in production
Value realized vs expected
Incidents and near-misses
Vendor and data risk
What’s being rolled back and why

Share

The 2026 Board Playbook: 6 questions that force execution

• Where did AI actually change outcomes this quarter (time, cost, growth), and where is it still theatre?

• Which workflows are now “AI by default” and which are blocked by data, culture, or governance?

• Who owns AI execution, and what is the operating cadence (weekly, monthly, quarterly)?

• What’s the talent plan: AI fluency by function, manager enablement, hiring profile changes?

• What’s our rollback plan if an AI layer breaks trust, accuracy, or compliance?

• If an AI-native competitor attacked our margin or customer experience, where are we most exposed?

If you are interested in sponsoring and receiving our media kit, message us here: https://substack.com/@theaileadershipedge

For Sponsors: Enquire about media kit

This week in the AI leadership Edge

• The Top 10 Books Every Board Member Should Read

• Are You Less Productive Than You Think?

• Benedict Evans’ latest “AI Eats the World Unpacked

• Tweet of the week: The Corporate AI Conundrum

If you are interested in sponsoring and receiving our media kit, message us here: https://substack.com/@theaileadershipedge

For Sponsors: Enquire about media kit

The Top 10 Books Every Board Member Should Read

I’m always on the lookout for great books, especially when putting together my holiday reading list. So I really enjoyed this list from the Enter the Boardroom team, a leading podcast for everyone involved in board work.

They compiled the books most frequently referenced by their guests — seasoned board members and leaders — that have shaped how they think about governance, leadership, risk, and long-term value creation.

Are You Less Productive Than You Think?

Many board members, executives, and investors don’t procrastinate with obvious distractions — they procrastinate with work that looks productive. Research, emails, news, decks, messages. All useful in isolation, but often a way to avoid the uncomfortable, high-stakes decisions that actually matter: strategy shifts, tough calls, people decisions, capital allocation.

This happens because board-level work is inherently ambiguous and risky. There’s no checklist for “set the right strategy” or “decide whether to back this CEO.” Our brains naturally gravitate toward responsiveness and information gathering because it feels safe, measurable, and identity-affirming — even when it quietly drains the focus needed for real judgment. The result is a loop where we stay busy, feel productive, but delay the work that creates value.

One practical takeaway: don’t rely on willpower, app deletions, or vague time limits. Instead, create explicit access windows for email, news, and updates — and block everything else by default. This shifts the burden from constantly resisting distractions to consciously opting into them at specific times. For board work, this simple structural change can dramatically improve decision quality, clarity, and follow-through.

Share

Benedict Evans’ latest “AI Eats the World Unpacked

Benedict Evans’ latest “AI Eats the World” report exposes a hard truth: while AI awareness is high, real transformation is still rare. Many people and organizations have tried AI, but far fewer have changed how they actually work. Inside companies, this shows up as endless pilots, experiments, and demos — with very little AI that people truly depend on day to day. The gap isn’t model quality; it’s adoption, integration, and leadership intent.

Every major tech wave follows a pattern — absorb → innovate → disrupt. Most companies are still stuck in the “absorb” phase, using AI to speed up existing tasks like support, coding, or content. These are useful training wheels, but they’re easy to copy and create no moat. Real advantage comes when leaders move beyond optimization and start redesigning products, workflows, and decision-making itself with AI built in, not bolted on.

The takeaway for boards and investors is clear: AI strategy isn’t a slide or a pilot count: it’s a roadmap. The winners won’t be those who automate yesterday’s work slightly faster, but those who ask the harder question:

What can we now do that was previously impossible?

Share

Tweet of the week: The Corporate AI Conundrum

Last quarter I rolled out Microsoft Copilot to 4,000 employees.

$30 per seat per month.

$1.4 million annually.

I called it “digital transformation.”

The board loved that phrase.

They approved it in eleven minutes.

No one asked what it would actually do.

Including me.

I told everyone it would “10x productivity.”

That’s not a real number.

But it sounds like one.

HR asked how we’d measure the 10x.

I said we’d “leverage analytics dashboards.”

They stopped asking.

Three months later I checked the usage reports.

47 people had opened it.

12 had used it more than once.

One of them was me.

I used it to summarize an email I could have read in 30 seconds.

It took 45 seconds.

Plus the time it took to fix the hallucinations.

But I called it a “pilot success.”

Success means the pilot didn’t visibly fail.

The CFO asked about ROI.

I showed him a graph.

The graph went up and to the right.

It measured “AI enablement.”

I made that metric up.

He nodded approvingly.

We’re “AI-enabled” now.

I don’t know what that means.

But it’s in our investor deck.

A senior developer asked why we didn’t use Claude or ChatGPT.

I said we needed “enterprise-grade security.”

He asked what that meant.

I said “compliance.”

He asked which compliance.

I said “all of them.”

He looked skeptical.

I scheduled him for a “career development conversation.”

He stopped asking questions.

Microsoft sent a case study team.

They wanted to feature us as a success story.

I told them we “saved 40,000 hours.”

I calculated that number by multiplying employees by a number I made up.

They didn’t verify it.

They never do.

Now we’re on Microsoft’s website.

“Global enterprise achieves 40,000 hours of productivity gains with Copilot.”

The CEO shared it on LinkedIn.

He got 3,000 likes.

He’s never used Copilot.

None of the executives have.

We have an exemption.

“Strategic focus requires minimal digital distraction.”

I wrote that policy.

The licenses renew next month.

I’m requesting an expansion.

5,000 more seats.

We haven’t used the first 4,000.

But this time we’ll “drive adoption.”

Adoption means mandatory training.

Training means a 45-minute webinar no one watches.

But completion will be tracked.

Completion is a metric.

Metrics go in dashboards.

Dashboards go in board presentations.

Board presentations get me promoted.

I’ll be SVP by Q3.

I still don’t know what Copilot does.

But I know what it’s for.

It’s for showing we’re “investing in AI.”

Investment means spending.

Spending means commitment.

Commitment means we’re serious about the future.

The future is whatever I say it is.

As long as the graph goes up and to the right.

By Peter Girnus

Share

Sponsorship opportunity: At 10.600 subscribers, we are now looking for sponsors to help us make the AI Leadership even better.

If you are interested in sponsoring and receiving our media kit, message us here: https://substack.com/@theaileadershipedge

For Sponsors: Enquire about media kit

Powered by BoardLens, the intelligence platform for board directors and investors. It ingests board decks, financials, and updates to build a living knowledge graph that tracks what changed, what was delivered, and what matters. Every portfolio company. Every quarter. Automatically. https://boardlens.ai

This week in the AI leadership Edge

• The Post-AI Org Chart

• The New Metric That Redefines Companies: Revenue per Employee

• Can AI Boards Outperform Human Ones?

• In a world where capital is a commodity, attention is the new advantage

  ---

  Join 10.300 investors, business leaders and board directors for decisive AI leadership tools and actionable insights

The Post-AI Org Chart
For 100 years, companies have been built as pyramids — one leader, layers of managers, many individual contributors.
AI breaks that model.

Two new structures are emerging:

1️⃣ The Short Pyramid
Managers oversee AI agents instead of large teams.
Headcount drops ~80%, but human leverage skyrockets.
Everyone moves one level up the value chain — from doing work to directing, auditing, and improving AI outputs.

2️⃣ The Rocket Ship
Every human becomes a player-coach, managing both people and agents.

They still execute work themselves but they also manage a set of AI agents that amplify their output. Instead of traditional hierarchies where managers only manage people, this model blends human execution + AI leverage.
Speed increases dramatically. Perfect for high-velocity teams and startups.

Boardroom takeaway:
Org design is strategy.
Leaders must ask:
Are we structured to capture AI leverage — or constrained by old hierarchies?

Share

For Sponsors: Enquire about media kit

The New Metric That Redefines Companies: Revenue per Employee

What the post- AI org chart hints at: AI is making massive teams obsolete.

The new metric to watch is revenue per employee. High-agency teams that move quickly and automate aggressively will outcompete companies weighed down by meetings and politics.

As Y-Combinator’s Aaron Epstein notes, with the right AI tools, ten exceptional people could build a hundred-billion-dollar company. In the AI era, speed and focus are the new scale.

Can AI Boards Outperform Human Ones?

A new Wharton and INSEAD study found that AI “boards” outperformed human ones across nearly every governance metric, making faster, clearer, and more data-backed decisions. Human boards, by comparison, struggled with bias, time management, and collaboration.

Share

In a world where capital is a commodity, attention is the new advantage

In a world where capital has become a commodity, the real battle is for attention, brand, and narrative power. That’s why Andreessen Horowitz just launched a New Media team — a move that signals something bigger than marketing: Founders, Partners, Investors gravitate to the strongest brands.

A16Z understands this better than anyone.
Their New Media play adds another layer of irresistible value:
distribution + influence + platform power on top of capital.

In the AI era, whoever controls attention doesn’t just win deals — they shape the market.

To all investors, leaders and boards: What is your unique media strategy?

Share

We just crossed 10,304 subscribers — thank you for helping push us past the 10k milestone. It’s incredible to see this community of leaders, investors, and board directors grow.

As we scale, we’re now looking for partners and sponsors who want to bring practical, high-impact AI leadership insights to this audience of 10,304 decision-makers. If you’d like to collaborate, please reach out.

For Sponsors: Enquire about media kit

The Top 10 Books for leaders, directors and investors Required reading for boards, the truth about productivity, AI’s macro disruption unpacked — and a viral reminder of how not to run an AI transformation.

The 2026 Playbook: Six Questions That Force Execution Six questions for 2026 that force execution, and what 2025 proved about AI-first operating models, metrics, and governance.

The Advisory Board Squeeze (And Why AGI Is Closer Than You Think) Does Your Advisory Board Make You Uncomfortable? Becoming AI fluent investor edition and the napkin test for leadership gaps