Friends,

By the time this lands I’ll be mid-air from Africa back to Australia. Transiting through the middle-east, so 🤞 nothing exciting happens there while I’m passing through 😬

I’m pleased to say I managed to see each of the Big 5, rounding out a peek at the elusive leopard in our second last day of safari. I also managed to (both learn about and) see all of the less affectionately known Ugly 5 — although I can confirm that Hyena babies are cute af.

You see pics of the crazy cool animals here and it just doesn’t do justice to seeing them in real life (but here’s one anyway) — they truly are remarkable.

Highly recommend you get over here if you ever have the chance.

It’s given me a renewed appreciation both for nature and for taking time off.

From next week I’ll be back in Aus and back to doin’ my thang.

In the meantime, enjoy ✌

Matt

(PS: I realised last week I butchered the link to my merch appearance in DRIP DROP, so here it is properly)

In case you’re new here (or just missed it) here’s the past three editions of the FNDN Series:

When startups start hiring internationally, the same questions come up every time:

Our founding team has sat in the People and Ops seats at fast-scaling startups. We've lived the midnight visa panic, the payroll surprises, and the compliance headaches six months later.

So we wrote the guide we wished we'd had.

Boots on the Ground is a free guide that covers the six scenarios where relocating someone genuinely pays off, how to identify who's right for a move, what it actually costs (spoiler: almost always less than not making it), realistic visa timelines by route, and the human side most guides skip - the conversations about partners, families, and expectations.

It's written for founders, People leads, and ops leaders making their first international hire or running a handful of moves a year and wanting to do it properly.

Know a startup Head of People looking for answers 🙋 why not forward this to them for some instant karma? ✨

This edition is based on the latest episode of the FNDN Series podcast, with Benjamin Langner.

Ben Langner is a global HR executive focused on the work behind the work. He’s known for a grounded, unfiltered perspective on modern HR, leadership, and building people systems that actually hold up under real pressure.

If your CEO has asked you "have you put it through AI yet?" more than once this month, you're not alone.

The pressure to adopt is louder than ever. The discipline of deciding what you won't automate is quieter than ever. Resolving that imbalance is the most important AI work the People function will do this year.

But of course, most of us haven't been given the extra time to do it.

Ben Langner came on the FNDN Series podcast recently to talk about it. He's an in-house People leader and one of the more thoughtful voices on what the function should and shouldn't hand to a machine.

The thing that stood out from this chat, was a discipline he runs before any AI conversation in his business starts.

He calls it his no-fly list.

The no-fly list is short. It names the moments where, however good a tool gets, he won't take the human out of the room. 

He builds it before he greenlights a single automation, not after.

This is the piece I'm seeing most People teams pressured to skip right now. Adoption is being pushed top-down by leaders who haven't sat in the conversations the technology would be handed. 

The conversation goes straight to "where can we use AI?" before anyone asks "where shouldn’t we?"

Ben's view: you can't answer the first question well until you've answered the second. Without that clarity, the human moments get automated by default. The pressure to adopt closes that gap whether you mean to or not.

What I like about the no-fly list is how cheap it is to build. A working session with your team gets you to v1. The output will be rough, and that's fine. v1's real value is the shared answer it gives your team when the AI question comes up next.

So what ends up on a no-fly list? Ben gives two examples relevant to any business, and he’s explicit on the kinds of categories that should be considered: performance, personal hardship, exits, and the moments that decide how trusted the People function is.

The first is performance management. 

AI is excellent at the front-end: surfacing the data and drafting the benchmarking analysis. Work that used to take People teams weeks now takes minutes. But the conversation that follows, telling someone they aren't meeting expectations, isn't a candidate for automation. As Ben put it, no business hires people in order to fire them. The conversation is a recovery effort, and recovery has to come from someone who can listen and respond in the moment.

But the second example was the one that stuck with me. 

Think about the time an employee comes to you with something deeply personal — like a miscarriage. They want to understand their time-away options and the support available in one of the worst weeks of their life. No one walking into that conversation wants a chatbot. They want a human who can sit with them and respond with care.

A useful frame here comes from medicine. A futurist I once heard speak argued that AI will get exceptional at diagnosis (because patients are often more candid with a machine than with a doctor). But the care doesn't change. People still want a human for the treatment. The same logic applies to people work: use the tools to surface what's true and keep humans for the moments where the response itself is the product.

Ben's approach has two phases.

Phase one is your People team only. Get the people who know the lived reality of these conversations in a room. Ask them where a machine running it would land badly, and where employees still expect a person across the table. I'd add two prompts to the round-table: which moments make us trusted, and which would feel like a betrayal to hand off? 

You'll come out with a working list inside an hour.

Phase two is taking the list to the AI-curious people in your C-suite, the early adopters already running prompts. Walk them through it and get them to pressure-test it. That conversation positions you as the one bringing structure to AI inside the business, a different posture from being seen as slowing it down.

This approach matters more than the list itself. If your CEO is pushing AI and you walk in with a "no" position, the conversation rarely goes your way. Walk in with "yes, here's where we'll move fast, and here's where we won't bend on the people experience." You've reframed the conversation.

That's the only AI policy that holds up when the pressure to adopt is loud.

Once your list is drafted, the work of keeping a human in the loop becomes behavioural. We outlined three habits that hold the line.

First, plan before you prompt. Ben starts with what he's trying to achieve and lets the model ask the clarifying questions. The interactive front-end produces materially better output than a one-shot brief. I've adopted this myself recently (using Claude to prompt me in Cowork), and the quality of what I get back has shifted. Ben reckons Claude has "changed a lot of lives in the HR space" over the last 30 to 60 days. This habit is a big part of why.

Second, read the thinking. Don't skip past the reasoning to land on the final answer. If you only read the output, you'll miss where the model has misinterpreted the inputs or invented something. The consequences of a wrong answer in people data aren't theoretical.

Third, fact-check the output. AI is exceptional at sounding confident about claims it can't defend. Treat its work the way you'd treat a junior analyst's first draft. Useful and fast, but not yet ready to ship.

Ben sees AI as a gift to People teams. He's right. The technology returns time and capacity, and it lets the function show up in the moments that matter most. But only if you've decided what those moments are.

If you don't already have a no-fly list, it might be something to consider. Pull your team together and draft v1. Bring it to your C-suite forum and make it the foundation of every AI conversation that follows.

If you enjoyed this post or know someone who may find it useful, please share it with them and encourage them to subscribe.

A roundup of the most interesting news from the week:

When you’re ready, here’s three ways I can help you:

1. Tools & resources
Resources and tools that give you what you need to build your own startup compensation practices.

2. Comp consulting
Building compensation practices that are clear, fair and competitive for startups.

3. Startup People Summit
A 1-day virtual event focused on creating the playbook for startup People practices. Grab recordings from past events, or subscribe to the newsletter to join the next event.