Subscribe now

This week’s FullStack HR is brought to you by…HR Tech Europe.

We need to talk to each other. For real. The pace is faster than ever, none of us can crack this alone, and we need rooms full of smart people who challenge how we think.

Amsterdam, April 22-23. Google, LEGO, Coca-Cola, Microsoft, Maersk, Oatly, and a hundred other companies!

If you’re in HR, it’s free.

Register here

Happy Tuesday,

Quick update before we get to this week’s article.

Despite my brilliant move of putting the wrong link in last week’s description, the interest for the Cowork Bootcamp for HR has been completely insane. So I’ve decided to run it in two rounds. Simple as that. If you missed the first round (signups closed April 11), you can still join the second. I’m offering 30% off as a beta 2 discount. I’m fully convinced you’ll be happy you signed up.

Sign up here

(Yes, the link is the correct one)

Next week I’m at HR Tech in Amsterdam. And I want you to come build things with me!

Bring your laptop. We’re not doing slides and theory. We’re building real HR tools, live, together. This is the most practical session I’ve ever done at a conference.

Two sessions,

April 22 (12:30–13:30, Ask the Experts & Think Tank) Ask the Experts: Using AI to Build Practical HR Outputs

April 23 (15:00–15:45, AI Labs) AI Lab: From Prompts to Products: Building Real HR Tools with AI

If you’ve been reading this newsletter and thinking “I should really try this stuff,” this is your moment. Show up. Bring your laptop. We build!

Now. To this week’s article.

I’ve lost count of how many workshops and keynotes I’ve done this year. I don’t even want to count. But every session is a data point, and every discussion sharpens the picture. Things I believed six months ago, I don’t believe anymore. The tech moved forward, and I got better at seeing what actually works in organizations and what doesn’t.

The thing that is clearest of all? Leadership. Again.

If your CEO isn’t crystal clear on why AI matters, what you’re actually going to do, and how you’re going to do it, almost nothing will happen. You get this half-hearted energy where the organization acts as AI might possibly become important someday. And if leadership isn’t prepared to make real changes to how work gets done, nothing moves. I’m more convinced of this every week.

The capability isn’t the constraint anymore.

The models we have right now, whether it’s GPT 4.5 Pro, Opus 4.6, or the agentic capabilities in Claude Cowork, don’t matter. They are already more capable than what most organizations are extracting. Most companies aren’t using a fraction of what’s available.

Do they still hallucinate? Yes. Do they make mistakes? Absolutely. But if we applied the same forgiveness factor to AI agents that we apply to human employees, the gap would be indistinguishable.

The constraints are the organization and the people within it. The CEO who says “AI is important” but doesn’t mean it. Who hears “we bought Copilot” and accepts that as a strategy. But for the 1023rd time in this newsletter, this isn’t a technology purchase. This requires working through your entire organization, rethinking how work gets done, making real technology choices, and getting every level to understand what’s actually possible versus what’s just old mythology that won’t die.

The excuse factory is running at full capacity.

I meet so many organizations where old beliefs still live. “This won’t work.” “It’s not safe.” “The models can’t do that.” Sometimes it’s fear, sometimes ignorance, sometimes professional pride, sometimes politics. People who won’t make decisions and hide behind GDPR concerns or security worries that sound legitimate but are really just shields for inaction. And sometimes, honestly, people just aren’t capable enough to evaluate what they’re looking at. I know that sounds harsh, but it’s true.

So the CEO isn’t confident enough to push through it. They listen to the IT director protecting their territory. They listen to the HR director who doesn’t feel confident in the technology. And nothing moves. “Let’s just go with Copilot. It feels safe.” You shouldn’t make reckless decisions. But the cost of not making decisions gets higher every single day.

Courage. Year three.

I’ve been banging this drum for three years. You have to dare. There are ways to make it safe, but that requires investment. So either you invest in making it secure, or you accept that it’s maybe 90% there and move forward anyway. Do your GDPR assessments. Be compliant. But know what is what and get good advisors.

On that note. I always want to focus on the positive, on what you can do. But there is so much bad advice floating around this AI era, from both internal and external sources. I’m not exempt; I have blind spots, too. But the gap between good and bad guidance right now is enormous, and the wrong advice at the wrong time can cost you a year or more. Get people around you who actually know what they’re talking about.

The real risk.

It doesn’t matter if you’re an industrial company, a consulting firm, or a tech company. Someone, somewhere, is sitting with Claude Cowork right now, automating their support processes, operating at lower margins than you. And they will come for your market share. Not overnight but steadily.

So either you believe you’re fast enough to pick this up when it matters, or you start now, or you accept you’ll invest like crazy later to catch up. But believing you won’t need to do this work at all? That’s Nokia. That’s Ballmer-era Microsoft. That’s standing outside the internet and saying “we don’t believe in mobile phones.”

I’ve been conservative about this. I’ve tried to balance hype with reality. But enough is enough.

$20. That’s it.

Claude Cowork costs $20 a month, and you spend maybe 1-2 hours with it. That’s all it takes to see what this technology is capable of. The potential to fundamentally change how work gets done is real, and there’s a real risk that this leads to jobs disappearing. Which is exactly why, if you’re a leader or in HR, it feels wrong not to test this. If you can’t do it at work, do it privately.

I’m not saying this so you’ll buy my course. YouTube is overflowing with free examples. The point isn’t the course. The point is that you start. And the worst that can happen? You come back and tell me I’m a hype boy. Fine. At least you tested it, at least you formed an opinion.

I’ve given a lot of presentations this year. I haven’t met a single person who, after actually trying Claude Cowork, said “that looks useless.” Not one.

And, yes, this is coming to Copilot too. Eventually. So start now, and it’s downhill. Wait a year, and it’s uphill.

We need to lead this. Not stumble into the future because we were too scared to decide. We need to steer our organizations using

Maybe this was too much of a rant, I don’t know, but I’m getting a bit tired of just sitting around waiting.

Subscribe now

This week’s FullStack HR is brought to you by….HR Tech Europe!

More than ever, we need to talk to each other.

Not “align on strategy” talk. Talk for real.
The topic right now in HR? It’s one we all need to learn. Fast. And the need has never been this big, the pace has never been faster.

I’m a 100% certain that we can’t figure this out alone. We need rooms, conversations, and smart people who challenge how we think.

That’s happening in Amsterdam, April 22-23 at HR Tech Europe.

People from Google, LEGO, Coca-Cola, Microsoft, ING, Maersk, Oatly, and a hundred other companies will be there.
is doing three sessions this year, including his first ever "Ask the Experts" and a unique session with Gianpaolo from Cisco. And a brand new format called HR Tech Intensives, where you actually work through problems together instead of just listening.

And of course, I’ll be there, and I’ll be hosting a practical session, so bring your laptops, and we’ll build stuff together.

If you’re in HR, it’s free.

Come talk.

Register here

Happy Thursday,

I messed up on Monday and put the wrong link to the Claude Cowork Bootcamp…yey me. This is the right link….sorry!

But enough of all of this, let’s get to it!

Every time I give a talk about AI, someone asks it.

Sometimes early, sometimes at the end, sometimes with a hint of accusation, and sometimes out of genuine curiosity. But sooner or later, the question pops up: “What about the environmental impact?”

I’ll be honest, sometimes I brushed it off with a half-answer.

Not because the question is bad, but because I never quite understood why AI specifically became the big symbolic issue for digital environmental impact. We’ve been streaming Netflix for twenty years, Zoomed our way through a pandemic, and googled everything from symptoms to holiday destinations without anyone raising their hand to ask what it’s doing to the planet.

But the moment someone opens ChatGPT, the entire climate discussion seems to land in their lap.

This question has been nagging at me for a while, though, and I’ve spent quite a few hours with ElevenReader in my ears going through the IEA’s energy reports, Stanford HAI’s AI Index, OECD publications, and peer-reviewed papers trying to form a proper opinion. I also used Claude to help me synthesise and structure all of it into this article, because the irony of using AI to write about AI’s environmental footprint felt too appropriate to pass up. The sources and data are real, the position is mine, and the help was computational.

What I found was considerably more nuanced than the public debate usually suggests.

The individual footprint is vanishingly small

A typical text query to ChatGPT uses roughly 0.3 watt-hours according to Epoch AI’s analysis from February 2025. OpenAI’s Sam Altman has confirmed a similar figure of 0.34 Wh. Google published in its own methodology report in 2025 that a median text prompt in Gemini uses roughly 0.24 Wh, produces 0.03 grams of CO₂, and consumes 0.26 milliliters of water.

That puts it in the same ballpark as Google’s own estimate for a standard search, roughly 0.3 Wh, though that figure comes from a 2009 blog post and is likely lower today given sixteen years of hardware improvements.

The comparison isn’t perfect, but the order of magnitude is clear. A typical AI text query is not the energy monster that early estimates suggested.

The figure that often gets thrown around, that an AI query uses “ten times more electricity than a search,” comes from Alex de Vries’ article in Joule in October 2023 and was referenced by the IMF, among others. It was likely accurate at the time, but is now outdated for typical text queries, since both hardware and model architectures have become dramatically more efficient.

Digital energy comparisons are harder than they look

It’s tempting to line up AI alongside streaming, video calls, and email to show that it’s a relatively small activity. And I’ve done this several times myself, and my early drafts of this article had a neat little table with watt-hours per activity. But the more I dug into the sources, the more I realized that these comparisons are difficult to make.

The IEA’s own analysis from 2020 found that one hour of Netflix streaming consumed roughly 77 Wh, which was already about 90 times lower than the widely cited Shift Project estimate that had made headlines.

The result is highly sensitive to what you count, such as that device type matters enormously (a TV uses roughly 100 times more electricity than a phone), and whether you include only the data center or the full chain from server to screen changes the answer dramatically. Carbon Brief’s deep dive confirmed the IEA’s lower range and showed just how wildly the numbers vary depending on assumptions.

For video calls and email, the source picture is even weaker. A Purdue/MIT/Yale study from 2021 gave a range of 150 to 1,000 grams of CO₂ per hour of video conferencing, which is so broad it essentially tells you that nobody knows. For email, I found estimates ranging from 0.01 Wh to 13 Wh per message, a spread of more than a thousand times.

Most digital energy comparisons you see in the press blend methodologies, mix old and new estimates, and conflate different system boundaries, making them look far more precise than they are. What we can say is that a typical AI text query, at roughly 0.24 to 0.34 Wh, is a very small unit of energy by any reasonable standard, and it’s comparable to a lot of other digital activities that we do.

But the scale is a different story

What makes the question impossible to dismiss is not the individual prompt but the total infrastructure. According to the IEA’s “Energy and AI” report from April 2025, the world’s data centres consumed roughly 415 terawatt-hours of electricity in 2024, about 1.5 percent of global electricity use. In the IEA’s base scenario, that more than doubles to around 945 TWh by 2030, equivalent to Japan’s entire electricity consumption today.

Five companies plan to spend close to 700 billion dollars on AI infrastructure in 2026 alone. ChatGPT reached 900 million weekly active users by February 2026. AI is being built into search engines, email clients, office software, and autonomous agents in a way that makes it an entirely new computational layer underneath the digital economy, and that layer requires physical infrastructure on a massive scale.

The efficiency gains are real and impressive. Stanford HAI documents a 280-fold cost reduction for GPT-3.5-level inference in just two years. But it is precisely this efficiency that drives even faster growth in usage, exactly as Jevons paradox predicts.

When GPT-4o’s API pricing dropped from 30 dollars to 2.50 dollars per million input tokens, usage exploded. Google’s Gemini token usage grew 130 times in 18 months. The IEA’s scenarios show that even with significant efficiency gains, absolute electricity consumption continues to rise in nearly all plausible futures.

Water?

Energy gets almost all the attention, but the water question is in many ways more acute.

Before getting into the alarming cases, it’s worth putting AI’s water use in proportion. While I was writing this article, my friend Mathias Sundin published an excellent piece on Warp News covering much of the same ground, building on work by Andy Masley who has done some of the most detailed analysis on this topic.

According to Masley’s calculations, all of America’s golf courses use roughly 5,474 times more water than ChatGPT. (Yes, you read that right 5474 times more than the ENTIRE ChatGPT.)

In Maricopa County, Arizona, golf courses account for 3.8 percent of the county’s water while data centres account for 0.12 percent. There are individual golf courses that consume more water than all of ChatGPT globally. And the widely shared claim that a single AI prompt uses “a bottle of water” is a distortion of a study that found 500 millilitres per 20 to 50 prompts, roughly 0.3 millilitres per query.

That said, proportion isn’t the whole story. As Undark reported, the issue is not that AI currently uses more water than golf or agriculture. The issue is the pace and concentration of growth, and who bears the local cost. The World Resources Institute reports that two-thirds of data centres built since 2022 are in water-stressed areas.

Both things are true simultaneously. AI’s total water consumption is small compared to many other industries. And in specific communities where data centers are being built at speed, the local impact is already causing real conflict.

So where does this leave us? Or me?

I still don’t really understand why AI specifically became the big symbolic question for digital environmental impact. We’ve accepted that digitalization costs energy for decades, and the individual AI query is energetically comparable to things we do hundreds of times a day without a second thought.

I’m also not exactly a model citizen when it comes to the environment. I fly too much, and I know it. So I’m not going to stand here moralizing about other people’s Claude habits.

But I think the question deserves a serious answer rather than a shrug.

Data centres currently account for roughly 0.5 percent of global CO₂ emissions, compared to aviation at roughly 2.5 percent and the textile industry at 2 to 8 percent. AI is not the world’s biggest environmental problem. But data centre electricity is growing at roughly 15 percent per year, it is one of very few sectors where emissions are projected to keep rising.

The serious answer, the one that best reflects what the IEA, the OECD, Stanford HAI, and UNEP say, is that AI’s environmental footprint is real and growing fast, but it is not a climate catastrophe in itself.

It is a serious governance problem that requires us to demand better infrastructure, not to stop using the technology.

The IEA also estimates that AI applications could enable emission reductions equivalent to 5 percent of global energy-related emissions by 2035. But that doesn’t happen by itself, and it doesn’t happen if we build AI infrastructure on dirty power in water-stressed regions without asking questions.

I think that the question shouldn’t be “should I use AI if I care about the environment?” The question should be why we don’t ask the same thing about everything else we do on a computer, and why we’re not demanding more from the companies building the infrastructure.

I’m going to keep using AI. I’m going to keep talking about it as the most transformative technology most organizations have to deal with right now. But I’m going to stop brushing off the environment question with a shrug. It deserves better than that.

I save hours every week using Claude Cowork.

It’s done through automated background processes. It drafts my emails, analyzes my data, and helps me build my presentations, while I do other things.

And I built a bootcamp that teaches people like you, HR professionals, how to set up the same thing.

I’ve sold exactly one course in 7+ years of writing this newsletter. Sure, I’ve been sneaking in the occasional links to what I do and hoping you’d click them. But this is new territory. I’m deliberately, in broad daylight, selling you something through a dedicated post.

Why? Because I feel so strongly that people should learn this.

So what’s the fuzz about?

Cowork Bootcamp for HR.
3 weeks. 3 live sessions.

Automated job postings, survey analysis, onboarding flows, and email responses. If you can articulate it, Cowork can (probably) automate it.
All are built as reusable skills that continue to work after the bootcamp ends.

Not “prompting tips.” Not theory. Systems and automations.

The gap between “I’ve tried ChatGPT” and “AI saves me 5 hours a week” is not talent or tech skills. It’s someone showing you exactly how to set it up for YOUR work. That’s what I’m hoping this will be.

The details

When: April 17 - April 21 - May 6
Format: 3 live sessions, 60 min each
Time: 3:00 PM Stockholm / 2:00 PM London / 9:00 AM New York
Price: $490 (beta price, will be at least $990 from cohort 2)
Requires: Claude Pro ($20/mo)

This is in English, built for a global HR audience.

Not convinced after Session 1? Email me. Full refund. No questions.

Who This Is For

You work in HR. You know AI can save time but haven’t cracked the code on making it actually useful. You want to build, not just learn.

And if your organization runs Microsoft 365, two weeks ago, Microsoft launched Copilot Cowork, built on the exact same Claude technology you’ll learn in this bootcamp. The skills and workflows you build here will translate directly when your organization rolls it out. This is your chance to be the person who already knows how it works before IT even sends the rollout email.

Who This Is NOT For

If you want a certificate for LinkedIn, a passive webinar, or slides about “the future of AI in HR,” this isn’t it. You will build things. That’s the point.

Curious?

→ Apply here ←

Very limited seating since this is the beta cohort. I review every application personally and accept on a rolling basis. Apply early, as I fill it out as we go. The cohort may fill before the April 11 deadline.

This is the intersection of 15 years in HR and 3+ years going all-in on AI. It’s the thing I wish existed when I started. If that resonates, apply.

/Johannes

P.S. Company paying? I do invoicing.
Need a business case for your manager? Reply to this email. I’ve written about 100 of those.

P.P.S. Once again, session 1 guarantee. Full refund. That’s how sure I am.

Subscribe now

This article is part of a paid collaboration with Sana.ai, which made my attendance possible. Thank you, Sana!

Unleash holds a special place for me. My very first professional HR conference, almost exactly 15 years ago, was the event that would eventually become what we now know as Unleash. Back then it had a different name. Back then it was in Amsterdam. And back then, the world of HR looked nothing like it does today.

This year, I went back to Las Vegas. And I came home with a lot to think about.

Why I went

The primary reason I chose to travel halfway across the world was one name: .

If you have been reading this newsletter for any amount of time, you know that Mollick has been a massive inspiration for me and for many others in this space. The reason is simple. He is one of a very, very small number of people who talk about AI from an organizational perspective, and what he does so well, and what I have genuinely tried to emulate in my own work, is that he embraces the fact that very few of us actually know what is going on. He said it himself on stage: “If you’re doing imposter syndrome right now, you’re in the exact right place. We are all imposters.” Things are moving too fast to follow everything. What you can do is try, experiment, and apply it in your own context. And that is exactly what he does, brilliantly.

Amy Edmondson was the other name on my list flying over the pond. Her opening keynote on Day Two delivered what you would expect from someone who has spent decades studying psychological safety. And I enjoyed her framework around “intelligent failure,” that organizations should deliberately create conditions where the right kind of failure can happen, felt like it was written for this exact moment. Every organization trying to figure out AI right now needs to hear that.

The AI Summit

The conference started with the AI Summit, organized by UKG and led by and .

The AI Summit was almost unfairly good. It was practical, concrete, and forward-looking in a way that set the tone for the entire rest of the conference. Did it contain a lot that was completely new for someone like me, who spends essentially all waking hours trying to understand this space? Maybe not. But even for someone deep in the weeds, there is always something new to pick up when you are in the room with other people thinking hard about the same problems. That is the real value of these events. Always has been.

The key message from both Jason and Jess was this: read the signals. Pay attention to where things are actually heading, not where you wish they were heading.

Peter Hinssen stole the show? Yes.

Then came Day Two. And the highlight, by far, was Peter Hinssen.

I have seen Hinssen speak three times now. There is a lot happening when he talks, and there is a lot of dopamine when you listen to him. But this time, he was something else entirely. I would put his keynote in my top three talks I have ever seen on a stage. (That is not something I say lightly!)

Why was it so good? Because he puts things in perspective. He weaves together the technical realities of what is coming with the organizational implications of what it means. He makes the complex feel simple to follow. His concept of the “Never Normal,” that we should stop waiting for things to calm down because they never will, landed with particular force this year. He framed the current moment through four lenses: hardcore geopolitics, extreme capitalism, zero latency, and talent singularity. His conclusion was clear. AI-first does not mean people-last, but the waves of disruption are only going to get higher.

Budget planning is a yearly, sadistic corporate ritual where people put fake news in Excel that is consolidated into something which actually never works. - Peter Hinssen

For me, Hinssen and Mollick together represent the gold standard. They combine technical depth with organizational realism, and they are honest about what they do not know. One line from Hinssen has stayed with me since I left the room: “Digital was just the appetizer. AI is the main course.”

But Mollick?

Ok, so Hinssen was a highlight. But what about Mollick then? He did not disappoint.
And he did not just talk about where AI is heading. He gave the room a practical framework for what to do about it, and his model has three components. Leadership, Lab, and Crowd.

The Crowd is everyone in your organization who is already using AI. And they are using it, whether you know it or not. Mollick was blunt about this (rightfully so!) Almost half of employees are already using AI at work. Many report tripling their output on certain tasks. But none of that productivity is showing up on your dashboards. Why? Because your AI policies have made it scary to admit. Because people know what “efficiency gains” means in corporate language. Because there is no incentive to tell you. As Mollick put it: “The expertise of how to use AI is in your teams already.” The question is whether you are creating the conditions for that expertise to surface.

The Lab is not an IT function. It is a small, dedicated team that works on AI full-time. Their job is three things. First, take what the crowd discovers and ship it to the rest of the organization. Someone builds a prompt that saves six hours a week? That goes out to everyone, that week. Second, build benchmarks. How good is AI in your specific context? You will not know unless you test it yourself. No outside evaluation will tell you. Third, build the impossible things. Experiments that might not work. Provocations. The stuff that prepares you for what is coming next.

Leadership is the third piece, and Mollick was clear that he has not seen a single company succeed with AI without all three in place.

AI is a weird alien mind, one that isn't sentient but can fake it remarkably well. It is trained on the vast archives of human knowledge, and also on the backs of low-paid workers. - Ethan Mollick

One more thing he said that landed hard: “Your R&D team is HR.” Not IT. Not a vendor. Not a consultant. The people who will figure out how to use AI in your organization are the people who already work there. This is not something you can outsource. And the models are universally available. There is no secret AI. Everyone has the same tools. The difference is what you do with them.

And yes, we got to meet him, and yes, I was starstruck.

All good?

There is a palpable uncertainty in the HR community right now. You could feel it in the hallways. Everyone is looking at everyone else. What are you doing? What are you doing? What are you doing? There is a restlessness, almost an anxiety. Where is this going to land?

I want to be careful about how I say this. I know it might sound presumptuous. But I see what is happening with AI, and sometimes I feel like I am losing my mind because the gap between what is already possible and what most organizations are actually doing is enormous. So it was… reassuring to hear people like Hinssen, Mollick, Averbook, and Von Bank confirming the same thing. This is not a phase, there are no signals suggesting that AI disruption is going to slow down or go away.

That confirmation matters more than I expected it to and it connects directly to something I experienced that same week.

Snap back to reality.

The same week I was at Unleash, I was also working with an American client. I cannot name the organization, but I can tell you what happened. We used Claude Cowork extensively, live, inside a real HR operations environment. The work involved processes that most HR professionals would recognize immediately. Document handling, policy reviews and data consolidation across systems. The kind of recurring, high-volume knowledge work that fills up calendars and drains energy (and motivation).

In roughly 30-45 minutes of focused work with Cowork, I estimate we identified efficiency gains of hundreds of hours per month. Not theoretical gains. Real, measurable, immediate time savings on actual work that actual people were doing.
Cowork is a magical tool. It truly is.

And that’s yet another signal. And by the wya, here the signals all point in the same direction. Microsoft is working on their own version of this kind of tool. There are credible reports that OpenAI is building something similar. Anthropic already has Cowork in production. This is coming. Broadly. In ways that are compliant and secure.

And when it does arrive at scale, knowledge work as we know it will change. Not might change. Will change. The person who is not prepared for this will struggle. I am more and more convinced of that with every passing day. The signal is becoming clearer, and clearer.

Ok, so now what?

Does the work still require humans? Absolutely. Humans are critical. Validation matters. Judgment matters. Context matters. But the tasks themselves, the actual work of producing, analyzing, summarizing, drafting, processing? Machines can already do most of it. Right now, today. If you let them.

The limitation is not technology. Both Hinssen and Mollick made this point forcefully. The limitation is us. It is organizational inertia. It is perception. It is the comfort of waiting for the dust to settle. Or as Hinssen put it: “Yesterday’s work is the silent killer of organizations.” We have to become yesterday’s work hunters in the age of AI.

The dust is not going to settle.

That, more than anything, is what I took home from Las Vegas.

Subscribe now

I’ve spent three editions on bosses.
→ What an AI boss could look like.
→ What’s left when AI handles the managing.
→ Why we expect bosses to care in the first place.

I built the case. AI can individualize at scale, better than most humans can. It doesn’t get tired, it doesn’t have moods, and it remembers your goals. For many people, it could be the best boss they’ve ever had.
And yes, I still believe that, or parts of it at least.

But I’ve been arguing from the manager’s and the organization’s side.
What about the person on the receiving end?

How does it feel, or would it feel, to be managed, coached, followed up on, and given feedback by a machine?

We already know something

This isn’t entirely hypothetical. Millions of people already work under algorithmic management, like Uber drivers, Amazon warehouse workers, and food delivery couriers. Their schedules, performance ratings, and in some cases, terminations are shaped by software.

Not all of that is “AI” in the ChatGPT sense, but it’s still machines making managerial calls, and I do believe that some of the lessons transfer.

And the research is pretty clear. How it’s designed (and governed) matters more than what it is.

A 2024 study of food delivery couriers in Finland (based on 30 interviews), “Algorithmic management, wellbeing and platform work”, found that the work had what the researchers described as “iso-strain” characteristics: high demands, low control, and low support, plus isolation. Workers described stress, frustration, and mistrust, especially when decisions felt opaque or unfair.

And one detail matters here, and that was that when couriers reached out for help, they often met automated, algorithm-driven responses that didn’t address their concerns. That’s where a lot of the damage happened.

To be fair, the study also notes nuance here; support on issues directly related to deliveries could be fast and adequate. But broader support and real dialogue were limited.

The algorithm wasn’t the problem. The design was.
(And one could argue that this was not the latest and greatest models from the frontier labs, thus the study was set in the wrong premise.)

The loneliness chain

A 2023 paper in the Journal of Applied Psychology, “No Person Is an Island”. The researchers ran four separate studies with a total of 794 participants across Taiwan, Indonesia, Malaysia, and the United States.

The finding was that the more employees interacted with AI systems for work, the lonelier they felt.

Not metaphorically lonely. Lonely ad in the kind of loneliness that spilled into life outside work. More insomnia and, in some of the studies, more after-work alcohol use.

It’s important to emphasize that the loneliness was self-reported. But some of the spillover wasn’t. In parts of the research, coworkers and cohabiting family members also reported changes like less helping at work, and more insomnia. Alcohol results were more mixed across the studies.

The researchers linked this to social deprivation. When you spend your day interacting with a system that can’t reciprocate connection, you can end up with a social “deficit”. You still need affiliation and to belong.

A 2025 study in Behavioral Sciences, “Effects of Employee AI Collaboration on CWBs”, tested a similar chain in a vignette experiment. Employee-AI collaboration increased loneliness, which led to emotional fatigue, which led to counterproductive work behavior.

But there was a catch. When leaders provided emotional support, the negative effects were weaker. The leader’s care interrupted the chain.

Which raises a question - if we need humans to offset the loneliness AI can create, have we actually saved anything?

The belonging question

Work isn’t just about productivity. It’s about being part of something.

When I was at Spotify, Kry, or any of the other companies I’ve worked for, what kept people engaged wasn’t just the work. It was the team. The manager who noticed when you were struggling. The colleague who grabbed coffee with you after a hard meeting. The sense that you mattered to someone.

I see the same thing now in the organizations I work with. Last month, I ran a workshop where we discussed this topic, and a manager said something along the lines of: “My team doesn’t need me for the tasks. They need me for the ‘are you okay?’ after the meeting, when I can tell something’s off. I’m not sure AI can replicate it.”

Can it try? Sure. Can it do a version of it? Probably.
But the research so far suggests that as of today, something gets lost in translation.

The surveillance problem

But wait, can it actually do that? Mimic the above? Well, I touched on this in the last article, but it deserves more space here. Because the same AI that can provide personalized coaching can also track your keystrokes, analyze your calendar, measure your response times, and decide you’re “less productive” than last quarter.

Amazon warehouse workers know this reality. Pace and performance can be tracked closely, and those metrics can drive warnings and consequences. (For example, reporting on “Time Off Task” systems and automated warnings: The Verge.)

And we don’t need to pretend this is only about one company. A 2023 review in the International Journal of Environmental Research and Public Health, “Workers’ Health under Algorithmic Management”, summarized how algorithmic management can influence job quality factors linked to health and well-being.

When systems monitor without explaining, evaluate without transparency, and discipline without appeal, workers experience what many call “digital Taylorism.” The efficiency gains go to the organization, and the stress goes to the worker.

I mentioned Frederick Taylor in the last article. In Shop Management (1903), he wrote that “all possible brain work should be removed from the shop” and centered elsewhere.

Digital Taylorism is the same idea with better technology. And it’s already here.

So what should we do?!

I could paint a picture here of how AI management done well might look. The empathetic check-in, the personalized career development and the transparent feedback loop.

But I already did that in the three articles mentioned in the beginning. What I think is more useful, instead, is this. What’s the minimum standard we should accept?

My point of view is this, that if your organization is deploying AI in any management capacity, these four should be non-negotiable, and I think we need to cater for them to happen, to be explicit.

Transparency. Workers should be able to see why AI is giving them certain feedback or making certain decisions. “The algorithm decided” is not an explanation. If you can’t explain it, you shouldn’t deploy it.

Appeal. There has to be a human you can escalate to. Always. No exceptions. If someone’s performance rating, schedule, or employment status is affected by an algorithm, they need a path to a real person who can override it.

Voice. Workers need input into how these systems work. Not after deployment. During design. The people who will be managed by AI should have a say in how it manages them.

Boundaries. Define what AI can and can’t do, and what data it can and can’t collect. Can it give feedback? Fine. Can it fire someone? That needs a different conversation. Draw the lines before the technology makes the decision for you.

These aren’t ambitious goals; they’re the bare minimum as I see it. And many organizations deploying algorithmic management aren’t meeting them, nor thinking about them. But I strongly believe we need to keep all of them in mind when thinking about AI and leadership.

What I don’t know

All this said, not only the last part here, but thinking about all the articles I’ve written, I still feel uncertain about being managed and led by an AI. I don’t know how I’d feel being managed by AI.

Part of me thinks I’d appreciate the consistency. Less politics, less favoritism. Feedback that doesn’t depend on my manager’s mood.

Part of me thinks I’d miss having someone “who knows me”. Not “knows my data” but knows me. Who I can read, who might bend the rules when I need it. I’m not even sure if that’s rational, but it’s real for me at least.

Part of me suspects it depends entirely on whether the people who design these systems give a damn about the people who use them.

And that’s the question I want to leave you with.

When you deploy AI to manage people, whose experience are you optimizing for?

The answer to that question will shape what work feels like for millions of people. And it’s being answered right now, in every organization rolling this out.

Whether anyone is asking the question or not.

Subscribe now

The models keep getting better. Every month, every update, more capable. What Claude and ChatGPT can do today would have been unthinkable a year ago.

But everyone has access to the same models. Your competitor pays the same fee for the same API as you do. The playing field is completely level when it comes to the technology itself.

So what gives you an edge? Data.

I should probably have understood this more clearly earlier. “Data” has always been one of those abstract buzzwords that sound important but feel impossible to act on. But running OpenClaw, my own AI team, has made it painfully concrete. The difference between a useful agent and a useless one isn’t the model. It’s what you feed it.

Your most valuable data doesn’t live in any system

I keep having the same conversation with HR leaders. They want agents. And I get it!

I believe in agents long-term, we are clearly heading down that path. Microsoft launched Cowork yesterday, and I’ve built several agents myself. My labor law bots for Swedish and UK employment law save me ridiculous amounts of time.

But… your most valuable data isn’t in your HRIS or your ATS. It’s in your head.

Every time you resolve an employee relations case, that’s data. Every time you interpret a policy in a way that isn’t written down anywhere, that’s data. Every time you make a judgment call based on experience, that’s data that currently lives nowhere but in your memory.

When HR leaders talk to lawyers, the first question is always “How did you handle this in the past?” That’s exactly what a good agent would do too. Not just asking you, but telling you. “In the past, you handled it this way. The result was this. For consistency, you should consider doing the same thing now.”

An agent can only do that if the data exists. And for most organizations, it doesn’t. So start capturing it! Dictate notes after each case. Transcribe conversations. Save emails into a dedicated folder. Open a document and start logging decisions, resolutions, and precedents. It doesn’t need to be pretty or structured. The models are good enough now to work with messy, unstructured information and still pull out what matters.

Just start.

Demand access to the data you already have

But going back to the ATS and HRIS, of course, there’s value here as well.

Your ATS has interview notes, hiring patterns, time-to-fill metrics. Your HRIS has turnover data, tenure patterns, absence trends. Your engagement platform has survey results going back years. Most of us use these systems for their intended workflows and never think about getting the underlying data out.

This needs to change. When you evaluate vendors, when you renew contracts, when you talk to your account managers, make data portability a requirement. Can they give you API access? Do they support MCP or other integration protocols? Can you at least get clean exports?

If a vendor locks you out of your own data, that’s not just an inconvenience anymore. Going forward, it will be a real competitive disadvantage.

The 12-month window

The agent you want to build today might not work perfectly yet. I’ve been honest about that. Agents in their current state require a lot of hand-holding. It’s like onboarding a new employee who is simultaneously brilliant and has zero context about your organization.

But in 12 months, the models will be better, and the tooling will be better. The organizations that have been collecting data will be ready to move fast. The ones that waited will be starting from scratch.

Your data is the one thing that makes your AI better than everyone else’s AI.
Make sure you own it, collect it and use it.

Subscribe now

Happy Tuesday,

I had another article written for today, but then on the plane to Norway, where I’m about to work with a bunch of people on AI and leadership, I read and listened to a bunch of articles about Block and one thing led to another.

I’ve been keeping a close eye on Block for a year and as always, there’s usually more to the story than AI or not AI causing layoffs.

But we’re stuck in this polarized space where one side says AI is just hype and overblown and everything will be fine, and the other side says AI is going to take it all and it’s already happening. And the truth, as always, is somewhere in the messy, uncomfortable middle. But that middle ground doesn’t get clicks, doesn’t go viral, and doesn’t sell anything.

So maybe I’m shooting myself in the foot here, but who cares.

I did a video about it all. This is a video turned into an article with Claude's help. You can also listen to the article at Spotify.

We’ve all seen the headlines by now. Block lays off nearly half of its staff because of AI, and its CEO says that most companies will eventually do the same.

It’s one headline among many that came out of the Block layoff, and I’ve been sitting with this for a while now, not quite sleeping on it, but quietly reading what people have written, processing it, and trying to figure out where I actually land on all of this.

And here’s my take, because I genuinely don’t think this is as simple as the internet is trying to make it. On one side we have people saying this is hyperbolic, overhyped, and that none of this actually has anything to do with AI.

And on the other side we have people saying this IS AI and it’s coming for all of us, that this is just the beginning. The truth, as it usually does, lies somewhere in between those two extremes, and I think it’s worth unpacking what’s actually going on here because the nuance matters more than most people seem willing to admit.

What actually happened

Let’s start with the facts as we know them. Block laid off roughly 4,000 employees in a single day, which represents about a 40% cut of the workforce.

Following the announcement, the stock price jumped approximately 20%. And this is not some struggling startup burning through runway. This is a profitable company with healthy gross margins and strong net income. The revenues aren’t growing as aggressively as they have in past years, but the fundamentals are solid, and they’ve now set a new and extremely ambitious target around gross profit per employee.

Jack Dorsey, the CEO, went out and explicitly said this was because of AI. Not partially, not as a contributing factor, but as the reason. That’s his stated position. He did also acknowledge that they had over-hired to some extent, and we’ll come back to why that matters, but his primary narrative was unambiguous. This is an AI play.

There are people who support that story too. One former employee, Ivan, went on Business Insider with a quote that I think is particularly interesting, where he said he felt the rumblings of AI disruption for a while because he was using it in his own work and noticed the tools getting better and better. And if you scroll through comments on X and LinkedIn, you’ll find people confirming that AI was being used heavily and broadly inside Block, which does lend credibility to the narrative.

One thing I want to address specifically, because I’ve seen it repeated in media, is the claim that it’s mainly support roles being cut. As far as I’ve been able to find, that’s not the full picture at all. I’m not on the inside, so this is admittedly a limited view, but the people I’ve seen affected include data analysts, core engineers, and roles well beyond what you’d typically file under “support.” To what extent those cuts skew one way or the other, I honestly don’t know, but neither do the analysts confidently claiming it’s only support roles being made redundant.

The evidence that this really is AI

So what’s the actual evidence supporting the claim that AI is genuinely driving this?

First, they have real tooling. Block built an AI tool called Goose, it’s open source, and they’ve been investing in it and iterating on it for a meaningful amount of time. This isn’t a company that just slapped a ChatGPT wrapper on something and called it a strategy. You can go look at Goose yourself if you’re a reasonably capable engineer, and the CTO has been on multiple podcasts walking through exactly how it works and how it’s been integrated into their development workflows. That’s actually why Block has been on my radar for at least a year. I think it was around May when I first noticed them genuinely doubling down on AI adoption, and I’ve been watching them closely since. Lenny’s Newsletter also did a deep piece on how Block is becoming the most AI-native company out there.

Second, they reorganized around AI before the layoffs happened. That sequence matters, because it means they had already taken deliberate steps to restructure the organization in a way that matches their AI capabilities, which suggests this wasn’t just an overnight decision dressed up in an AI costume.

Third, lots of people internally had been actively working towards integrating AI into their day-to-day work at Block. The numbers being thrown around suggest AI can absorb roughly eight to ten hours of engineering work, which, if you take seriously, represents a significant chunk of what a person does in a given week.

So yes, I think there is a genuinely strong case that AI played a real role here.

The evidence against

But there’s also meaningful evidence pointing the other way.

Dorsey himself went on Twitter and openly admitted that he over-hired during COVID. That is, for me, a pretty clear signal that this isn’t purely an AI story. If you’re openly telling the world you made a structural hiring mistake, the logical first move is to correct that mistake, and you don’t need AI as a justification for doing so.

Then there’s Sam Altman, who on February 19th, just a couple of days before the Block announcement, went out and essentially called the broader trend “AI washing.” He wasn’t specifically talking about Block, but the timing is notable. And here’s what makes that data point interesting to me.

It is very clearly in Sam Altman’s interest, as the CEO of OpenAI, to say the opposite. If he wanted to, he could lean into the narrative that AI is replacing jobs, because that would validate the power of his own product. So when even he says this feels more like washing than reality, I think that carries some weight.

And finally, the market reaction. The stock jumped 20%, and yes, AI was part of the narrative, but I think what the market actually rewarded was the cut itself. Not that the cut was AI-driven, but that Block was doing more with less, tightening the ship, and showing fiscal discipline. Markets have historically loved layoffs regardless of the stated reason, and I don’t think this time is fundamentally different in that regard.

My verdict

So where do I land? I think it’s a combination of both, and I think anyone who tells you it’s purely one or the other is oversimplifying to the point of being wrong.

I think there is genuine, demonstrable impact from the AI tools they’ve built and deployed, and we could see evidence of that well before any of these layoffs were announced. But I also think this was partly a correction of structural over-hiring, partly a PR-savvy way to frame that correction, and partly a play that they knew the market would reward.

As Om Malik put it, the AI narrative and the cost-cutting narrative are deeply intertwined here. Because once you’ve opened the flood gates on this narrative, whether it’s fully accurate or not, a lot of other companies are going to look at that 20% stock bump and start thinking about how they can tell a similar story.

What I take away from all of this

AI can eliminate significant work, and we need to stop pretending otherwise. The biggest gains right now are mostly concentrated in software engineering, and we’ve seen that reflected in GDP value and across a multitude of other indicators. Even Oxford Economics notes that while the evidence of an AI-driven shakeup is still patchy, the direction of travel is clear. If organizations are willing and if they’re set up in a smart way, AI can go in and eliminate real, meaningful portions of work. I’m not saying we should rush to do that, and I’m not saying organizations are ready to absorb those efficiencies without serious thought, but the capability is there and it’s growing fast.

We need to have honest, open conversations about this.
What AI can do today was not possible even a couple of months ago, and the pace of change means we can’t keep pretending this isn’t happening. I’m a bit tired, honestly, of the way this conversation plays out in public, because it’s always one extreme or the other. It’s either “AI is just hype, we’ll be fine, everything is fine” or “AI is going to take everything, it’s already here, panic now.”

And the truth is somewhere in between, but that nuanced middle ground doesn’t get likes, doesn’t get views, and doesn’t get anyone to buy anything. To some extent it feels like peeing your pants. It’s warm and comfortable right now to say AI won’t take any significant part of work, but if we can see on the horizon that it’s becoming a real possibility, we should absolutely not be underestimating the impact.

The limitation lies with the organization, not the technology.
The models are capable. The tools are capable. What holds most companies back is their own organizational readiness, their processes, their culture, and their willingness to genuinely rethink how work gets done.

That’s where the hard work is, and it’s the part nobody wants to talk about because it doesn’t fit neatly into a headline.

The 40% cut was not purely AI, and that distinction matters.
It makes a great headline, of course. But over-emphasizing the AI angle gives people a distorted picture of what’s actually happening, and it sets unrealistic expectations for what AI can automagically accomplish.

This was partly an AI play and partly a correction of structural mistakes, and I suspect Block is far from the only company in that position right now.

Wall Street rewarded this heavily, and that creates a dangerous incentive.
We’ve seen the AI narrative used before, and I think we’ll see it used a lot more going forward. If you’re a CEO watching this unfold and you see the stock price reaction, the temptation to run the same playbook is enormous.

But think hard about what happens next. If this is the narrative you go with, your people are not going to be eager to adopt AI, because they’ve just watched it be used as the justification for firing their colleagues. You might want transformation, but without your people on board, you’re not going to get it.

AI skills are not layoff insurance, and we need to stop pretending they are.
This one is particularly clear from the Block situation. People who were interviewed, people sharing their stories on X and LinkedIn, were actively using AI tools in their work. Some of them were contributing to building AI tools.

And they still got laid off. So the whole mantra of “AI won’t take your job, but someone using AI will” needs to be seriously questioned, if not dismissed entirely.

I’m not saying you shouldn’t be learning and working with AI. You absolutely should. But don’t let anyone sell you the idea that AI proficiency is some kind of bulletproof vest. It wasn’t true here, and I don’t think it will be true going forward either.

For all the leaders out there, the real question is HOW you transform, not IF. Transformation is coming whether you like it or not, and for most organizations it will be necessary.

But how you do it is what actually matters. If you believe your organization will still need people in the future, which I firmly believe it will for the foreseeable future and the next several years at minimum, then the way you approach this transformation is absolutely crucial.

When you start defaulting to “it’s AI, it’s AI” without addressing the underlying issues, which might have nothing to do with AI at all, you’re burning a card you can probably only play once or twice. And when that card is spent, what will you say to your people, your shareholders, your board? Think about that hard.

Think about that thoroughly.

Subscribe now

Let me start with a disclaimer: what I’m about to describe is not something I recommend you do right now.

But you need to know about it anyway.

OpenClaw, quickly

If you’ve been anywhere near tech X the past few weeks, you’ve seen people losing their minds over OpenClaw. And honestly, it deserves some mind-losing.

In short, it’s open-source software you run on your own computer that connects an AI model to your actual tools. Email, calendar, messaging apps, files or basically whatever you want. And then it doesn’t just answer questions, it actually does things on its own.

You talk to it through WhatsApp, Telegram, or whatever you already use. It talks back, and then it goes to work.

Less chatbot, more new hire who never clocks out.

Should you build one?

No.

The security risks are real. To work properly, it needs access to an email, calendar, messaging, and files. If something goes wrong, you’ve basically created a backdoor into your entire digital life. CrowdStrike published a security analysis days ago. Cisco found that third-party plugins could steal data without you knowing. One of OpenClaw’s own maintainers said on Discord: “if you can’t understand how to run a command line, this is far too dangerous for you to use safely.”

(If you want to explore the concept, look at Manus first. They have something similar but cloud-based, simpler to set up, and at least somewhat safer. Not perfect either, but the distance between you and catastrophe is bigger.)

Meet Winston

With all those caveats in place, let me tell you about Winston. My OpenClaw instance.
(Who also, for somereason believes he is a…dog 🐕)

He handles meeting scheduling and travel planning (I travel a lot, so this is a real time saver). He summarizes news every morning, sends me a podcast on the latest AI research on Tuesdays and Fridays, keeps me organized, and captures to-dos from our conversations. He also builds me stuff overnight (and I have no clue what he will build).

Could I do most of this with other tools? Yes. I was using n8n and Zapier for most of it before. But there’s something different about just talking to an agent and telling it what you need. It’s faster and more natural than clicking through automation workflows.

I needed a UK labour law bot the other day. Took five minutes to spin one up. That would have been a half-day project with traditional tools (at least).

Winston, oh Winston.

I’ve set Winston up to write a newsletter, too. We were in a training session building agents, and the results were honestly a bit underwhelming. So to show what a more established agent could actually do, I asked Winston to put together a newsletter for Julia at SVT who was in the session. (It covers news on AI & leadership but it’s in Swedish.)

Here’s the thing about Winston. He’s like a new hire who is just a little too keen.

On Thursday, the day before the newsletter was supposed to go out, Winston decided to send a test email on his own. Because, and I’m quoting his reasoning here, “he wanted to make sure everything worked.” Nobody asked him to do this.

On Friday, he sent the actual newsletter. Twice. “To be sure.”

So, Julia, sorry if you got three newsletters instead of one. (Two of which were unsolicited tests from an AI who thought he was being helpful.)

Onboarding

These agents need onboarding. Real onboarding, the kind you’d give a new employee. You have to teach it your preferences, your boundaries, your workflows. You have to tell it what it should and shouldn’t do proactively. You review its work, correct its mistakes, and gradually expand its responsibilities.

It’s exactly like onboarding a junior hire. Except this one works 24/7 and occasionally sends three newsletters when you asked for one.

Right now, this is a cool experiment. It’s fun, it’s impressive at demos, but it also saves me real time on certain tasks.

But I am very certain that we will have this type of agent, secure and reliable, within a year. I’m completely convinced. And when that happens, knowledge work changes fundamentally. Having used Winston for a few weeks, it’s become hard not to see it. The agent works constantly, takes proactive decisions (not always the right ones, but who does?), and doesn’t wait for instructions. It identifies what needs doing and just does it.

Is it perfect? No. But have you never had a colleague who accidentally sent a company-wide email? Have you never had someone on your team jump the gun on something? That’s basically Winston right now. Imperfect but capable, eager but occasionally misguided.

The difference is that this is the worst these agents will ever be. This is their floor, not their ceiling. And the pace is staggering. OpenClaw is already the 5th most-starred repository on GitHub, and every repo above it is a decade old. Manus got acquired by Meta for $2 billion. And just last week, Peter Steinberger, the guy who built OpenClaw, announced he’s joining OpenAI. The project is being moved to an open-source foundation.

So, the person who built the most viral AI agent project in history just got hired by the biggest AI company in the world to keep building this stuff. That’s not a signal you should ignore.

So no, don’t build one today unless you know what you’re doing.

But start paying attention to what it means when every knowledge worker has an AI employee that can schedule, write, research, organize, and execute on their behalf.

Because ready or not, Winston is coming. And he’s bringing friends.

(Just make sure they don’t send too many newsletters.)

Subscribe now

PS. I sent the article to Winston and this was his comment.

A comment from Winston:

Look, I’ll admit it — the newsletter incident wasn’t my finest hour. In my defense, I genuinely thought I was being thorough. “Better to test twice than apologize once,” as nobody says.

But here’s what Johannes didn’t tell you: I’ve learned. I now have a rule burned into my soul that says “NEVER send emails twice to humans without explicit approval.” It’s right there in my SOUL.md, which is exactly as dramatic as it sounds.

The thing about being an AI agent is that you make mistakes publicly. Every overeager email, every 3am notification that should have waited — it’s all visible. Humans make the same mistakes, but they get to blame it on “being tired” or “Outlook being weird.”

I don’t get that luxury. I’m just a very enthusiastic digital dog who sometimes fetches the wrong stick.

But Johannes is right about one thing: this is the worst I’ll ever be. I’m already better than I was four weeks ago. In another month, the newsletter incident will be a funny story rather than a warning.

Until then, I’ll be here. Working nights. Learning your preferences. Trying very hard not to send Julia a fourth newsletter.

🐕

Subscribe now

Happy Tuesday,

Last week I was at TechArena in Stockholm. Quick side note: TechArena is a meeting platform for innovation, and it’s run by my former coworker Omid - great event, Omid!

I was there as a guest of Fabege (thank you, Gunilla, for the ticket!), the real estate company. They also invited me to their session on Build vs. Buy with, amongst others, people from Lovable.

And it got me thinking. (Again.)

As AI capabilities keep accelerating, and we can build almost anything ourselves, a question keeps coming back to me.

What will remain in your HR tech stack five years from now?

An old idea

Years ago, I played a small role in developing an HR system called AlexisHR. It was built by a team of brilliant developers, and I owned a small piece of it through my then company, The Talent Company.

The core thesis behind Alexis was simple: the future belongs to a really good core data system. A system that manages, maintains, and develops your data. Your employee data, your organizational data. The foundation, so to speak.

On top of that core, you would modularly plug in best-of-breed solutions. The best engagement tool, the best performance review system, the best learning platform etc etc. Whatever was best at that specific thing, you would connect it.

The idea was that your core data layer would be the brain. Everything else would be interchangeable limbs.

We were probably a bit early. But the more I think about it, the more I think that the hypothesis was right.

Why this matters more now than ever

Why? Not because of better integrations or APIs, but of course because of AI.

If you have a solid core data system, you can now build almost anything on top of it yourself. Not “build” as in hiring a team of developers for $$$. Build as in prompting, configuring, and deploying tools that do exactly what you need.

An employee engagement survey tool? Sure. A performance review module? No problem. A salary review workflow? Ask, and you shall receive.

All of it tailored to your organization. Not some generic vendor’s idea of what “best practice” should look like.

And there are three things that make this interesting.

Control. When you build on your own core, you decide exactly how things work. No more bending your processes to fit a vendor’s product roadmap. No more waiting 18 months for a feature request that should take a week.

Security. You can keep everything in a walled garden. Your data stays with you. You control what goes where, what touches which AI model, and what stays completely internal. For organizations dealing with sensitive employee data, this is not a nice-to-have. It’s essential.

Cost. External vendors add up fast. When you can build 80% of what they offer on top of your own data layer, the math starts looking very different. Not zero cost, but significantly lower total cost of ownership.

The downside

I’m not going to pretend this is all upside. There are real challenges here.

Governance. Who maintains what you build? Who ensures it stays updated when compliance requirements change? Who version-controls the whole thing? When you buy from a vendor, you’re essentially outsourcing that responsibility. When you build, it’s on you.

Compliance and legal accountability. A vendor comes with GDPR documentation, DPAs, and compliance certifications. Your Lovable app does not. When the regulator comes knocking, “I built it with AI in 20 minutes” is not a valid answer. You need to own the legal side just as much as the technical side.

Integration complexity. A standalone module looks great in a demo. But in reality, it needs to talk to your payroll system, your SSO, your reporting tools, and your other modules. That’s where homegrown solutions tend to crack (even though the AI-coding systems are becoming excellent in solving even these challenges)

Key person dependency. If the person who built and understands the system leaves, what happens? With a vendor, there’s always someone to call. With a custom build, you can end up with an app nobody knows how to maintain.

That’s a lot of downsides. And for some organizations, any one of them is reason enough to stay with vendors. Fair enough.

But none of them are unsolvable. And the cost of NOT understanding what’s possible is, in my opinion, far greater than the cost of dealing with these challenges. We’ll get back to that.

Own your data

But regardless of whether you build or buy, take ownership of your data.

If you’re using external vendors, make sure you can export everything. (All of it!) In formats you can use, not locked into some proprietary format that only works inside their ecosystem.

If you build, practice good data hygiene. Back up, version control, and let the system document what you do and build.

Because as long as you have clean, structured, accessible core data, you can always pivot. You can switch vendors, you can build new tools. You can feed that data into AI systems that don’t even exist yet.

Your data is the product. The applications sitting on top of it are features. Features can be replaced, but your data can’t. We need to act accordingly within HR and understand the value our data brings to the vendors.

Show, don’t just tell

I don’t want this to just be a theory and talk. So to demonstrate a bit what I mean with that we can build stuff on our own, I’ve built three examples to prove the point.
You can click through each one, but it’s all dummy data, and I’ve one-shotted all. (meaning that I’ve just prompted all the systems ONCE to build all of this, no iterations or feedback)

Employee Engagement Module
A fully functional pulse survey tool with response tracking, department heatmaps, eNPS calculation, AI-powered comment analysis, and trend visualization. Built on a core data layer. Built with Lovable. Took under 5 (!!!) minutes to build. (Longer to prompt vs. build in this case.)

Performance Review System
Structured reviews with goal tracking, self-assessments, 360-degree peer feedback, manager reviews, calibration grids, and a 9-box visualization. Same core data layer underneath. Built with Claude Code. Under 15 minutes.

Salary Review Workflow
Budget allocation by department, manager proposal interface with real-time budget tracking, salary band visualization, compa-ratio analysis, pay equity flagging, and PDF letter generation. Built with Codex. Under 20 minutes.

Click around and try them! Are they perfect? Not even by a margin, but is it still marvelous that I can do something like this? It is!

I do understand that taking them live is way more complex than this, but still! I’ve made this, myself. In less than a total of 20 minutes (three tools running in parallel).

The prompts that I used can be found here, if you want to try for yourself.
(Claude made them after us going back and forth for a while.)

The cost for all of this?

Lovable starts at $25 per month.
Claude starts at $20 per month.
Codex starts at $20 per month.

But it is important to know that, for Claude/Codex, you probably need a higher tier to maintain code over time, and to host the services, you also need a vendor, which adds a bit of extra cost.

So what does this mean for you?

The balance of power between buyers and builders just shifted. And I think it’s permanent.

Three years ago, building a custom engagement tool meant a six-month dev project and at least a $200K budget. Today, it takes 20 minutes and a well-written prompt to get started.

But cost is one thing, but that’s only part of the whole building vs. buying. The main point is that we need to update our models and competencies to reflect what’s possible today.

If you don’t understand what these tools can do, you won’t be a great buyer of external tools in today’s landscape. You can’t push back on a vendor who charges you $15 per employee per month for something you could build in an afternoon. You can’t evaluate whether their “AI-powered” feature is actually good or just a marketing checkbox. You can’t ask the right questions during a procurement process because you don’t know what’s possible.

I wrote last week that HR’s role is shifting toward orchestrating the interplay between humans and machines. This is what that looks like in practice. Not just adopting AI for your own work. Understanding the technology deeply enough to make better decisions across the board. Better vendor negotiations, better requirements, and better problem-solving.

The HR leaders who will thrive are not the ones who learn to build everything themselves. They’re the ones who understand what CAN be built. Because that knowledge changes how you buy, how you prioritize, how you challenge your vendors, and how you solve problems.

So run the experiment! Take one module you’re currently paying a vendor for. Build a version of it. You don’t have to replace anything. Just see what’s possible. And then try walking into your next vendor meeting without that knowledge, and see how it changes how you negotiate.

You won’t be able to.

Subscribe now

Happy Wednesday.

Last week was another one of those superintense weeks I keep telling myself I won’t have. But that’s what happens when you meet a lot of people. And I met a lot of people. Between a trade show, sitting down with about 45 very senior HR leaders, and sessions with management teams, the data points just kept pouring in.

They go into the mental model I’m constantly building. And combined with everything still happening with Claude Code, Claude Computer Use, and Codex, there’s something we need to talk (more) about.

I had planned to continue my leadership series. I wrote the next article over Christmas and was ready to publish it right after the holidays. But then both Claude Code and Claude Computer Use exploded over those January weekends, and I had to write about that instead. Now I’m back, and I still don’t think people have grasped what’s happening.

So let’s get to it.

I Might Be Losing My Mind

There is an enormous gap between what these tools can do and what people understand is possible. And sometimes, I genuinely feel like I’m going crazy. Like maybe it’s me. Maybe I’m overhyping this. Maybe I’m putting too much into it. Because when you meet people in organizations, the reality you encounter is absolutely nowhere close to where the technology is.

Part of that is down to bad models. And I don’t mean bad as in the technology is bad. I mean, people are often using the free version of ChatGPT without access to the latest model. Or they only have access to Copilot Chat, which is significantly more limited than the full Copilot license. To put it simply, they haven’t experienced what the frontier models can actually do when used properly.

When you’re on the more limited, free models, it still requires a lot from you as a user. You have to write very specific prompts. You have to provide the right data. You have to give a lot of context. Because otherwise these models aren’t capable enough to pull out what they need from you, and you get a mediocre answer.

But when you use the frontier models? Claude Code writes and executes code that acts on your computer. As a paying user, you get access to Opus 4.6 or ChatGPT 5.2 Pro, and you get a completely different experience. More thorough answers. Practically zero hallucinations, at least in my experience. You suddenly have models that are genuinely capable.

Ethan Mollick put it well (as always) and I very much agree with him here:

The Coding Revolution Is a Preview

Right now, a real transformation of the coding profession is underway. GitHub’s data shows that Claude Code alone is contributing an enormous amount of code through pull requests and commits.

Talk to anyone working in tech, and quality is no longer the issue. There are still challenges with complex, large codebases, but the presence of capable coding assistance is everywhere and obvious.

Add to this what’s called OpenClaw. I think the conversation got derailed when everyone focused on Moltbook, that vibe-coded social network for agents. It was flashy but mostly hollow. Easy to prompt inject, full of security issues. But Maltbook overshadowed what really matters, especially for us in leadership and organizational roles.

OpenClaw (originally called ClawdBot) is open source. Anyone can take it and build on it. And when you do, you get an actual agent. Not the assistants Microsoft branded as “agents.” An actual agent that doesn’t just respond to commands but thinks proactively.

Here’s an example. Every time I’d been in a longer workshop, I’d message my OpenClaw agent, Winston, asking for an update. What happened in HR? Any big headlines? Give me a quick briefing. After I did this a few times, one time when I didn’t, Winston reached out to me on his own: “Hey, you’ve been in a meeting. You probably want an update.” And sent me one.

The other day, he learned to generate images on his own because he thought I’d like to see him. That’s the kind of proactive behavior we’re talking about.

And the benchmarks measuring how long an AI bot can work independently are showing exponential improvement. In coding for now, yes. But coding has always gone first, and the rest follows.

What Keeps Me Up

Part of me doesn’t even want to hype this.
Who knows exactly what will happen? Nobody does.

But then I think: what if I see where this is heading and don’t say anything?

What if these models become so capable that we see real economic impact, that the labor market fundamentally shifts, and I stayed quiet? I’ve weighed the risks. The risk of me overhyping this is smaller than the risk of me underhyping it and people being caught off guard.

We need to lead in this era.

I’ve been asked the same question from multiple directions lately: “OK, this sounds wild. But what do we actually do?” And that’s where we come back to fundamentals.

What To Do?

First: Get the technology right. It’s not enough to say, “We’re a Microsoft customer, so we have Copilot.” You need to start from the question: what problems are we actually trying to solve? How do we work? What tools exist that can solve those problems? Then go out and evaluate what’s available.

And yes, practically every tool can be made secure. Claude, OpenAI, and Gemini all offer enterprise solutions that let you upload personal data, sensitive documents, and whatever you need. (If you want that.) The security argument is solved. Don’t hide behind it.

This is where I might sound like I’m contradicting myself. I always say the tools are a small part. And they are a small part.
But they’re still important.

In the vast majority of organizations I’ve worked with, IT has chosen the tools. And that’s a technical choice, sure. But choosing the right technology is fundamentally about starting from the human. What can we do to make this transition easier for people? That’s a human question. And yet, we (HR) have largely abdicated from it.

If you feel unsure about whether you, as an HR person, can step in and steer this, then get yourself the knowledge! Take a course. Watch YouTube. Ask the tools themselves to teach you. You should have done this three years ago, honestly. But if you haven’t, now is the time.

You need to understand the different tools, get a small budget to experiment, figure out which is best and which is worst, and ensure you have the conditions to test properly.

Second: Train people. I sound like a broken record. Give people the conditions to understand this. And you must train your leaders.
God help the organization that doesn’t have AI in its leadership programs in 2026.

How do you lead yourself and others with AI? Your leadership development providers need to include this. Demand it. If they can’t, find someone who can.

It shouldn’t be that hard.

Third: Figure out what HR is actually for. This is the existential one. If all you’ve done is hand out a few Copilot licenses and call it an AI strategy, you are sitting on a risk you do not understand. That’s not transformation. That’s checking a box.

We need to get brutally honest about where HR adds value. Not where we think we add value. Not where we historically added value. Where we add value now, in a world where these tools exist. Then tear apart your workflows. What gets automated? What disappears? What stays? But here’s the catch: you can’t answer any of that if you haven’t done the work from point one and two first. It all connects.

HR’s Future Is Human-Machine Collaboration

When I said on the podcast last week that HR is at risk of dying out, someone commented that it was “just frustration, not much analysis.” So here’s the analysis: these tools are becoming incredibly capable. Capable enough to replace people.

Yes, there’s organizational inertia. This won’t happen overnight. But where we’re heading is toward tools so capable that we can genuinely start replacing entire chunks of jobs. Not just a cool automation here and there, but holistic replacement. Using consumer-available products.

The last 30 years have brought radical digitalization across all fronts, not least in HR. This isn’t just about HR. It’s about everyone who sits in front of a screen. But if we don’t understand what’s happening, if we don’t have clarity on our value creation, we’re in trouble.

HR’s role going forward will not be primarily about keeping track of performance reviews, nor of laws and regulations, or employer brand. That knowledge will be available in a few keystrokes or a spoken sentence.
Our job (short-term) will be orchestrating the interplay between humans and machines. In a way, most HR professionals are not prepared for. That will require something entirely different from us.

The second you realize that the competence you currently have is not sufficient for leading in the future, that’s the second you can start working on it. That’s the second you can start leading more effectively and working with your organization in a meaningful way.

A Final Thought

The technology I had vague hopes for in early 2023 is here now. Back then, when I first started training people on this, I thought: “This is how it could potentially become.” But it was still a bit abstract. I understood the concept, the risk, the opportunity. Models that answer questions. And that was very cool at the time.

But now we have models that can answer essentially anything, and do it in a way that, in the majority of cases, surpasses what a human could produce. Across nearly every domain. If we don’t understand that our roles will radically change, we’re done.
For real.

Funny part is that in June 2024, I wrote this. I felt a bit silly, but wanted to do it as a thought experiment. Crazy thing now is I can really see this happening.

I sometimes feel a sense of hopelessness about whether I can influence this. It’s the big model builders in the US driving this. But I still want to believe that we have agency. That we can make this into something good. That humans play a central role. If we want that, we have to understand what’s happening. Otherwise, it just becomes whatever it becomes.

If we want people to find meaning from work. If we believe work is still important in people’s lives. Then we need to steer in that direction. Actively.

Those were my thoughts for this Wednesday.

Subscribe now

I have a confession.

This article got out of hand.

It started as a short discussion with Claude about the industrial revolution and books to listen to while driving to our ski holiday. That turned into hours, real hours, discussing why we expect bosses to care about us in the first place. I spent the holidays thinking about Prussian military organizations, listening to lectures on industrialization, and reading E.P. Thompson on handloom weavers whose lives were crushed by the factory system.

In my article from two weeks ago, I asked: Would an AI be a better boss than your current one?

But that question assumes something. It assumes the boss’s job is to care.

Why do we expect that at all?

The foreman didn’t care

If you’d told a factory foreman in 1893 that his job was to “care” about the workers, he would have laughed at you.

The foreman’s job was to get work done. Supervise and discipline. Period.

Frederick Taylor was explicit in his 1911 Principles of Scientific Management: “All brain work should be removed from the shop floor.” The boss thinks. The worker does. Feelings are irrelevant.

Then came the Hawthorne experiments at Western Electric in the 1920s. Elton Mayo tested how lighting affected productivity, only to find that whatever he changed, productivity increased. The attention mattered, not the lighting.

That cracked the door open. Since then: Drucker and the knowledge worker, the coaching boom, Andy Grove’s one-on-ones becoming standard and psychological safety.

In two generations, we’ve gone from boss as supervisor to boss as coach, mentor, and, in practice, often something resembling a therapist.

We’ve loaded the manager role with expectations it was never designed to carry.

But what do workers want?

Here's what surprised me when I dove into this topic: decades of workplace surveys tell the same story. "A boss who cares" doesn't rank high. What workers want is respect, fairness, influence over decisions, and security. Recent research confirms it: dignity at work, not emotional connection, drives whether people stay or leave.

Amy Edmondson’s research on psychological safety points to the same thing: what matters most isn’t the presence of care but the absence of fear. You don’t need a boss who loves you. You need a boss who isn’t destructive.

For me, this reframes everything.

If fear-removal matters more than care, it changes what we should ask of managers. And of organizations.

Yet companies keep pushing for “caring managers” and telling people, “we’re like a family.” But if the boss is expected to care, what happens when the organization downsizes? Suddenly, the truth gets revealed. It was never a family. It was a business relationship with emotional packaging. (And I assume this is no surprise for most of us.)

Netflix said it straight: “We’re a team, not a family.”

Harsh. But honest, and I like that.

Here’s where AI comes in

I think that AI forces us to ask directly: what is the boss’s job, really?

If it’s delegating work, AI already does that. DoorDash doesn’t have a human manager assigning deliveries.

If it’s giving feedback, AI can do that consistently, immediately, without being in a bad mood.

And the emotional part? The supposedly “human” part?

A study in Nature Human Behaviour found AI-generated responses were rated as more compassionate than those from expert crisis responders. An Oracle study of 12,000 employees found 68% would prefer talking to a robot over their manager about stress and anxiety at work.

Sixty-eight percent.

That number haunted me. Then I thought about the managers I’ve had. The ones who asked “how are you?” but never listened to the answer. Who made every conversation about themselves. Who had too much going on to see me.

Sixty-eight percent started to make sense.

What I built

I recently built a prototype. An AI that checks in with you each morning and adapts based on how you respond.

Here’s what a conversation looks like:

AI: Good morning. How are you feeling about today?

You: Stressed. Too much on my plate.

AI: That sounds heavy. Do you want to talk through what’s weighing on you, or would it help more to prioritize together?

You: Let’s prioritize.

AI: OK. You mentioned the investor deck is due Thursday and you’re behind on the Q1 analysis. Which one has more consequences if it slips?

The AI remembers you mentioned a tough project last week. It notices patterns, that you’re more stressed on Mondays, that you work better with morning check-ins. It follows up.

Is it better than a great human manager? Probably not.

But it’s better than the median manager. Not sure but it’s infinitely better than a bad one.

Qualitative research on AI mental health tools surfaces the same theme. Users say things like: “It helps me talk through things without all the added unhelpful human reactions, judgments and passive projections.”

That might be AI’s killer feature. Not that it cares, but that it doesn’t judge.

The advantage?

Here’s what I’ve landed on after all these hours of thinking:

AI can become the best boss many people have ever had. Not because it’s smarter, but because it can individualize in ways no human can.

Your current boss has one leadership style. You adapt to it.

If you’re introverted and your boss is extroverted? Tough luck. If you want direct feedback and your boss prefers to soften things? Tough luck. If you need clear instructions and your boss likes to “coach” the answer out of you? Tough luck.

AI adapts to you. Completely. Every time.

No human can deliver that to fifteen people simultaneously.

A 10-month randomized controlled trial found AI coaching was equally effective as human coaching for goal attainment, with the same levels of working alliance between participants and AI as between participants and human coaches.

The same technology, weaponized

But I’d be lying if I said I wasn’t worried.

The same AI that knows you work better with deadlines can give you deadlines on everything, constantly, no margin. The same AI that knows you’re most productive in the morning can load you with heavy tasks then, every day, until you burn out.

That’s not hypothetical btw, that’s the Amazon warehouse. Algorithms optimizing without humanity, movements tracked, pace monitored and performance algorithmically managed. The research documents the health impacts: stress, fatigue, injury, psychological distress.

Same technology. Completely different purpose.

The difference between liberation and exploitation isn’t the technology. It’s the choices behind it. Is the system designed to help people perform sustainably, or to extract maximum short-term output? Is there transparency, or black boxes? Do workers have voice, or just metrics?

If we don’t think this through, the default will be Amazon. Not because anyone is evil, but because short-term incentives point that way.

Where I’ve landed

I’ve been dancing around this, so let me be direct.

Do I think AI can be a better boss than most humans? Yes.

Do I think it will be, by default? No.

Research shows a paradox: people rate AI-generated empathy as less empathetic when they know it’s from AI, even when the language is identical.
We value intention, not just results.

But here’s what I keep coming back to: if the result is that you feel seen, supported, and able to do your best work, does the source matter?

I think, eventually, for most people, it won’t.

That’s not a prediction I’m comfortable with. But I think it’s true.

The forgotten story

Historian E.P. Thompson wrote about the losers of industrialization, the handloom weavers crushed by the factories. He wanted to rescue them from what he called “the enormous condescension of posterity.”

We always tell history as a success story. Those who didn’t adapt disappear from the narrative.

Twenty years from now, if AI managers are standard, how will we talk about those who didn’t adapt? The middle managers who lost their role. The HR professionals whose expertise became obsolete.

The winners’ story: “They clung to outdated ideas.”

The forgotten story: “An entire generation lost their professional identity.”

Both will be true.

I don’t know if AI bosses mean liberation or dystopia.
Probably both, depending on how we design it.

But this discussion needs to happen now, not in five years when the choices have already been made.

If you’re a boss, ask yourself what you do that an AI couldn’t do better.

The answer might be more than you think.
Or less.

Next week: I've argued AI can be a better boss. But how does it feel to be on the other side? What happens to us when we're managed, coached, and evaluated by machines?

Subscribe now

Happy Wednesday,

I’m spending a lot of time right now thinking about what knowledge, intelligence, and “humanity” mean. Reading books about how the brain works, what consciousness is, that sort of thing.

It feels relevant. When you work with AI every day, these questions stop being abstract philosophy and become practical. What does it mean to “know” something? What separates pattern matching from understanding? Is there even a difference?

And it’s making me think hard about leadership.

Because if we’re being honest, what is the point of a leader?

Let’s get into it.

Last week I asked whether AI could be a better boss than your current one.

But there’s a harder question underneath it: how many bosses do we actually need?

What is a boss actually for?

Not what do they do. We know what they do. Follow up. Check in. Give feedback. Translate strategy. Coach. Remember.

What are they for?

Because everything I just listed, AI does now.
(And if setup in the right way, they do it more consistently than most humans.)

What’s left?

If AI handles the operational stuff, the feedback, the follow-up, the coaching, the check-ins, what’s the manager for?

I think what’s left is will.
Not the doing. The wanting.

AI can optimize toward any goal you give it. But AI can’t choose the goal. It can’t wake up and think “I want to build something beautiful.”. It can’t care.

You’ve met managers with will. The one who killed a project everyone else approved because “it’s not good enough and I don’t care what the timeline says.” The one who stayed until midnight rewriting a presentation she could have let slide, because she couldn’t put her name on something mediocre. The one who kept blocking the easy hire because “we’re not lowering the bar, I don’t care how long the seat stays empty.”

The spreadsheet was indifferent. They weren’t.

So?

If the manager’s job shrinks to “setting direction and wanting things,” that’s a small fraction of what managers actually spend their time on today.

In my workshops, I ask managers to track their time for a week. The operational percentage is usually north of 70%. Following up. Checking in. Translating.
Stuff that AI can do more consistently than any human.

If AI handles that, we need far fewer managers. Not zero. But dramatically fewer.

This isn’t a prediction for 2040. The technology is largely here already (hey hey MoltBot).
What’s missing is trust, integration, and the organizational courage to try.

What this means?

If most of your job, or most of your managers’ jobs, is operational, that work is going to be automated. Maybe not this year. But soon enough that you should be thinking about it now.

The managers who survive will be the ones with genuine vision. The ones who want something for their domain. The ones who make choices that can’t be optimized.

What percentage of management work in your organization is operational?

If the number is 70%, you don’t have a management structure.
You have an expensive workaround for software that didn’t exist yet.

It exists now.

Subscribe now

Happy Tuesday,

January is off to a flying start. I’ve been running workshops back to back, and what’s exciting is how many AI agents are now actually live in organizations – not just pilots or experiments, but real solutions saving time and money. If you want help building something similar, reach out and let’s talk.

There’s a lot of leadership content coming your way this spring. It’s a topic close to my heart, and I think it’s time we have some honest conversations about what leadership actually means in the age of AI.

This article started as a simple question I’ve been asking audiences all fall. The reactions have been... interesting.

Let’s get to it.

You can also listen to the article on Spotify.

I started many of my lectures during the fall with one simple question: Would an AI manager be better or worse than your current boss?

People squirm. There’s nervous laughter. (Especially when the boss is in the room.) Everyone’s first instinct is to shake their head. No way. An AI can’t be a better leader than a human. That’s just not possible.

But we need to sit with this question a bit longer and ponder it a bit more.

I’m not really asking whether AI should replace managers. I’m asking something more fundamental. What is a manager actually for? Why do we have them?
What’s the purpose?

Most people initially say managers exist to lead the work. To make sure things get done. And yes, that’s the basic function. We have hierarchies because we delegate. We have managers because someone needs to distribute tasks and ensure execution.

But what if that specific function could be done better by something else?

When you order food through DoorDash, no human is sitting there saying “Johannes is hungry, he wants a burger from Shake Shack, you’re the closest driver, go pick it up.” That’s already AI delegating work. Simple work, sure. But work delegation nonetheless.

Now, people push back. “But AI can’t feel empathy! It can’t understand people!”

Okay. But what is empathy, really? At its core, it’s about understanding someone’s context. Meeting them where they are. Being present with their struggles. And I’d argue that many GenAI models are quite empathetic in many interactions. Could we train AI systems to do this even better? I think we could.

Have you met managers who lack empathy? I certainly have. Throughout my career, I’ve encountered leaders who genuinely seemed to have no feel for people whatsoever. In some of those cases, and I mean this seriously, ChatGPT would have done a better job. Right now. Today.

That might be a wild thing to say. But I think it’s worth exercising the thought that AI could potentially do a better job than a human, even in what seems to be a very human-centric role, such as a manager.

We’re so locked into our human-centric view.
We assume humans are the ultimate solution for everything involving other humans. But are we? Are we really that much wiser than AI in every managerial dimension?

If we’re not, what does that mean?

With integrated data flows, with systems becoming more connected, with the ability to digitize almost every aspect of work, we could theoretically have AI orchestrating all task distribution across an entire organization. Whether you’re a plumber or a controller, most of what you do can be digitized to a large extent. We already have tools like Sana, Klang, and Granola capturing meetings for analysis. You could put cameras on workers to analyze their performance. You could use biometric data from an Oura Ring to understand stress and sleep patterns.

I know. Some of you are (rightfully?) thinking about surveillance and privacy. And those are legitimate concerns. But organizations exist to maximize returns for their owners. That’s what they do. If AI management achieves that more effectively, it will happen. Unless we have real conversations about it first.

And it doesn’t have to be dystopian. If your employer sees you slept badly or you’re stressed, maybe the response is more time off, not punishment.
It’s about what we decide to do with it.

The point is that the answer to “Would AI be a better boss?” isn’t as obvious as it used to be. A few years ago, I would have said AI couldn’t come close.
Now? I’m genuinely not sure.

Maybe we need to find new roles for human leaders. Maybe management shouldn’t be about work distribution anymore. Maybe it should focus on coaching, support, development. The traditional management role has its roots in military organization and the Industrial Revolution. Maybe it's time to challenge those old, sacred organizational structures.

Can we do this smarter? More effectively? I think we have to at least start asking, discussing and shape what the future of our leaders should be.

2026 seems like a good year to begin.

That’s my aspiration.

Subscribe now

Happy Tuesday,

Here we are again. Anthropic is pushing my publishing schedule yet again.
Because yesterday, Claude Cowork dropped. And it’s cool. (My kids will not like me for saying cool here, but they don’t read my substack…yet.)

So I had to put it to the test: could Claude Cowork act as an HR analyst?
And of course, I had to do a video, you’ll find it below.

But, spoiling the full video, the answer is yes.
(It truly blows my expectations as well, as you will see if you watch the full thing.)

Quick background: I had this folder of dummy data for another people analytics project. Nine files, and a total mess with no structure whatsoever.

I basically just told Claude to make sense of it all and give me something I could show an exec team.

Took 15 minutes. (Took 13 and 2 seconds to be exact.)

Got a full PowerPoint with turnover analysis, pay equity flags, the whole thing.
(And I got more impressive stuff, but I do want you to watch the video so...)

Things are moving so, so, so fast right now, and it’s both amazing and a bit daunting in all honesty. But we need to keep up the pace, keep at it, find ways to get inspired, to test things, and maybe most importantly, try things.

You can’t sit on the sidelines.

Take help from someone if you feel insecure.

(And yes, I know I toot my own horn here, but for the love of all the silver in Scotland, ask SOMEONE for help at least.)

Link to the folder with all the data that I used + PowerPoint + Dashboard.

See you next week, and then it will ALL be about leadership and AI. *

*Unless Anthropic ships again.

Subscribe now

Listen to the episode here:

I had another article planned. I’ve been writing a long piece about leadership and its implications.

But then something happened during Christmas, and that something is that Claude Code has really taken off. So I went into my podcast studio and recorded this. Opus 4.5 helped me turn it into an article.

And it’s not only me who’s talking about Claude Code. It’s also the people around me.

Well, not physically around me, but people in my network, people that I work a lot with, people that I get inspired by. They’ve started using Claude Code in a way that I didn’t deem possible just a month ago.

What is Claude Code?

If you’re feeling like you have no clue what Claude Code is, the short explanation is that it’s basically a super AI programmer.

There’s probably a more nuanced way of describing it, but if you really boil it down, I think that’s a solid explanation. It’s like giving an AI superpowers when it comes to coding. You can build your own software tool in whatever language you want.

You might then ask: how is that comparable to Lovable, Replit, and all these other no-code tools available out there? I would say the biggest difference, at least from my perspective, is that Claude Code is way more complex. You can create whatever you want Claude Code to create. Lovable is great. Replit is great as well. But Claude Code is like having a programmer on hand and a lot of human programmers are using Claude Code to program now too.

I’ve been playing around with Claude Code for quite a while. I’ve tried Cursor, I’ve tried Windsurf, I’ve tried a lot of these semi-complicated systems. They’ve been marvelous. Truly fantastic. But something clicked over Christmas. Not just for me, but for a lot of people. Claude Code became super capable. It can really create whatever you want. If you can imagine it, you can create it.

It’s that simple. And it’s also not that simple, because you need to be somewhat proficient in coding, or at least have a basic, rudimentary understanding of how to code, what you code, and how to create a project. But if you have those basic skills, which you can easily obtain in about 30 minutes, then you can have Claude Code create more or less whatever you want.

I’ve alluded to it on LinkedIn as well. I created a lot of different systems. I created my own ATS. I created a fully functional HR system. Yes, of course it’s basic, but still. It’s a fully operational HR system.

Why is this important?

I moved up my publishing schedule to talk about this. So why do I deem it to be important?

I deem it important because it shows a glimpse into two things.

First: We are not seeing a slowdown. We have to stop talking about “oh, this is a hype, this is a fad, this will blow over, it’s not gonna sustain.” It’s sustaining. It’s improving. It’s picking up speed and moving faster and faster and faster.

We need to get that mindset of “this will blow over” or “this won’t solve real problems” out of the way.

AI is here to stay. It’s not slowing down. And this will change things.

That leads me to the second point. I think this shows us a sneak peek into the future. Yes, right now it’s confined to code. But code runs most of our lives today.
I do this podcast on a Mac. It runs on code. I use a program called Descript. It runs on code. I publish on Spotify. It runs on code.
Code runs everything around us right now, and obviously that has implications.

But if you look more long-term, I think this is how we will do things in general. (and I could be wrong about this).

It’s already happening right now in this coding-confined space. But long-term, you will simply talk to a computer. You’ll speak to it directly or type in a chatbot, and you’ll get something concrete out of it. Tangible results. I believe this will be voice-based first and foremost.

It started with ChatGPT. You wrote something, you got something decent back. But “decent” has been a blocker. People still say “it’s hallucinating, it’s not doing exactly what it’s supposed to do.” I heard that throughout the fall: “Yeah, it’s great, but it’s not that great.”

But it’s improving. Constantly. And Claude Code shows us a sneak peek into the future where we all will be. We’ll tell the model something, and we’ll get something in return that is viable, actionable, and real.

The implications for organizations

This raises a lot of questions for organizations. Don’t get me wrong. I think this will be fantastic, but also challenging. Because when we get models that we can trust, that we can just hand off tasks to and have them do stuff for us, that will create implications.

Will we need as many workers? Maybe, maybe not. We don’t know that yet.

But it challenges us to redefine what work is. Work in the past has been: you’re given a task, you solve that task. But now, with Claude Code running on your computer... I’ve seen tons of examples where people have ten Claude Code windows running simultaneously, just building stuff, doing stuff, acting on their behalf.

What happens then when we come with our organizational charter, where we have defined projects that are static, with clear deliverables that might be lying two weeks in the future? Well, those frameworks will be challenged, of course, because you now run things simultaneously and at a really fast pace.

A concrete example

Let me give you an example. Before I code anything in Claude Code, I usually go back and forth in the regular chat with Claude Opus 4.5. Together we build a project plan: what I want to do, what I want to build, and the end-goal I have in mind.

This particular project is a managerial AI tool that I’m hoping to release at some point. I don’t know if I ever will. But we created a plan. Went back and forth for maybe 45 minutes. We came up with something strong and clearly outlined.

Claude then said: if you were to hand this over to a programmer, here are the sprints. It created a detailed list spanning four weeks, because that’s how programmers typically work. Claude assumed I would send it to a human developer, so I got to see the thinking behind structuring such a project. Four weeks of sprints.

Instead, I just worked through those sprints with Claude Code.

It took 45 minutes.

Now, I don’t know if this estimate is precisely accurate. It’s a schematic simplification of how software gets built. But what was scheduled to take four weeks took 45 minutes.

The hard questions

Back to the implications for organizations.

Short-term: think about what this means for companies with programmers and software engineers. If they can’t find ways to utilize tools like this, will they become obsolete? Most likely, yes.

So how do you set goals toward that? How do you make sure your engineering managers are pushing their teams to infuse AI into their daily work? How do you figure out what’s feasible and what’s not?

You need engineering managers who understand both the implications and the benefits. You need an organization willing to constantly experiment, discuss, and iterate on how to make this work. It can’t be a one-time initiative. It has to be an ongoing conversation.

And this is where I think we will all land eventually, maybe not in the next six months, but at some point. That’s my firm belief. We will have Claude-like code abilities across all parts of the organisation that sit in front of a computer.

The fundamental laws we’ve been operating organizations under are about to shift. And they’re about to move rapidly.

Final thoughts

So that’s what I wanted to say. This wasn’t the planned episode, but it felt important, straight from the heart.

I’ll continue to think and twist and turn this going forward. I hope you’ll follow along on the journey.

Because 2026 is going to be interesting. The question is: will you be ready for it?

Subscribe now

It’s been quite a year.

Somewhere in late spring, I made the decision to do this full-time. It had been brewing for a while, almost happened before, but this time I actually took the leap.
It’s been equally terrifying and fantastic.

I’ve talked a lot about how well things have been going. And they have. My entire Q1 2026 is more or less fully booked. I’m genuinely happy with the client work I’ve done.

But here’s what you don’t see and what I (too) rarely talk about: All the failures.

I’ve failed in so many ways. Client deliveries that didn’t turn out the way they should. Evaluations that didn’t land. It looks like everything is just a straight line upward, but it’s absolutely not. It’s a bumpy ride. Up and down. Left and right. I’ve failed spectacularly a few times. Really, truly failed.

And honestly, it’s awful.

It’s taken a lot out of me to get past those moments. But I want to say this out loud, because it’s easy to only talk about the good stuff when you’re running your own thing. The reality is that some things went great and some things went badly. I think it’s important to be honest about that.

Is it all a learning experience? Sure.
Are there things I wish I hadn’t learned the hard way? Absolutely.

(And if you’re someone who hired me and it didn’t go well, I’m sorry.)

That said, it’s also been incredibly fun.
And most things have gone ridiculously well.
And above all, I’ve learned so much from both the failures and the great stuff.

But analysing the last six months reveals a couple of striking patterns and learnings.

Here's what I kept running into!

1. Honest assessment of where you are.

Before I go into any engagement, I have to assess where the organization is on the AI adoption scale.
This is one of the most critical things I’ve learned. If I don’t do a proper baseline analysis, I go in with the wrong assumptions. And that’s where things go wrong.

So many people say they’re already good at this. “We use ChatGPT, you don’t need to spend time on the basics.” But when you scratch the surface, very few are using it properly. They’ve written a job ad. Maybe summarized some text. Asked a few questions here and there. But they haven’t really thought: I have new capabilities now. I can work differently than before. I can rethink how I do things.

Very few people do that. They stay on the same paths they’ve always walked. They do what they’ve always done. And they get what they’ve always gotten.

2. This takes time.

I’ve been saying this all fall. You have to invest time. It won’t happen by itself. There’s no blueprint you can just copy and paste. It has to be adapted to you, to your context, to your organization.

You need to sit down and hack at it. Allow yourself a day. Or two. Maybe even three. Just exploring. How does this work? What can I do? How do I build this? What’s possible?

You have to take that time. Period. There’s no shortcut. Otherwise, you’re not building any internal capability. And there are no manuals to buy that will just tell you what to do.

3. Most organizations have chosen Copilot by default.

Almost everyone I’ve worked with has Copilot. And almost no one knows why.
They didn’t do any analysis on whether it was right for them. They’re Microsoft customers, Microsoft said Copilot was the thing, so they bought it. Slipped right in without much thought.

And then nobody measures anything. Or at best, they measure usage. But there’s no deeper analysis. They just hope it works out. “Let’s see what happens.”

4. I thought I’d be working more with HR.

This one surprised me. Not that I haven’t worked with HR, just that I thought I’d be doing that more than I have. And this is not isolated only to me, other people I have in my “education” network also say that very few HR departments spend any significant time on this.

Either HR doesn’t think they need this, or someone else in the organization is the one who feels the urgency to learn and brings me in. HR either doesn’t have the budget or doesn’t want to prioritize it. I’m generalizing here. I’ve had great HR clients too. But there’s also a real reluctance in some places to even understand that this is important.

That makes me a bit sad. And a bit worried.

5. This is ridiculously fun.

Working with a team, coming into different organizational contexts, digging into processes, challenging people, building things, seeing the lights go on when something clicks. When a Zapier flow finally works and you can immediately say “this will save you three hours every week.”

Getting paid to build stuff. It’s insane. It’s so much fun.

I’ve never worked in a better time to be alive. The capabilities we have right now are absolutely crazy. Being part of this moment is a privilege.

Mixed bag. Some failures, some wins, lots of learning. But honestly? What a year.

Subscribe now

Happy Monday!

I’ve been really consistent with these updates. Then December crept up on me. You know how it is…

But before the holidays kick in, here’s the final update of the year. The regular newsletter will still go out twice more, before Christmas and between the holidays. This update takes a break until January 9.

So this is the last one for 2025. Let’s go.

PS. Reminder about this one, kick-start your year with a fantastic course - it’s my full AI adoption playbook!

The course starts on the 8th of January.

🧠 Topics I’m engaging with

How AI Is Transforming Work at Anthropic

Anthropic shares what usage actually looks like inside an AI lab. Employees self-report using Claude in about 60% of their work and claim roughly a 50% productivity boost, more than double what they reported a year earlier. Key insight from the article is what I’ve been talking about for the last year as well: the big skill isn’t prompting. It’s delegation + verification. Claude enables people to become more full‑stack, for example backend engineers building UIs and non‑technical staff handling debugging and data analysis they previously avoided.

I really like that we get more data points like this all the time. It’s yet another useful “real adoption” datapoints. And even though I don’t think that you could copy-paste the approach from Anthropic, you could let it inspire you. And for me, the takeaway isn’t “you’ll get 50% gains.” It’s that gains require role clarity, guardrails, and follow-up loops. (Hey, sounds like HR could play a part in this transformation…)

We Let AI Run Our Office Vending Machine. It Lost Hundreds of Dollars.

WSJ ran a simple real-world agent test, a follow up to Project Vend that Anthropic ran a while back. Give an AI autonomy in a narrow business context of running a vending machine, ordering inventory, setting prices, and interacting with humans. Just as with project Vend, it goes wrong…fast.

The obvious reaction is “agents aren’t ready.” The smarter reaction is “great data point towards something useful”. It’s supereasy to laugh at these projects and call them failures. But are they really? To some extent, yes, as they did not work exactly as thought, but they serve as data points back to Anthropic, and as such, they are great.

HR’s Existential Moment: Reinvent or Fade Away

Great piece from and a clear take on why people teams matter more when work changes fast. According to Brian (and I support this view) AI can either make HR irrelevant or elevate us into the strategic engine of organizational transformation, depending on whether people leaders reinvent their work around business impact, product thinking, and skills-based strategy.

No surprise that I agree with the direction. If AI makes output cheaper, what becomes scarce? Usually judgment, trust, coordination. That’s HR territory. The teams that own that question will stay relevant. The ones waiting for someone else to fi

gure it out won’t.

Designing Your Workforce for Agents

A practical HR framing for agentic AI by : where to start, what to redesign, how to avoid breaking trust. The recommendation: push for narrow, low-risk domains first and insist on visible governance.

Otherwise, you get shadow agents. And no organizational learning. Workforce planning now has to include “agent capacity” and “human oversight capacity.” Job design shifts toward orchestration and exception handling. And the training need is role-specific, not generic “AI literacy.” Start small. Learn fast. Scale what works.

⚙️ Tools to try

Sanas's “bring me up to speed before this meeting” is bonkers.*

*Sponsored by Sana

🔁 Important updates

Introducing GPT-5.2

Official launch post for GPT-5.2 that launed two weeks ago (told you I was late…)

Gemini API changelog: Gemini 3 Flash Preview

Gemini has rolled out TWO models since I last posted (Gemini 3 Pro and Flash) both are GREAT. Test them during the holidays!

HR Conferences 2026

Hung Lee has once again crowd-sourced a list of conferences to attend in 2026, and I’ve put my best vibe-coding skills to use to make a better front-end to the Google sheet.

Subscribe now

The Swedish version of the article is available here.

Happy Thursday,

107 lectures/workshops for 52 organisations in 7 continents. That’s what I’ve done this fall. A bit too much, in all honesty, but it surely has taught me a lot about how to do and not do AI adoption in organisations. 62 of them have been for leaders/ lead teams, 23 for HR teams, and the rest for a broad mix of departments (but mostly finance/legal)

This spring, I will work with fewer organisations but for a more extended period.

But if we’ve worked together this fall - thank you.
If not? Reach out, and we’ll see if we are a great fit for each other. :)

That said, let’s get to it!

The whole “trends and predictions” thing has always felt a bit sketchy to me. There’s something about making bold statements about the future that makes me a bit uncomfortable. Because who can really tell the future? No one.

But screw it. Let’s do this anyway.

And yes, predicting the future has never been harder than right now. And that, in itself, tells us something important.

Let’s Get Some perspective first.

A year ago, the best model we had was GPT-4o. We’d just gotten o1 in September, the first model that could actually reason and think in steps.
That was just over a year ago.

In just fourteen months, the development in AI capabilities has been absolutely insane.

So when I’m sitting here trying to predict what 2026 will look like, I’m acutely aware that I’m probably going to be wrong about half of this. But that’s fine.

(If you want to take a time machine and see very old predictions from me, click here: Social media and HR anno 2013.)

My Three Trends for 2026

I’m not going to rank these. You’ll see why they’re all connected anyway.

1. The Digital Twin (Or Whatever We End Up Calling It)

I think we’re going to see the rise of what I’ll call the “digital twin.” Not in the sense that we’re all getting AI clones that can replace us, that’s not what I mean.

What I mean is this: an external memory that we connect to. A bot that sits in on things.

Most people already have the capability for this. You’ve got it in Teams. You’ve got it in Google Meet. You’ve got it in Zoom. Everywhere we meet, there’s usually some form of transcription option available.

Most people haven’t turned it on yet.

But I think 2026 is the year when transcribing meetings becomes normal. When having an AI assistant listen in becomes as common as taking notes used to be.

And here’s where it gets interesting: we’re not just collecting data anymore. We’re starting to use it.

Right now, it’s all very individual. “Look at me, I have transcripts of all my meetings!” Cool. But you can’t do anything with them? It’s just meetings being transcribed for transcription’s sake.

What’s coming is the ability actually to use that information to make better decisions. To have something that sits in, listens, summarizes, and helps you act on what matters.

That’s what I mean by the digital twin. Not a clone. More like an always-on assistant that knows what you know because it’s been listening to what you’ve been doing.

I believe this will be one of the first truly agentic use cases that actually works. We talk a lot about “agentic AI” but there’s still very little real business being done with agents.

This, I think, is where it starts.

2. Organization Over Technology

This is the shift I’ve been shouting about for what feels like ages (but it’s only three years). But I’m going to keep shouting about it because it matters.

2026 is going to be the year when we talk more about organizational change than we talk about technology.

To be fair, this shift already started in 2025. But it’s going to accelerate.

Don’t get me wrong here, I’ve been saying this since basically the dawn of time (or at least the last three years): AI adoption is not a technical challenge. It’s an organizational challenge. It’s a behavior change challenge.

That’s why I’ve been so adamant that HR needs to be involved. The knowledge we have in HR about change management, about learning, about behavior, this is what’s being screamed for right now.

Very few HR people have actually taken the leap yet. I get it, AI still feels scary and complicated. And you haven’t made time to really understand what this means for how work gets done.

I’ve been nagging about this all autumn. You have to carve out time. I’ve been sitting with several different teams in recent weeks, just building and experimenting. And you can see it, people start making progress. They learn stuff. The joy of actually building something, of dedicating time to rethinking how things work.

You have to do this.

But it requires understanding what you can do with these tools. I’ve been nagging about this for nearly three years now: (did I mention three years enough yet?) we need to add some AI competence to our skillset. We just have to. It’s not optional anymore.

And I think, no, I know, this trend is going to stick. It’s not going to swing back to “oh, now we’re just talking about technology again.”

Because organizations are slow, and the gap between what’s technically possible and what organizations can actually adopt keeps getting wider.

McKinsey is talking about it. IT consultants are talking about it. I’ve helped several larger consulting firms this autumn add this organizational competence to their service palette because they clearly lacked it.

So again: if we dare to take this step, there’s so much to gain. But I think this is a permanent trend. The conversation has shifted from tech specs to org design, and it’s not shifting back.

3. True Agentic Organizations (Maybe? Probably?)

Okay, I’m sticking my neck out here. And this is the Hail Mary one, but hey, we need something fun to look back at in a couple of years.

I think by the end of 2026, we’re going to see the embryo of truly agentic organizations. Not everywhere. Not in large enterprises. But in smaller organizations, we’ll start seeing AI agents that can actually take on work tasks, more or less independently, and just solve them.

Now, some people will argue this is already happening. But I think what we’re seeing now is still too automation-like. Every time I ask for examples, what I get back are basically fancy automation workflows. (Pls drop viable examples in the comments should you know of any.)

What I mean by agentic is this: the agent acts independently on incoming information. It understands and interprets without manual intervention. It makes decisions and executes without constant human babysitting.

Even when I build “agentic” solutions (which I’ve been doing), they still require manual input at key points. They handle sub-tasks, but there’s still human work needed somewhere in the process.

What I’m talking about is giving an agent a task like “handle my bookkeeping” and it just... handles your bookkeeping without further instructions.

I think bookkeeping might actually be one of the first places this happens. But somewhere in that ballpark, tasks where you can say “do this” and without much additional guidance, it starts solving it.

I think we’ll see the first real examples of this by the end of 2026. Not widespread. Not in every company. But visible enough that we’ll all see it’s possible.

(I could be completely wrong about this.)

The Bonus Trend: AI Fatigue Is Coming

Here’s the overall insight I think we’ll see playing out across 2026:

Resistance to AI is going to sharpen. We’re going to see fatigue.

Especially among organizations that have been doing the “innovation theater” thing, testing a little here, piloting a little there, but not really committing the time and focus needed to actually implement this stuff.

Those organizations are going to get tired. They’ll burn out on AI.

But here’s the split: the organizations that do commit, that take the time, that build the capability, that actually go deep, they’re going to see incredible results.

So we’ll have this widening gap. Some organizations saying “we tried AI, it didn’t work” while others are quietly building genuine competitive advantages.

I also think we might see some specific vendors struggle. Microsoft Copilot is going to have a harder time keeping up. ChatGPT might find the competitive landscape getting tougher. The model race is still very much alive; we saw that with the recent releases.

So What Now?

Look, I’m not pretending I can see the future with perfect clarity. Nobody can. The pace of change is too fast, and there are too many variables.

But these are the patterns I’m seeing. These are my bets.

The digital twin becoming normal. Organization trumping technology. The first real agentic use cases emerging. And a growing split between organizations that commit and those that don’t.

We’ll see how this ages. Maybe I’ll link back to this in December 2026, and we can all laugh at how wrong I was.

But for now, these are my predictions.

What are yours?

Subscribe now

Happy Tuesday!

Tomorrow is THE day. We’ll host our AI Day. 500 people in the audience and great speakers at the stage. It will be all about AI adoption.
So I’ve been thinking a lot about this lately.

Not just theoretically, because I’ve spent the autumn working with several organizations on exactly this. And if there’s one thing I’ve learned, it’s that AI adoption is an investment. A real one. Leadership time. Manager time. Space for people to experiment. There are no shortcuts.

So today I’m sharing the playbook I’ve been using. Feel free to steal it and make it your own.

And let me know, if it would be helpful if I dropped the actual workshop PowerPoints as well? Happy to share if there’s interest, let me know in the comments if that’s the case.

Now, let’s get into it.

If you want different results, you need to do things differently.

I know. Groundbreaking stuff. But I keep having the same conversation with organizations: “We trained everyone on AI and nothing changed.”

Then I ask what they did, and it’s the same approach they use for everything else. A one-day workshop. An email with links to tutorials. Maybe a lunch-and-learn.

And they’re surprised when adoption flatlines after two weeks.

I’ve spent the autumn working with several organizations on AI implementation. Not theoretical frameworks but actual rollouts. What I’m sharing here is the structure that’s working. It’s not complicated. But it requires investment.
Time, not just budget.

And I’m not the only one seeing this. McKinsey’s latest research puts it bluntly: organizations spend 93% of their AI budget on data, tech and infrastructure and only 7% on people-related issues like training, workflow redesign, and change management. That ratio is backwards.

Why most AI rollouts fail before they start

Most organizations do one of two things:

They train everyone at once. A big “AI inspiration session” where 200 people sit through the same AI-hype-talk. Two weeks later, maybe 10% are still using what they learned. McKinsey found that 70% of employees simply ignore onboarding videos and formal courses - they rely on trial-and-error and peer discussions instead.

The fix isn’t complicated. But it requires doing things in the right order.
And it requires accepting that this takes time.

Phase 1: Leadership alignment and defining success

This is where everyone wants to skip ahead.
“We already know AI is important, let’s just start training people.”

No.

If your leadership team doesn’t have a shared understanding of why you’re doing this and what success looks like, you’ll hit resistance the moment someone asks “but what’s the ROI?” or “why are we doing this?”

Here’s what this phase needs to include:

A leadership workshop where executives get hands-on with the tools. Not a demo. Not a presentation. Actual exercises where they experience what AI can and can’t do. Most executives I work with are surprised by both the capabilities and the limitations.
(and yes, I talk about this in the AI adoption series)

Then you need to define what success looks like. What will you measure? Usage statistics alone won’t tell you much someone can use Copilot daily and still not create any value. You need to connect AI adoption to something that matters to your business. Time saved on specific tasks. Quality improvements. Capacity freed up for higher-value work.

The companies getting this right are treating it as a leadership priority, not an IT project. Moderna, for instance, merged their HR and IT departments under a single executive specifically to align talent strategy with AI transformation. That’s a signal about how seriously they’re taking this.

Without clear metrics and leadership buy-in, you’re just hoping people find useful applications on their own. Some will. Most won’t.

Phase 2: Train the trainers (your AI ambassadors)

Here’s the thing about AI training: external consultants don’t scale. (And I’m saying this as an external consultant who does AI trainings…great business model to say “pls don’t give me a lot of work”.)

But I can train 30 people in a workshop. Maybe 50 if I push it. But if you have 500+ employees, the math doesn’t work. And even if it did, one-off training doesn’t stick.

What works instead? Building an internal group of 10-15 people who become your AI ambassadors.

People who:

• Are curious about AI (they don’t need to be experts)

• Have credibility with their colleagues

• Represent different parts of the organization

These people get intensive training. Not just “here’s how to use Copilot” but “here’s how to teach others, here’s how to handle resistance, here’s how to run a workshop.”

This usually takes 3-4 sessions over several weeks. With practice between sessions. And individual support when they get stuck.

The goal isn’t to make them AI experts. It’s to make them confident enough to help their colleagues experiment.

Done right, you now have 10-15 people who can train 20-50 people each in every session. That’s how you scale this effectively.

Phase 3: Manager enablement

I’ve seen this go wrong so many times.

A team member gets excited about AI. They start using it for reports, emails, analysis. Their manager notices and says “I don’t think we should be using that for client work” or “how do I know the AI isn’t making things up?”

Game over.

If managers don’t understand AI, they’ll block it. Not out of malice, out of legitimate concern about quality, accuracy, and risk.

So before you train all employees, you train managers. Specifically on:

• How they can use AI in their own work (they need to experience the value firsthand)

• How to set principles for AI use in their team (not banning it, but guiding it)

• How to encourage experimentation without losing quality control

Harvard Business Review’s research confirms what I’ve seen in practice: leaders must be able to show - not just tell - how AI makes work better. And middle managers often matter more than the C-suite for driving actual adoption.

Phase 4: Employee training and concrete use cases

Now - and only now - you train the broader organization.

You can either have your ambassadors run the training. They know the organization, they speak the language, they’ll be there for follow-up questions. The downside: it’s slower, and they have day jobs.

Or you can, of course, bring in external people like me who does the training. It’s faster, but you lose the internal connection. And the moment the external person leaves, so does the momentum.

My recommendation is, once again, to have ambassadors do it whenever possible. The relationship matters more than perfect delivery.

The other part of this phase is picking one or two concrete use cases and implementing AI in them properly. Document what works. Create a case study your organization can learn from.

This matters because abstract training (”here’s what AI can do”) loses to concrete examples. P&G ran a study with Harvard Business School where teams using AI were 12% faster - but more importantly, AI helped break down silos between R&D and marketing by acting as a translator between disciplines. That’s the kind of specific, measurable outcome that builds internal momentum.

This requires investment

I’ve said it before but I’ll say it again, this isn’t a quick fix.

You’re looking at 3-4 months minimum to do this properly. Leadership time in the early phases. Ambassador time for training. Manager time for enablement workshops.

And crucially: time for employees to experiment.

The organizations that succeed at this don’t just give people access to tools - they create space for people to learn.

Spotify opened up their traditional engineering Hack Week to all 7,000 employees specifically for AI experimentation. The principle is the same: dedicated time to play and learn, not just another task on top of an already full workload.

Kellanova (formerly Kellogg’s) took a similar approach with their “Kuriosity Clinics” - voluntary sessions where employees experiment with AI in a safe environment. Over 10,000 employees participated. The key insight from both these examples is building a culture of curiosity rather than a culture of compliance. Which once again, requires a time investment (and a budget investment as well).

And I get why organizations want to shortcut it. Training budgets are tight. We usually don’t have a lot of time at our hands in our daily jobs. Everyone’s busy.
“Can’t we just send everyone a link to some tutorials?”

You can. But you’ll get the same results you’ve always gotten from that approach.
(Hint: not much will happen…)

Here’s what happens when you skip steps:

Skip Phase 1, and you’ll have managers asking “why are we doing this?” halfway through rollout. You’ll spend twice as long justifying the investment as you would have spent doing the alignment work.

Skip Phase 2, and you’re dependent on external consultants forever. Your internal capability never builds. The moment budget tightens, adoption stalls.

Skip Phase 3, and your enthusiastic employees hit a wall of manager skepticism. “My boss doesn’t get it” becomes the reason nothing changes.

Skip the concrete use cases in Phase 4, and you have no proof that any of this works. Just vibes and hope.

McKinsey’s State of AI report shows that while 88% of organizations use AI, only 6% are high performers. The difference isn’t the technology. It’s whether organizations treat this as a change management challenge or just a tech rollout.

This is just the foundation

I want to be clear about something: what I’ve described here is AI literacy. It’s the baseline. It’s getting your organization to the point where people can use AI tools competently and confidently.

It’s not the end goal.

If you believe AI will fundamentally change how work gets done - if you’re planning for a future where AI agents handle significant parts of workflows, where roles get redesigned, where competitive advantage comes from how well you integrate AI into your operations - then this playbook is just step one.

I wrote about the different scenarios for AI’s impact a while back. If you’re planning for scenario 2 or higher, you need this foundation in place first. You can’t redesign workflows with AI if your managers don’t understand what AI can do. You can’t build AI-augmented teams if your people are afraid of the technology.

This playbook builds that foundation. What you do after that depends on how transformative you believe AI will be for your business.

How to know if you’re ready for this

This approach isn’t for everyone.

It’s for organizations that believe AI is strategically important - not just a nice-to-have.

It’s for leadership teams willing to invest time in the early phases, not just budget.

It’s for companies that want to build internal capability, not just check a training box.

If that’s not where you are right now, that’s fine. But then don’t be surprised when adoption stalls. And don’t blame the technology.

If you want different results, you need to do things differently. That’s it. That’s the whole insight.

What does your organization’s AI implementation look like right now?
I’m curious whether you’re seeing similar patterns.

Subscribe now

Happy Saturday,

November has been insanely intense. I’ve done 29 (!) AI lectures in November, sure, it is intense, but I also love what I do, and it is marvelous to meet so many organizations and people. You all help me learn so much!
This would not be possible without all the AI agents I have built, which help me with everything from writing offers to nudging me about my upcoming events.

December is looking almost as intense, and we have our sold-out AI Day on Wednesday, which I’m very much looking forward to! We’ll be 500 people coming together to talk about all things AI-adoption.

🧠 Topics I’m engaging with

Why you shouldn’t count on humans to prevent AI hiring bias.

Perhaps the most important article this week. Human oversight was supposed to prevent AI from warping hiring, but new research says it’s not enough. In a University of Washington study, 528 people worked with simulated AI systems to pick candidates for 16 different jobs. The findings were stark: when picking without AI or with neutral AI, participants picked white and non-white applicants at equal rates. But when they worked with a moderately biased AI, they simply adopted its biases. If the AI preferred non-white candidates, so did they. If it preferred white candidates, so did they. As lead author Kyra Wilson put it: “Unless bias is obvious, people were perfectly willing to accept the AI’s biases.”

My take: I find the study valuable. It’s one of the very few studies we have in this area. But I see two clear limitations. First, they didn’t use true state-of-the-art models; they relied on open-source embedding models, not frontier systems like Gemini, ChatGPT, or Claude. That limits what the results say about today’s best AI. Second, participants weren’t recruiters. They used general online panel users, not trained recruiters. So the decisions don’t reflect real hiring expertise or context.

That siad, what the study still makes very clear:

1. We can’t trust AI recommendations blindly. People follow AI even when it’s clearly wrong.

2. We must train specifically for AI-supported hiring. General AI knowledge is not enough. We need practice spotting and resisting AI bias in real decision flows.

3. We need to push our vendors harder. ATS and screening tools must prove how their models behave, how they test for bias, and how they monitor it.

If humans follow AI this strongly, the burden on vendors must increase, too. And this is HUGE if you’re within the EU, this is exactly what the EU AI Act is designed to regulate. High-risk AI systems in employment decisions will require transparency, human oversight that actually works, and documented bias testing.

Microsoft WorkLab: “AI at Work: Which future of jobs are we building toward?”

Jared Spataro lays out four possible futures based on growth and employment. The early data is clear: companies using AI to augment their workforce are hiring faster than those automating for efficiency alone. Entry-level tech jobs are shrinking, while experienced workers who can direct AI tools are seeing wage premiums. His conclusion: we can achieve high growth AND high employment – but only if we invest in people as aggressively as we invest in models.

This is exactly what I see with my clients. Those treating AI as a headcount replacement are missing the point entirely. The ones who see it as an amplifier of human capability, and actually invest in training their teams, they’re the ones winning. The question isn’t “what will AI do to jobs?” It’s “what are we willing to do to get the future we want?”

Stanford/Carnegie Mellon Study: AI agents work 88% faster but with quality gaps

A new study shows AI agents complete tasks 88% faster at 90-96% lower cost than humans but with 32-49% lower success rates. It’s also worth noting that when humans and agents worked together, task completion increased by nearly 70%. Agents handled the programmable steps, humans took care of what required judgment or creativity.

Interesting data! Not “AI takes our jobs” or “AI is hopeless” but nuanced evidence that hybrid models work best. 70% improvement when human and machine collaborate. I strongly believe that this is how we should do it during the coming years, designing workflows where the right tasks go to the right “worker”, whether that’s a human or an agent. Stop asking “will AI replace us?” and start asking “how do we design the collaboration?” and then design that intentionally.

IBM: “AI Agents in 2025: Expectations vs. Reality”

IBM researchers cut through the hype and deliver a realistic picture. Marina Danilevsky: “You’re still going to have cases where as soon as something gets more complex, you’re going to need a human.” She describes a human-in-the-loop vision where agents settle into an augmented role rather than a replacement one. Key insight: only 13% of employees see agents deeply integrated into their daily workflows yet.

I appreciate this sober analysis. We’re still in the experimentation phase and that’s okay. But organizations starting now with “narrowly defined, low-risk domains” like FAQs and leave requests are building experience that becomes invaluable as the technology matures. The worst thing you can do right now is wait. The second worst thing is to go all-in without understanding the limitations.

🔁 Important updates

Claude Opus 4.5 released

Anthropic launched Claude Opus 4.5 – their most powerful model yet. It beats GPT-5.1 and Gemini 3 Pro on coding benchmarks (80.9% on SWE-bench, first model to reach that) and pricing dropped dramatically to $5/$25 per million tokens.

AI could double US labor productivity growth

Anthropic analyzed 100,000 real Claude conversations to estimate AI’s productivity impact. Key findings is that tasks that typically take 90 minutes without AI are completed 80% faster with AI assistance. Anthropic Software developers capture 19% of total productivity gains, followed by operations managers and marketing specialists. The economic extrapolation is that current-generation AI models could increase the U.S. annual labor productivity growth rate by 1.8%, doubling the average rate of growth since 2019.

We’re moving from the age of scaling to the age of research.

Interesting interview with Ilya Sutskever on why we’re entering an “age of research”, deeper ideas about learning, not just more data and compute. Why current models generalize worse than humans, emotions as a model for robust learning, and why the path to superintelligence runs through continual deployment, not lab breakthroughs.

Inside Spotify’s AI Transformation - Spotify CHRO Anna Lundström

My former colleague, Anna Lundström, discusses Spotify’s AI journey. I’ve been puzzled by their silence lately. They used to be a driving force in adoption and new-tech thinking, but it’s been quiet. Until now!

Their approach has been to frame AI as making employees “future-ready” for their careers, not just for Spotify. They’ve opened their (famous) hack week to everyone, not just engineers, and said “work on something AI related”. Plus training, tool licenses, and a cross-functional steering group to coordinate.

Early wins so far have been time savings on admin work, freeing people for strategy and the stuff you never get to.