Confidence with AI doesn’t come from training.
It comes from proximity.

In my podcast conversation with leadership coach this week, what stood out wasn’t a tool or a technique. It was how people actually build confidence in practice.

Fay said she’s noticed that her clients become more confident using AI when these things happen first …

1. They see how others are using it.

2. They ask questions, sometimes very basic ones.

3. And eventually, they try it for themselves.

That sounds simple. But in most organisations, all three of those steps are missing.

If you want something practical to try as you read, I’ve created a short companion tool you can download. It gives you a simple way to take control of AI outputs and start building confidence straight away. You can use it yourself or to build confidence across your organisation. I’ll share the link later in this article.

Subscribe now

Why it still feels overwhelming

Overwhelm is a real thing.

You sit down with something you need to get done. You open an AI tool because you know it should help. But instead of feeling clearer, you get something long, slightly off, or just harder to work with than expected.

So you tweak it. Or ignore it. Or close it completely.

That’s not a capability problem.
It’s not even really a confidence problem.

It’s a lack of control and a lack of a clear place to start.

Subscribe now

What actually builds confidence (in real life)

What Fay described reflects what’s happening in practice everywhere. Confidence builds through exposure.

Seeing

People need to see it used in real work.

Not polished demos. Not ideal use cases. Just something familiar:

• a draft email

• a meeting agenda

• a summary that gets you part of the way there

When you see AI applied to something you already understand, it becomes easier to engage with.

Asking

People need permission to ask simple questions.

• “What did you put in to get that?”

• “Could I use this in my role?”

• “Is this the sort of thing you use it for?”

These questions are often where confidence starts to build. But they are rarely surfaced explicitly.

Trying

Then comes the most important step.

Trying something small.

Not a transformation. Not a big workflow. Just one task.

Because confidence does not come before using AI.
It comes after it.

What’s missing at work

Most organisations skip these steps.

They move straight to:

• tools

• training

• policies

All important. But none of them create confidence on their own.

What is often missing is:

• visibility of how people are actually using AI

• shared learning, including what did not work

• space to try things without getting them perfect

One simple way to take control

If you are feeling overwhelmed, it is often because you have given the AI too much space.

So try this.

Whatever task you are using AI for this week, add one line to your prompt:

Make sure your answer does not exceed [X] words.

That’s it.

A small constraint changes the experience completely.
You get something shorter, clearer, and easier to shape.

And you move from reacting to the output to steering it.

Reduce the effort each time

There is a second shift that starts to happen once you have tried this a few times.

You notice you are repeating yourself.

• Explaining your role.

• Explaining the context.

• Explaining what “good” looks like.

When that happens, I suggest you don’t keep typing it over and over.

If you find yourself repeating the same context each time you use AI, consider creating a short set of context instructions. This becomes a single source of truth you can reuse rather than starting from scratch. Different tools handle this in different ways, but the principle is always the same.

That one document, note, or page becomes the thing that keeps the AI grounded and reduces the effort every time you use it.

A simple way to start (with someone else)

One of the simplest ways to build confidence is to not do this alone.

Share a simple task with a colleague. Ask them:

“Have you used AI for anything like this?”

If they have, ask them to show you:

• what they did

• what they put in

• what came back

Then try it yourself. Most people are only too happy to help because they are flattered that you trust them enough to ask.

You don’t need a perfect prompt.
You just need a starting point.

Where this goes next

If this still feels harder than it should, or you are not quite sure how to shape what comes back, that is completely OK. This is much more common than you may think.

It’s also exactly what I will be working through in a 90-minute session on 10th June:

AI Without Overwhelm : A workshop for Neurodivergent Professionals

Click this link to register on Eventbrite.

Not tools. Not hype.

Just how people move from:

• feeling overwhelmed
  to

• feeling in control

In a way that actually fits their day-to-day work.

Download this confidence booster

If you want to get a head start, you can download this simple table that guides you through:

• how to control AI outputs

• how to build from a first response

• what to watch out for

• how to reduce repetition over time

Ai Confidence Booster

159KB ∙ PDF file

Download

Download

A final thought

Confidence with AI is not something you switch on.

It builds, quietly, through:

• seeing

• asking

• trying

And if those three things are not present in your environment, perhaps it’s time to introduce them!

You can tune into the full conversation with Fay Wallis here on Substack or wherever you get your podcasts.

Thanks for reading as always, and if you find this article useful please share it with a friend as it may just be what they’re looking for!

Share

You can catch my solo podcast tomorrow where I reflect on my week in light of what I learned from my guest and share some insights which you may find useful. See you there!

Before anyone builds systems or develops deeper expertise, they usually start by simply seeing how someone else is using AI in their day-to-day work. Drawing on her experience as a coach, Fay describes this shift clearly, from hesitation and overwhelm to curiosity sparked by watching, asking, and being shown.

I’m looking forward to exploring this with you, and if you’d like these posts straight to your inbox, you’re very welcome to subscribe.

Subscribe now

TRANSCRIPT

(00:00:00):

Welcome to the Sovereign Career Podcast.

This podcast, if you don’t know already, is part of the Sovereign Career Hub,

and its job is simple. It’s to gather honest, practitioner-led field evidence about how AI is reshaping work, what’s changing, what’s breaking, and what people are figuring out in practice.

What surfaces here becomes raw material for the rest of the hub, deeper articles, practical solo audios, and ongoing community conversations.

So no pressure on me then! I am joined today by the wonderful Fay Wallis and it’s fantastic to have you here Fay I’m going to of course ask you lots of wonderful questions but first I’d really like it if because we don’t have everybody watching the video I presume and there’s a lot of people might be listening I hope there are anyway!

Then I thought it would be a nice thing if you could just describe the room you’re sitting in, where you are located maybe in the country and really what you enjoy about that lovely beautiful office that I can see behind you. I’m used to looking at it because we often chat online but I’d like you to just describe it to others if you could. What makes you happy about it?

(00:01:14):

Of course. Oh well, thank you so much for inviting me to be on the show, Carolyn. It’s really exciting to be here and as far as where I am in the country, I am in a tiny little town called Banstead, which most people have never heard of, which is in Surrey, so it’s just outside of London in the UK. And I’m joining you today from my home office, which also doubles up as our spare bedroom, haha, and when I took over this room I ended up completely redecorating it because it was really, really dark and I do so much of my work online I thought, oh, I don’t really want people thinking it’s a bit dark and gloomy whenever I see Fay. So I spent ages. It’s a sort of duck egg blue and behind me there’s a grey minimalist sofa which doubles up as a sofa bed so that if we have got anyone coming to stay it means they’ve got somewhere to sleep, so those lucky people get to sleep in here with all of my office equipment. And I’ve got an abstract painting on the wall above the sofa as well and that’s probably what makes me happy because it takes me ages to choose any sort of pictures. I don’t know why, I find it quite difficult decorating my house or finding paintings. And I loved this one when I saw it and saved up for it, so I felt super excited when I got to put it up on the wall.

(00:02:30):

Oh, that’s beautiful. Oh, that’s lovely because actually it means more when you’ve invested in it yourself, doesn’t it? And, you know, you’ve saved up for something like that. Oh, that’s wonderful. Well, just to say thank you for that, and also just to set the tone for the conversation for everyone who’s listening, Fay, you’re not here as an AI expert today, are you? Instead, we’re interested in how AI shows up in real working life and the choices people are actually making, you included, of course, but also because you speak to so many HR practitioners, professionals, HRDs, et cetera, in the work that you do. And we will be talking about that as we go through, I’m sure, because we’re bouncing off some of those wonderful things you do, including, for example, the Inspiring HR Leadership Programme. But I thought it would be really nice to ask you to start off. I do this because it helps people to place you. Where would you place yourself on the AI spectrum right now? Can I ask you that?

(00:03:34):

I think if the spectrum is going from zero with no confidence about AI at all all the way up to 10 where you’re an AI expert I’d probably put myself at about a seven. I do feel fairly confident with trying out different AI tools. I’m fairly excited about it a lot of the time, although of course sometimes it can be a bit scary as well. I keep fairly up to date with watching videos and things like that and reading resources about AI so I am certainly not an expert, but I’m a happy dabbler and happy user of AI.

(00:04:09):

Well, that’s nice to know. I mean, do you find, while I’m just on that quickly before we leave that point, do you find that is fairly normal for people to be at that level or do you feel that you’re probably a little bit more advanced than some of the people you’re speaking with a lot?

(00:04:24):

I think it’s probably not normal and I’ve learned that the hard way actually when I have talked about AI with other people and then just seen this absolute terror and fear on their face and kind of shut down, no I don’t want to talk about it or even consider it for my work. So that made me realize, oh gosh, I’ve got to go a bit more slowly here and check with people when I’m speaking to them what their thoughts are and their feelings are about AI because I don’t want to be off-putting when talking about it with people. So I’ve always been quite techy. I quite like tech stuff. My sister’s a website designer and just from her helping me with my website she ended up teaching me how to do loads of it. I’d probably be a terrible website designer if that was my job, but I’m certainly not scared of getting stuck in there and testing trying new things on my website and with other aspects of my work as well. I’ve always liked embracing technology. When I was in my HR career I remember the company started using Microsoft SharePoint for the first time and they asked for volunteers to get extra training and become SharePoint super users and I put my hand up straight away and absolutely loved it and ended up building a whole HR intranet on SharePoint at the time. So I think I have always been quite into tech and I just see AI as the next evolution of tech really, but I mean there are so many people who know so much more than I do.

(00:05:48):

That’s really interesting because obviously your work with Bright Sky, Bright Sky HR, which is your consultancy, is really giving you access to a lot of people with a wide range of seniority, I would expect, and also responsibility. And so do you feel that they are starting to talk to you? Are people starting to talk to you about, oh, you know, a little bit of insecurity maybe about, because I know you talk about strategy a lot. So then that’s great, you’re coaching people, your coaches, to be more strategic if that’s what they want. And so in that pursuit, do you find that AI enters the conversation?

(00:06:35):

I think the vast majority of my work nowadays is delivered through my Inspiring HR Leadership Programme and because AI is such an important topic and impacting the workplace so much at the moment it’s something that I do talk about as part of the program and as we work our way through it I might share tools, so different AI tools and suggestions of how they could be used, and that really is where I realised that not everyone is comfortable with AI because I got very overexcited and the first time I started introducing some of the tools I could just see real resistance from a couple of people where they hadn’t started embracing AI yet and they were quite cautious of it.

(00:07:18):

How long ago was that, Fay? Yeah, and have you noticed, because obviously this is a very fast-paced development, isn’t it, so are you noticing any change maybe? I mean, I was listening to a podcast, one of my favourite podcasts this morning on AI, well I say favourite podcast, it’s why I do my research, a lot of it, and they were saying that’s been a change in the last six months, I think it’s since November, where are we now, you know, since November 25 there’s been a change in how people, this is an American survey by the way but it’s by KPMG or someone like that, and it’s millions, I think they actually, I’m sure they said they’d interviewed over a million people, which seems crazy, or at least surveyed over a million people. And the result is that there’s been a drop in the amount of people who feel very fearsome, who are fearing AI. And I think it’s quite a big drop. It’s gone from 23% down to about 17%, which is pretty amazing, that’s in that six months, and of course before that it was higher and before that it was higher but it’s going quite fast. Are you noticing that change?

(00:08:31):

I’m not sure that I work with, you know, that’s a million people they’ve surveyed and I definitely don’t have a million people who I’m able to work with on a one-to-one or group basis, but I would say it’s the same as any adoption of a new thing. I remember reading a book by Seth Godin, who’s a marketing expert, and he talks about when a new product is released you always have people approaching it at different stages, so you will have some people who just love anything new and shiny and can’t wait to get their hands on it and try it, and then you’ll have people who are happy to give it a go and interested in it but they probably want to have heard a bit more about it from those super excited people, and then you have people who are slower to adopt it, and then you’ve got people who are just a bit fearful and resistant to change or to try a new product. And I think it’s exactly the same as anything else to be honest, so as time goes on I guess more and more people are going to become comfortable with it and certainly when I talk about it in my program I would say the majority of people are at least comfortable with it and maybe using it a bit, and we might have one or two people who are really advanced and really, really into it, and there might just be one or two people who aren’t massively keen on it. So I think that spread probably reflects what Seth Godin talks about far more articulately than me in that book, that you know it’s human nature, that’s how we tend to respond to any sort of new product whether it is technology and AI or whether it’s something else. I mean I remember when the iPhone first came out and my husband bought me one and I said why have you got me this, like my Nokia’s fine, it’s all because you’re gonna... oh sorry, he disappeared for a minute, they’ve come back now though, and he said oh well because it’s got apps and I was like well why do I care about apps? He loves anything new and will always get his hands on it immediately and I’m probably that one person down where I’m quite comfortable with tech but I probably need to be shown it a bit before I start really getting into it.

(00:10:34):

Yes, we are all at different stages, aren’t we? I mean if you think about the adoption curve, how people are, you know, some people are laggards, or they’re even called laggards, or they were when the adoption curve first came out, that bell curve thing that we’ve seen, and crossing the chasm used to be a thing that technology had to do. It had to leap over this chasm from the early adopters, the enthusiasts, the pioneers who, you know, like your husband and possibly me to some extent, would be clicking away and not worrying if it was ......... the White House or whatever. This is what we used to worry about. Will it ......... the White House was a thing people said. If I press this button, what will happen? So don’t worry, it’s not going to. But I think if we really sort of just are so enthusiastic about using it, it can be a bit overwhelming for people who are just trying it for the first time, and we can be a bit too enthusiastic and they’re sort of like it’s overwhelming. And I think that now we see so much on social media, like LinkedIn for example, which is where I spend a lot of my time, I know you do, and you know we see it there, is it a good thing, is it a bad thing, is it evil, is it good, and you have this crazy kind of dichotomy going on and it’s overwhelming to people who are not yet part of the conversation and that can send them in the other direction. However, I would encourage people not just to use tools but to get to know about AI a bit more, to understand it a bit more, because clearly that’s probably something that you do, isn’t it? You understand a bit more about how it works and have you thought about the ethics and things that go with it? Do these things occur to you from time to time, you know?

(00:12:28):

Oh yeah, very much so, and I think where the majority of my work is with HR professionals, you know, we’ve got to be so careful and mindful of ethics as a profession, and where I’m a trained coach as well there’s a huge part about making sure that your practice is ethical. So it’s definitely something I think about all the time as far as using it is concerned.

(00:12:52):

Yes, and how it might impact. I mean, there’s two narratives going on. There’s the big one, the societal one. How is AI going to impact our society and us as people and jobs as a whole? That’s the macro level. And then you’ve got more of the micro level within the organization where HR is quite concerned about what it’s going to do with the data and so forth. And, you know, have you got the guardrails in place in relation to your own organisation? Sometimes in relation just to your own family. I mean, you’ve got to start, you know, you look at all different levels of it. It’s relevant at all levels. But that can be, I think others, I mean, I’m working with several groups of people. HR is one of them. And I find that HR is quite reticent by comparison. It’s risk averse, as we all know, process driven as we know and when you look at some of the other groups that I work with who are more kind of like well we just have to get this thing done let’s just do it you know we take the risk they’re more interested in what AI can do for them and you know what the outcome the results if you like more of a results based approach does that bear out what you know about how AI is affecting and impacting the HR teams that you work with?

(00:14:16):

I think the discussions that I tend to have wouldn’t necessarily be talking about it in that depth because when I’m talking to people or sharing tools that I think could be useful it’s very much thinking about how it can be helping them in their day-to-day work or with their team. We don’t necessarily do a deep dive into, you know, how it’s being used throughout the organization. One thing I have been conscious of is the fact that people have to be really aware of what their organization’s policy is when it comes to AI. So at first, I think when I was really super enthusiastic about it, I said, oh, look, everyone, I found this great tool, you know, if you want to give it a try. And one person said, well, no, I can’t touch it because actually it’s outside of, I can’t remember which AI system is that the right word, I don’t think system’s the right word, which AI platform they had approved within their organization. I know lots use Copilot, for example, lots use Gemini, and this would have been outside of it and so, you know, of course I don’t, oh yes, of course, you know, I work for myself, I have my own business so I can make decisions as to, you know, personally if I think I’m not sharing any confidential information in that tool, I’m just trying it to see if it can, you know, create a slide in a certain color, I’ll give it a go, whereas of course you haven’t got that luxury if you’re working for an organization where a decision’s been made that we need to make sure we’re being careful with what people are using. So we’re not just going to say people can just use these tools. So I think that probably my explorations of it with the HR community, I’m not working with them as an AI specialist looking at how they’re going to be using AI throughout their entire organization. I’m working with them as an individual and as part of the program that I have may share certain tools that I think could be helpful for the topics that we’re covering on the show rather than necessarily talking about okay what do we want to do as an HR community when it comes to potential job loss because of AI. If someone brought that as a challenge to the group coaching that we have we could talk about it, but it’s not something that’s actually shown up that someone’s wanted to talk about so far in the sessions, so at the moment it’s more about HR getting on board as individuals at the moment. In my sphere it’s just very much that my program looks at certain aspects that I think are helpful for HR leadership, so we look at confidence, we look at influencing skills, we look at building strategic skills and as we work our way through that if I think actually there’s a prompt or there’s a tool that I think could be a helpful addition to this session that people might like to know about if they want to learn more about AI or start feeling more confident using it or this just could be handy then I’ll share it with them. So that really is it, but as part of the program as well there are opportunities for one-on-one coaching and group coaching and you can bring whatever challenge you’re experiencing at work to that conversation, so if someone wanted to bring AI to that conversation they could, but so far that hasn’t happened.

(00:17:32):

That’s really interesting because you know my work with Jobscaping™ looks at how we make those decisions as to which tools we’re going to use and for what type of work, and I break down based on tasks because I think if you look into workflow rather than a department, so if you look at a workflow you might say onboarding a new member of staff for example, well that’s going to cross over maybe a few departments for example, so if you’re breaking down the tasks and then you’re looking at it and saying which of these things are above the line below the line in terms of what I’m prepared to give to AI and what I’m going to keep as a human task, you know, you’re making a decision really whether or not you’re handing it off completely to AI, whether or not you collaborate with AI or whether or not it’s a no-go, no AI, thank you. Do you have any no-go, no AI tasks that you would say absolutely not, this is a human judgment thing and I’m doing it for myself?

(00:18:27):

I don’t know that I’ve thought about it from that perspective because as far as how I use AI for myself in my business of course I’m not an HR practitioner anymore. I am a coach and a trainer and have my own business so the ways that I use AI are probably slightly different. So things that I might use it for that are incredibly helpful are I’ll be posting on LinkedIn for example, I know that you already mentioned that you like being on LinkedIn, so do I, and in the old days I would have completely written the post by myself whereas now I might just do a brain dump and say this is my very rough draft of what I would like to say in this post, please can you improve it for me, so it’s like having my own personal copywriter that I have with me. If I need to create a graphic to go with the post I’ll say this is what I would like to have in the graphic, can you create it for me? Here are my brand colors. And then I can use that. I have a podcast, HR Coffee Time, so for that I might have interviewed somebody and then I’ll have the transcript and I will put it into AI and say this is the transcript of my episode. I want to record an introduction and I want to record an outro where I entice people about why it’s an interesting episode to listen to. Can you give me some suggestions? Here are some key points I want to include. So, you know, I’m very much using it in a practical way to support me rather than thinking deep. Obviously I’m interested in the impact of AI within workplaces, but I am conscious that I’m not an AI specialist. The challenges that HR professionals tend to come to me with aren’t necessarily around deep use of AI and the ethics etcetera, so I’m just tending to look at it from a very practical how can this help me within my work or how do I think this could help my clients actually if I find a new tool that looks like it could be quite handy, whether that’s as a thinking partner or refining a presentation for them or helping them draft a board report for example. So yeah, I think that I strongly suspect, or I know, you will have such a deeper level of expertise about how a whole organization may be implementing AI and having recommendations for how they can use it, but that’s not really what falls within my wheelhouse.

(00:20:57):

Exactly and it’s really interesting because that is exactly the perspective we want, you know, it’s your perspective. I think a lot of the stuff that we hear about how organizations are implementing AI, quite a lot of it is PR-based and PR-driven. And in reality, what is changing is much more interesting. What is changing in terms of how you and I and other people use AI for ourselves? I mean, I use it for brainstorming a lot, okay? And I do use it as a thinking partner, but I push back a lot and I think that’s important, so that would be my contribution to a conversation like this would be yeah, I do use it a lot but I push back a lot as well because I don’t accept everything it tells me. Yesterday, or last week actually, speaking with Sharon Green and she was, I mean Sharon loves to write, doesn’t she, and she said yeah okay AI can write really well probably lots of stuff but I don’t want to give it that task. So is there anything that you know AI can do but you don’t want to give it, or there might not be, but I’m thinking about your coaching practice now because it’s a very human-to-human thing, isn’t it, coaching? So is there a point at which you sort of have to say no AI here at all or do you find that with almost all your coaching practice there’s no AI involved? I’m trying to think if there would be any, like if as part of the coaching someone’s saying something and I think oh this AI tool might be useful for you, I might share it with them, but I’m not using AI in the coaching itself at all. Do you use it for something like a bit of research maybe in your coaching practice? If somebody had brought something up and you thought that was interesting and you started to look at it after the conversation you’d had with them, after the coaching session, would you use AI to help you research anything or do you still do that yourself?

(00:22:59):

I might use it to research something. For example, if I’ve got a topic for my podcast I want to talk about then I definitely prefer it to just using Google because it’s much more likely to give me good links. I am, just like you said about pushing back, if I want to share a theory or an idea I want to make sure it’s a valid one, and I find it’s easier with AI tools to say okay make sure that you’re giving me the references for this and give me the links so I can read those references myself. But I think as far as coaching is concerned, when it comes to coaching I take quite a purist perspective with coaching and so from a purist perspective what that would mean is as a trained coach rather than telling people what to do or giving them mentoring and advice about a situation I’ll normally be asking them questions and helping them explore their own thinking. So within that kind of framework I’m not normally going to be using AI in any capacity, but the only time there are times with coaching where I am more directive and I will give more advice and so that might be in something like interview coaching, and in interview coaching yes I may use AI. So if someone says to me I want to prepare for an HR business partner interview for example, I’ll say okay do you want to first of all ask the employer, ask the interviewer, if they’re happy to share the questions with you because people are always surprised that often companies will do that, not always but there definitely is a chance they will. And if not I’ll say ask them for the full job description and what context do you know about the organisation, and then I’ll be able to put that job spec into AI and say I have a client who’s going to be interviewed for this role, you can see in the job description what the key competencies are, can you generate some potential competency-based interview questions for me to practice with them? And then I’ll be able to say to them, because if you say to someone what questions are you worried about they’ll often sometimes go what, I don’t really know, or I haven’t had an interview in ages, I’m not really sure what to expect, but if I’m able to say okay well here’s some potential questions they’ll often go oh yeah, that one, that one would be really hard, I’m not sure what I would say, so then I’m able to say okay well let’s have a practice. So that is one way actually I suppose within my coaching practice that I do use AI.

(00:25:30):

And it’s great for role play as well, isn’t it? I mean, I don’t know if you’ve ever used it for that, but if any of your clients have ever used it for that, that’s becoming quite popular in the sense that if you download, for example, ChatGPT to your phone, you could just have a chat with it about the job you wanted to apply for, tell it to interview you. You could even give it those questions that you’ve just mentioned and then say that you want it to give you some feedback on your answers because obviously that’s the whole point, so you want to move forward, so that’s an interesting way that it’s being used. And I read that it’s because people feel much less judged when they’re just talking with a machine and of course it’s being completely objective, isn’t it? Well, as objective as it can be based on the information it’s being given of course, but I mean objective in the sense that it isn’t looking at you and judging you based on your age or your location, your demographic, your race, anything like that. It’s just basically helping you with that role play you’ve asked it to do.

(00:26:43):

Well, and I think with that example it’s probably why lots of coaches are worried about AI doing them out of jobs. People are so comfortable with just asking AI for support, like whether coaching is still going to exist as a profession. I like to think it definitely will because although yes, it’s helpful to be using ChatGPT for these things, I think it’s not quite the same as speaking to an actual person.

(00:27:11):

That’s interesting, isn’t it? Because this conversation about the expert comes up quite often, not just in relation to coaches but in relation to all sorts of subject matter experts, and I’ve been looking at it in terms of learning, for example learning design. So yeah, it doesn’t replace the expert I’m told, but what it does do is speed up the process dramatically and let them get to the interesting bits more quickly. There’s a lot of the stuff that the subject matter experts are really pleased that they don’t have to sit through, hours of drudgery about a topic that frankly they’re never ever going to care about or see or do anything with in the future, but which they absolutely need to get a handle on in order to write the content. So their real skill is different. The designer has to listen to the subject matter expert telling them all this, but the designer’s real skill is something else and they don’t need to know all this stuff from the subject matter experts, so they can use AI for that side of it now which is really helpful. But the expert in that loop is really, there are two experts in that loop, and the subject matter expert I find usually doesn’t want to be there any more than the instructional designer in that conversation because they’ve both got other things they would like to be doing and which they’re very good at. So maybe AI is just helping us to cut through some of the boring grunt work maybe in that respect. I certainly wouldn’t want to sit down with an AI coach, oh my goodness me, it would be awful. I mean it makes all sorts of things up for a start and I wouldn’t have a clue if it was going in the right direction or not. No, I think you’re safe there. I think your skills as a coach will be required for many many years to come, let’s hope so, yes let’s hope so.

(00:29:10):

I think although as you say you’re not in HR now, you are in the sense that you interview a lot of people on your HR Coffee Time podcast, which is doing extremely well. As I mentioned earlier at the beginning of this interview, actually is it top two percent? Do you want to tell us a little bit more about that because I’m in awe?

extremely well as I mentioned earlier at the beginning of this interview they

(00:29:33):

Oh gosh, thanks Carolyn. Yes, so my podcast HR Coffee Time is ranked in the top 2% of all 3.3 million podcasts in the world, which I still can’t quite believe. I was in complete shock when I found out. And it’s a podcast that focuses on professional and career development for HR professionals, and so it’s a combination of solo episodes from me talking about a particular topic that I think might be helpful, whether that’s confidence in speaking up in meetings or how to do well in an interview or what skills are going to help you step up into an HR director role as examples. And then I also have interview episodes which might be with experts who have got expertise I don’t have. So as an example I’ve got someone coming up who I’m interviewing about employee experience. I just interviewed someone about employee communications, there are too many words with employee. So they’re two topics I don’t have deep expertise in, but they do, and they’re helpful things for people who work in HR to be aware of and they’ll be able to share lots of tips and advice about how to do that well. So there are lots and lots of HR podcasts out there, more coming out all the time, and we’re really really lucky that there’s such great choice. When I started HR Coffee Time though I felt like there wasn’t really anything that was very career-focused and that looked at it in that way, so that’s why I decided to take that approach for that.

(00:31:06):

Yes, yeah, quite. I mean, that is, I wonder how much confidence, because you mentioned confidence there, and I wonder how much confidence is, or lack of it, or being a bit insecure about your confidence perhaps, is holding people back. I know it held me back when I was at a certain part of my career. People used to say to me, oh you’re ever so confident, but deep down I wasn’t. And sometimes that, I’m sure you must come across that, don’t you?

(00:31:32):

Oh, all the time. I’d say that’s the core thing and the longer that you’re coaching for I think when I first started as a coach I was really surprised when I realized just how many people, if not almost everybody, it’s confidence that they often want to work on. They just want to build their confidence, whether that’s in one particular area or confidence-building generally. So when I was mentioning the Inspiring HR Leadership Program earlier, one of the core focuses of it is on building confidence.

Yeah, and of course anybody just sort of fly on the wall would be so surprised I should imagine at the people that are feeling that way when really others might look at them and think oh I wish I was as confident as they are. Oh absolutely. People say that about me when I’m working with AI, you know, they sort of say well of course you know all about it and everything, but it’s so fast moving, Fay, you can’t stand still for two minutes. So I’m very much aware that I need to stay extremely vigilant and I certainly wouldn’t say that I was confident about it, but nevertheless I take great reassurance from the fact that nobody else knows either, and I think that’s one of the things, because it’s all moving so fast, isn’t it, and that’s one of the things that this podcast is about. It’s about understanding what’s changing in the way we work. I mean, if you looked back four years to your practice four years ago, or your daily work, just your work really because there’s a lot involved with running your own business, it’s not just doing the delivery, I mean I don’t know how much time you would allocate, 50% something like that to delivery, 50% to keeping the show on the road really, I don’t know, but when we look at that, how much has that changed due to AI entering into the picture? How has it changed for you? Can you tell me how that’s changed when you look back at that picture from four or five years ago compared to now?

Oh yeah, hugely. If I think I have an email newsletter that I send out every week that used to take me hours and hours and hours and hours of agonizing over getting the wording right, whereas now I’ll write the draft and put it in AI and say can you make this better, or I have a tendency to be quite waffly so I’ll be like can you please condense this down and improve it so it’s more engaging, or start with a better question. I mean that has just saved me hours every week. With almost all aspects of my work that aren’t the actual coaching part of it, it’s just been an absolutely amazing enormous help at just making me much faster at doing things or meaning that I don’t necessarily have to pay someone else to help me with stuff that I might have done before.

hugely if I think I have an email newsletter that I send out every week that used

(00:34:34):

Yes, I was going to ask what are the benefits of it because just talking with Tom McDowell a couple of days ago about the efficiency trap, it’s easy to think oh we all need to be more efficient, we all need to do things faster and quicker and everything else, but actually it’s a bit of a trap, it can be, for various reasons he was saying. But I think if we look at what we’re doing with the time that we save, what does it mean to you? Does it mean you save costs, does it mean that you can do more stuff that you want to do, or do you just have time off? I bet you don’t.

No, I probably should think about having more time off. I really do think the main thing is just the amount of time that it saves me, all those little things are really helpful, but also other aspects. So I started having my podcast on YouTube last year and that was really daunting because it had been audio only for years, and the idea of being on video and oh I don’t know, I definitely didn’t feel confident about that, I had to really work on my confidence. And for YouTube things like keywords really matter and your title really matters and your thumbnails, so that’s the picture that represents your video that people can choose whether to click on it or not, all those things really matter, and I knew nothing about YouTube. I wasn’t really a big YouTube user, but there are AI tools that will give recommendations for titles and thumbnails and keywords to be including, and I’ve just found that so helpful. Even YouTube’s got AI built into it now so I can think oh I think this is the title I want to use for the podcast, but I can do A-B testing with it and put in two different titles and then it will let me know which one is more popular and then that’s the one I can tackle on.

(00:36:28):

Yeah, well I mean it’s taken me years to get to that point. Yeah, but you know, you wouldn’t have had the time, would you, a few years ago, to do A-B testing on all this stuff. I mean even if you did you wouldn’t have had time to tweak it because that’s a marketing department type thing, wasn’t it? Yeah, yeah, yeah, that’s it. I suppose that’s where it helps me the most, with my marketing. I would say that’s definitely where I’m using it the most, and even choosing titles for my podcast, I mean I do still spend ages on it, but I used to spend hours agonizing over every title. There’s a tool that will analyze your titles for you and tell you a ranking as to how likely it is. It’s called, I’ve just forgotten what it’s called, I think it’s called Headline Analyzer or Headline Co-Schedule. I pay, it’s not very expensive, I pay an annual subscription for it, but you’ve got to think of the title to put in there in the first place and then it will give you recommendations on improving it. So now I can go on to ChatGPT and say this is my episode, this is the focus of it, what would be some compelling good titles to use, and I can then put them into the headline analyser tool and it just makes it start shaping up what the final title is going to be so much quicker than how I was doing it before.

(00:37:50):

Is it having an impact? Is it actually, are you seeing the results of it?

(00:38:00):

Well I know the impact it has on me really is just the time it’s taking because I will keep on going until the headline analyser tells me I have a strong title. So before I might be spending at least two hours on picking one title whereas now I might be able to do it in half an hour.

(00:38:21):

But you’re getting the same high score and you’re getting the same results with people clicking presumably because that’s the goal, isn’t it?

(00:38:27):

Yeah, hopefully.

(00:38:30):

I know they are clicking because you’ve got quite a high profile on LinkedIn now. So what I want to ask you before you go, because there’s been so much that we’ve covered, but I’m rushing to this because I’m hoping you will be okay with it, I’m wondering if you would give listeners some advice about how they can join the conversation about AI, not in terms of being, you know, just like the conversation we’re talking about now, just being a bit more confident to try stuff out. How would you advise people to be just a little bit more confident in themselves to start trying stuff and using it?

(00:39:12):

My advice really would be about asking people how they’re using it and asking them if they’ll show you. That’s what I found the most helpful thing. I have a friend who’s really great at using AI, absolutely loves it. She’s ahead of me. I know you said I had to rank myself and I said a seven, she’s probably at least an eight. And so she’ll often try stuff out before me and be all excited about it and talking about it, and I’ll say oh could you show me how you did it? And she’ll just share her screen with me and go so I went in and asked it this and then I did that, and I think as soon as someone shows me how to do something or they tell me how they’re using it, I think oh that’s how you do it, but I wouldn’t necessarily have had the idea on my own. And I know for other people, someone said to me once well could you just help me get the tone of this email right? And I said oh sure. So, you know, had a look at it and then I went onto ChatGPT and said how could we tighten this up or make this point clearer? And they went oh, that’s how you use it. And I was very lucky that I got to sit in on an HR team away day and they had brought in an internal IT person to talk everyone through how to use Copilot and when he showed us all just on the screen, he was a really really good trainer, really good, and I just thought oh my gosh, I didn’t realize it, I hadn’t even realized I’ve got Copilot. I do watch quite a few TikTok videos as well, there are loads of great little short videos which can be helpful, but what I find with those is they’re often geared towards, it’s about finding the right one, they might be recommending stuff they’re wildly excited about but you think I’m never going to use that in my work or I don’t really see why that’s such a big deal, so you do have to be intentional otherwise, because I mean we don’t have time to consume all this stuff do we? The irony of it all really, we’ve got less time. But I’ve definitely learnt the most from other people being generous enough to show me how they’re using it. So that would be what my recommendation is.

Yeah, people are flattered when you ask them. I mean, you know, they’re quite sort of helpful, I think. They say oh yeah, of course, watch this or let me help you or whatever. So it’s always a good thing to ask. People always love it because it shows you value what they can do and you value their expertise.

Sorry?

(00:41:19):

Yes, absolutely. Absolutely. Fay Wallis, it’s been a pleasure speaking with you. Thank you so much for opening up your office to the Sovereign Career Podcast. I appreciate it very much.

You’re so welcome. Thank you so much for having me on.

Not at all. We’ll see you again soon, I hope. Take care.

Get more from The Sovereign Career Hub in the Substack app

Available for iOS and Android

Get the app

Here’s the cheat sheet if you’d like to use it directly:

Intentional Ai Use Power Verbs Cheat Sheet

89.4KB ∙ PDF file

Download

Download

I’ll be sharing one of these each week. If you’d like them in your inbox, you’re very welcome to subscribe.

Subscribe now

Most professionals I talk to have made their peace with AI. They’ve found the tools that fit their week. They’re in the game.

And yet.

There’s a quiet dissatisfaction bubbling underneath the productivity gains. A sense that the output is fast but rarely finished, that the work leaving their desk doesn’t quite sound like them, that something has been gained but something has also, subtly, been lost.

That’s not an adoption problem. Adoption, it turns out, is the easy part.

It’s a direction problem. And almost no one is talking about it. Almost no one, that is, except Sharon Green and me, on Episode 1 of the Sovereign Career Podcast last week.

How work ends up in the ‘workslop’ pile

Here’s what I see constantly. A capable professional sits down to use AI. They type draft, create, generate, and the output comes back polished, complete, standalone. They tweak it. They ship it. They feel productive.

What they didn’t notice is that the tool just made a series of decisions on their behalf. It chose to produce a finished artefact rather than extend something already in motion. It chose tone. It chose structure. It chose what to foreground and what to drop.

None of those were neutral choices. They just weren’t the author’s.

This is how workslop gets made. Output that looks finished but carries no one’s judgement. It ships because it’s polished. It underwhelms because no one directed it.

And here’s the part that’s easy to miss: AI doesn’t wait for you to be intentional. It fills the space. Every decision you don’t make, it makes. Every verb you don’t choose, it chooses. The moment your attention lapses, even slightly, it steps in and authors the gap.

Which means there is no neutral position. You’re either directing the tool, or it’s directing the work that goes out under your name.

And the cumulative effect, over months and years, is a body of work that’s progressively less yours, but for which you remain entirely accountable.

What direction actually looks like

In the first Episode of the Sovereign Career Podcast, I get to chat with HR Interim, Sharon Green who describes herself as an intentional AI user. The phrase is easy to nod at and easy to underestimate, so it’s worth slowing down on what she actually means.

She uses AI for specific things, not everything. She treats it as a critical friend, not an authority. She triangulates across tools because hallucination is real and she’s not willing to put her name to something she hasn’t checked. She anonymises client data before it goes anywhere near a system. She avoids platforms whose ethics she doesn’t trust.

And she doesn’t use AI for writing. In fact, every post she publishes carries the same endearing sign-off: “AI was not used in the writing of this post. All mistakes are the author’s.”

Notice what’s not in her reasoning. She isn’t claiming AI can’t write. She isn’t comparing speeds, weighing trade-offs, or defending the choice on craft grounds. She kept ‘writing’ for herself because she enjoys it. Whether a tool could do it faster is, to her, beside the point.

That’s a more sovereign position than it first appears. It’s a professional refusing to let the efficiency conversation decide what she does with her own time and skill.

Five different practices, one underlying move: she decides what the tool is for before she opens it, and what it isn’t for, on grounds that are hers to set.

That’s direction. And once you see it, you can’t unsee how rare it is.

Subscribe now

A boundary is a declaration of ownership

Sharon’s writing boundary is the one I keep returning to, because it’s the one most likely to be misread.

It would be easy to file it under AI-sceptic or old-school. It’s neither. It’s a professional making a deliberate decision about what stays hers, on grounds the productivity conversation simply doesn’t have access to.

That’s not a weakness in her AI practice. It’s the strongest part of it.

A boundary, held for reasons you can name (even reasons as simple as I enjoy this), is not an admission you couldn’t keep up. It’s a declaration that you’re still the one running your career.

Which is the whole point. Intentional AI use isn’t a virtue in itself. It matters because it serves something larger: the career you’re actually building, the work you want your name on, the judgement you want to be known for. Direction without that frame is just better prompting. Direction with that frame is sovereignty.

Use it or lose it

If ‘direction’ sounds abstract, here’s the most concrete place that ‘direction’ shows up: the instruction word you use when you ask AI for help.

Never was the phrase, ‘use it or lose it’ more apt!

For example…

‘Draft’ tells the model to produce a complete, standalone artefact.

‘Extend’ tells it to continue something already in progress.

‘Pressure-test’ tells it to interrogate.

‘Sketch’ tells it to offer options rather than commit.

‘Refine’ tells it to improve what’s there without starting over.

Most professionals rotate through two or three verbs and wonder why the output keeps overshooting what they needed. The verb isn’t a stylistic choice. It’s the instruction that determines what kind of work the model does on your behalf, and, by extension, how much of the gap-filling it does for you.

Choose it deliberately, and you’re directing. Choose it by default, and the tool makes the choices - and that’s why it’s called ‘workslop’.

The professionals who’ll come out of this era with their authority intact won’t be the ones who used AI the most. They’ll be the ones who, at the end of any given task, can explain and stand behind every decision the work reflects, including the ones the tool would have made if they hadn’t been paying attention.

That’s not about prompts. It’s about whether your career is still being authored by you.

Leaving you with these two things…

I’ve put together this one-page cheat sheet, Intentional AI Use: Verb Signals, that maps the most common instruction words to what they optimise for, what they tend to produce, and what to watch for. It’s great if you’re serious about being more intentional. It’s free to use and sits next to you while you work. Download it here:

Intentional Ai Use Verb Signal Cheat Sheet

89.4KB ∙ PDF file

Download

Download

So as well as leaving you with the Cheat Sheet - I’d also like to leave you with this question to ponder…

‘which parts of your work are you currently letting the tool fill in, because you haven’t decided what you’d like to keep?’

You can tune into the full conversation with Sharon Green here on Substack or wherever you get your podcasts.

Episode 01 with Sharon Green

Thanks for reading as always, and if you find this article useful please share it with a friend as it may just be what they’re looking for!

See you for episode 02,

Share

Sharon Green is a professional HR interim, coach and consultant - and one of the quieter connectors in the independent HR world. She founded Chiara Consultancy after a career spanning senior people, project and management roles across the NHS, the not‑for‑profit sector, and international law firms. Alongside her client work, she runs a thriving peer community for independent people professionals - because, as she puts it, it gets lonely out there.

Long before AI became the story of the moment, Sharon was already watching waves of workplace technology arrive with bold promises, reshape things partially, and settle into something far more complicated than the hype suggested. Earlier in her career, she helped introduce one of the first online graduate recruitment tools at her organisation - and she’s been watching technology promises collide with workplace reality ever since.

Share

So when AI came along, Sharon didn’t rush to the front of the bus.

She rates herself a 3–4 out of 10 on ‘the AI spectrum’. Not because she’s behind - but because she’s an intentional AI user. She uses AI for specific things. She treats it as a critical friend rather than an authority. She triangulates across tools because she knows hallucination is real. She anonymises client data before it goes anywhere near a system. And she won’t touch platforms where she doesn’t trust the ethics or values of the people behind them.

There’s also one thing she doesn’t use AI for at all.

Writing.

Not because she can’t - but because writing is a craft. It’s something she loves. Something bound up with her professional identity. For Sharon, that isn’t a gap in AI literacy. It’s a boundary.

And that raises a question we don’t ask often enough:

What are you not willing to give to AI?

In this first episode of The Sovereign Career Podcast, we talk about what it actually looks like to engage with AI intentionally rather than enthusiastically - and why that quieter, more considered stance may be the wiser one for people professionals right now.

We also get into the messier, practical side of what Sharon sees in her client work: organisations keen to adopt AI in their people function before their processes are clear or their data is ready. Her view is simple - you can’t just drop a tool onto an unprepared system and hope it works. Foundations matter. Context matters. People need to be met where they actually are, not where leadership wishes they were.

Throughout the conversation, Sharon keeps returning to the role HR plays in shaping how AI shows up at work. Sometimes she sees AI treated as a tech project and handed over accordingly. But people, technology and experience are deeply interrelated and perhaps because HR is a broad church - AI adoption can be patchy. If HR isn’t part of the conversation early, something important tends to go missing.

This isn’t an episode about keeping up.
It’s about staying grounded.
About making conscious choices.
And about protecting the parts of your work - and yourself - that still need to stay human.

What comes up in this conversation

• Why Sharon rates herself a 3 to 4 out of 10 on the AI spectrum even though she is using AI regularly and comfortable with the tech.

• What being “a curious sceptic” looks like in real client work

• Why using AI well sometimes means slowing down, not speeding up

• The difference between ‘tech-licensing’ conversations and the conversations that actually matter

• Why poor data and unclear processes derail AI efforts in people teams

• The ethical lines Sharon won’t cross - and why drawing them matters

• The one part of her work she consciously protects from AI

• Her advice to HR interims wondering whether it’s worth engaging with AI right now

About Sharon Green

Sharon Green is a professional interim, qualified coach and consultant, and founder of Chiara Consultancy. She has held senior people, project and management roles across the NHS, the not‑for‑profit sector, and international law firms.

Sharon specialises in people change and transformation, people technology and people experience. Alongside her interim and consultancy work, she coaches clients at career crossroads — supporting professionals navigating transition, identity shifts and complex decisions about what comes next.

She holds a Masters in HRM, is a Chartered FCIPD, a certified change manager, an Agile® and PRINCE2® project manager, and an ICF‑trained coach.

As a #payitforward passion project, Sharon also runs a global LinkedIn peer community supporting over 3,000 independent people professionals, including interims, consultants, coaches, contractors and freelancers.

Connect with Sharon

• LinkedIn: https://www.linkedin.com/in/sharongreenchiara/

• Website: http://chiaraconsultancy.co.uk

• X (Twitter): https://x.com/SharonGChiara

• Bluesky: https://bsky.app/profile/sharongchiara.bsky.social

• Instagram: https://www.instagram.com/sharongchiara/

Subscribe now

Get more from The Sovereign Career Hub in the Substack app

Available for iOS and Android

Get the app

TRANSCRIPT

EP 01 The Sovereign Career Podcast

Use AI on your terms. Protect what’s yours.

Carolyn Shepherd in conversation with Sharon Green

Carolyn:
Well, hello and welcome to the Sovereign Career Podcast.
It’s the first one.
And this is my guest, Sharon Green — the lovely Sharon Green — who was there right at the beginning of my AI journey. And I’m sure we’ll come into that a little bit in a few minutes. But that was three years ago.
But we’ve only really bumped into each other occasionally live, haven’t we, Sharon?

Sharon:

Yes.

Carolyn:

Mainly it’s been an online kind of relationship, or friendship as well as colleague. I like to think of it as almost friendship now.
And because I came from an HR interim background myself, and I know that’s really why I was drawn to you and your amazing network that you’ve got, I wondered if you’d tell us a little bit about yourself and the network.

Sharon:
It’s interesting when you meet people in real life after you’ve connected online. I always love that.

I’m Sharon Green, and I would always say that I’m a professional interim. That’s what I do for a living. I used to have a permanent career and many years ago I stepped off to run my own business.

In that business I go into clients to add capacity or capability to their people teams. Usually, not always, but usually their people teams.

The three areas of work that I focus on are around people change and transformation. That could be running big projects or smaller projects where people want to make changes.

People change and transformation, probably over the last 15 years, usually involves technology. So my second area of work and specialism is people technology, which links very nicely to the AI conversation.

And then for me, the third area is people experience — how we learn, how we develop people, how we give them a great experience and put the human into work.

Those are the paid elements of my work.

My passion project, my side hustle, is running a community for people professionals who work independently. That could be coaches, consultants, interims, freelancers, solo entrepreneurs — whatever they want to call themselves.

It’s really about peer‑to‑peer support, networking, learning and sharing, because it can get a little bit lonely when you’re on your own.

Carolyn:
Can’t it? I mean, I know. And it was your WhatsApp group that I came into quite quickly, although I know it’s also on LinkedIn. It’s a vibrant group.

I joined it when I was still looking for work — this was three years ago — but the HR interim work dried up for me. It had to be 100% remote for me because of where I live. I think people looked on the map and thought, “That’s a long way away. What are we going to choose her for?”

A lot of other things happened as well. I just thought, like you, a lot of my working life has involved HR tech projects. So it was a natural extension for me.

I was fascinated anyway. I started experimenting with people at work and they were very willing to join in.

Before we go into more detail about AI and how you use it, can you paint a picture of how you work? Are you working from home at the moment?

Sharon:
It really depends on the client. I’ve done fully remote work, particularly with remote‑first startups and scale‑ups who maybe started without an office.

Some clients expect more in‑person work. When you’re an external person, it really helps to build a cultural picture to see people in their own environment and not just transact through screens.

This is my Chiara HQ — the centre of the business.

Carolyn:
Because meeting in person, a lot of people listening will be thinking they go into work some of the time — maybe a couple of days a week — and there’s a lot of value in that peer‑to‑peer contact. It’s social as well. I miss out on all that. I work here on my own quite a lot — just me and the dog, my German Shepherd.

I know you’re all about the real stuff, and I think that’s what people are drawn to. It’s certainly why I’m drawn to you. As much as I love AI, I love what I call the Zone Three stuff — the human stuff. Without that, something feels missing from the relationship.

Do you think that’s why CEOs — who often get criticised — want people back in the office? Or do you think it’s about control and suspicion?

Sharon:
I’ve always had the capacity to work remotely since I set up my own business. The acceptability of that from clients has varied over time.

For me, it’s similar to my approach to AI or technology — it has to be purposeful and intentional. It has to serve a use and a purpose.

There’s now more acceptability for remote‑first work, but there’s still an intentional aspect to being in an office or with people. Coming back to your question about CEOs and senior leadership teams, there are additional layers — scepticism, concern, and sometimes a darker side around control.

Managing a dispersed workforce can test leadership skills. There’s often more going on than what’s presented on the surface.

Carolyn:
I don’t work much in experience or culture myself, though I see it and hear about it. It’s not something I prioritise in my day‑to‑day work.

So here we are: you’re responding to client needs in HR transformation, HR technology and people experience. You’re working as a human‑first person. I know you often say AI wasn’t used in the making of your posts and all errors are your own.

Where do you sit on the AI spectrum?

Sharon:
I remember you asking me beforehand whether to give a label or a number. If I used a one‑to‑ten scale, I’d say I’m probably around three to four in how I think about AI and how it relates to my work.

That reflects how much I know, how much I use it, and how much there still is to learn. Lifelong learning has always been important to me.

In terms of a label, I’d describe myself as a curious sceptic. That’s quite a complex definition. I think about AI on a macro level — its impact on work, the technology marketplace, how challenging it is for clients to choose AI‑enabled tools, the sales hype around it.

AI feels like the latest stage on the technology bus. I’ve always been interested in technology. Early in my career I introduced online graduate recruitment tools and learning admin systems. I’ve always been an early adopter of things like LinkedIn.

AI feels like part of that journey, but I’m not first on the bus. It has to respond to a need. It has to have intent and purpose.

I’ve worked with a lot of tech and information security professionals, so I’ve learned there’s always more to consider beyond the surface level.

Carolyn:
I agree. This isn’t our first rodeo. I remember when the internet arrived and everyone predicted paperless offices. A few years later my boss said his paper bill was bigger than ever.

That said, I feel this time is different because of the pace. Previously you had an adoption curve — you had to drag people along. Now people are using AI at home and bringing that expectation into work. There’s social pressure when colleagues are producing better outputs faster.

That worries me, because employers seem to be investing in training only at a tools‑use level. They’re not focusing on understanding, ethics or guardrails. I suspect we’ll see disappointment later when competitive advantage doesn’t materialise.

How do you see AI being embedded into workplaces?

Carolyn Shepherd in conversation with Sharon Green

Sharon:
I was on a conversation last week where we were talking about the use of AI tools within a certain part of HR — looking at development, particularly around assessment and development — and having a kind of Chatham House Rules conversation about how people were using it, what considerations they have, what tools they’re looking at.

I felt privileged to be in that conversation because that’s not something that has come up in some of the clients I’ve worked with. What tends to happen is you’re starting with where they’re at.

A recent client was talking about wanting to use more AI in their people function. I said I get that’s the goal, but we’ve got to start with where you’re at, which is: at the moment we need to clarify processes, consolidate those processes, and make sure all of the data that your AI tool of choice will feed off is accurate, up to date, in the right place, so that it can work within the system that you’ve created in your organisation.

You can’t just plonk a tool on and hope that’s going to work. There are some foundations that you need in place and you’re not there yet.

Sometimes clients are further on and they’re looking at experimentation. They’ve got a particular problem they want to solve. They want to go out into the market to work out what tools they can have — whether it’s within their Microsoft suite, for example, or whether it’s a tool of choice — what’s in their tech stack to maximise, or whether they’ve got to go out to market and create a brand new tool.

There are people and organisations at varying stages of that journey. You can already see by the thought leadership papers that get published, particularly by the big consultancy companies who are investing money in that, that they’ll be at a different stage to your startup scale‑up, or in my experience, private banks or regulated businesses, which will be totally different.

They’re all going to be on a different journey in some senses.

Some of the other aspects make sense to me. I do people change, I do people tech, I do people experience — they are interrelated and connected. Sometimes clients won’t see that, or they’re starting with the technology but they’re not thinking about the experience or the change impact of that. Some of that adoption curve might be less for some people and more for others in your workplace.

On the surface, it might look like: let’s just use Copilot, how many licences do we have, who gets the fancy licences, who gets the basic. Do we use ChatGPT or Gemini or Claude or Perplexity or whatever other tools are out there.

Once you get into it, it feels like a potential for a richer conversation if you’re willing to have it. Some people will be at the surface level of tools. That’s where it’s our job as people professionals, or whatever we’re coming at it with, to start where they’re at and maybe add some additional insights from our experience.

Carolyn:
So do you think then that HR is doing a good job in terms of dealing with this age of AI? Or do you think on the whole it’s standing back a little bit and saying, well, look at all the problems, guardrails and all that stuff we need to think about?

I wouldn’t be on the evangelist side of AI, nor would I be on the other end of the spectrum. It’s there, it’s coming. The rapidity of it — the fast‑pacedness of it and the plethora of it — makes it unescapable, so we’ve got to be in the conversation.

Are we as people professionals always in the front of the conversation around technology? Probably not. There’ll be a lot of organisations defaulting to the tech team and thinking about it as a tech project.

Carolyn:
Is that a mistake, do you think?

Sharon:
It’s usually a mistake to think of it through one lens.

I did a great project where I was the people person in a tech function. They were doing an agile transformation change project and I was in the tech function. It was great — I learned an awful lot about technology. I felt I was pretty tech‑savvy before I got there, but I learned a lot.

Carolyn:
Because you’ve got your agile background as well, haven’t you?

Sharon:
Exactly. So I could talk their language and get their trust.

I do hear people saying HR is looking to tech to lead it or saying it’s really about the licences and things. If I compare different types of clients — sometimes I’m working with HR, sometimes I’m not — HR seems to be much further behind the general uptake of AI. They want to know how it can work for them, they probably want a quick fix, but I’m not sure they really want to get stuck in.

Carolyn:
Is that because of the sort of people it attracts, or because it’s fragmented into two levels — senior leadership and operational — and there’s been a fragmentation? There’s a lot in that question, but what’s the flavour you’re getting?

Sharon:
It’s part of my business and how I get work to be in that space. The projects I’ve been involved in have meant I’ve had exposure over a number of years to the technology side of the people function — interest in data, interest in systems and how they work.

I remember a recruiter telling me once: “Why are you not advertising this as part of your service, Sharon? Because I don’t see many HR professionals who are that tech‑focused, that tech‑savvy.”

On one level, it’s a broad church, the people function. You’ll get some people for whom that stereotype of “we don’t do this” or “we’re not great at tech” is there. That helps me get work.

However, I do think there’s too broad an agenda of what the people function is doing, and there needs to be a consolidation so we can work out what the focus is.

Even if you’re not coming at it from a technical perspective, it’s about stepping into conversations where we might not know the answers. You’ve got to be in the conversations across the organisation. This is an organisation‑wide matter.

Coming at it from pure tech, or being in the people team and only getting in touch with the tech team when you need something from them — neither is the right approach. Collaboration is better. Co‑creation.

Carolyn:
It’s very agile, isn’t it? Collaboration, co‑creation — it’s a good thing. You get buy‑in and all the other things you need to get projects like this off the ground.

It’s been absolutely fascinating talking with you. It has opened up so many thoughts for me. I’ve made some notes as we’ve gone along. We could go into this again and again — tech stacks, redesigning the HR function — fascinating topics.

Before we go, what would your advice be to an HR interim who’s thinking: should I spend time getting to know more about AI? Is it worth it?

Sharon:
Definitely worth it. It’s not going anywhere. We’re not quite sure what impact it’s going to have on the world of work or our individual workloads, but unless you get curious about it, experiment with it, have a point of view and an understanding of it, we can’t make informed choices.

When I use AI, it’s intentional. It’s about what I need it to do. You mentioned I don’t use it for my writing because I like writing — it’s a craft.

People feel they should be using it for everything. There’s pressure that they’re going to take the joyful bits away — the bits they enjoy doing. Use it intentionally.

The way I use it is for specific things that I need to do quickly. It’s like a critical friend sometimes.

I would also say: think about your values, your ethics, which tools you want to use. Pick a few tools. Triangulate because of hallucination.

Carolyn:
Would you suggest using different types of AI assistants? I heard you mention four or five different assistants. Would you suggest experimenting with ChatGPT plus a Perplexity, for example — with permission, obviously — to get a feel, because they behave differently?

Sharon:
Not necessarily without permission, but doing that so you get the feel, because they do different things.

The key thing for me is whether you’re using them internally or externally. I don’t put my own IP in there. I don’t put client data in there. I anonymise things if I use it. I use it as a baseline. I’m critical of the information I get from the tool. I ask it for references. Good prompting to find good information.

There are some tools I don’t touch because I don’t like the ethics of the founders. Other people might not have that point of view, but that’s mine. That’s about my values.

It is about being open, but also listening to people who are critical of it and concerned about it, and striking your own point of view. Getting your own point of view is really critical.

Carolyn:
Thank you very much indeed, Sharon. I’ve really enjoyed this talk with you.

The message I’m taking away is that you value intentional use of AI, and you advocate being well informed. It doesn’t matter where you are personally on the spectrum. Join the conversation, be part of things, and then you can decide.

I love that. I’ll vote for you at the next election. Great to speak to you again soon. Take care.

Before you go

The Sovereign Career Podcast is a place for thoughtful conversations about work, judgement and identity at a time when roles, tools and expectations are shifting fast. It isn’t about keeping up or having definitive answers. It’s about noticing what’s changing, making conscious choices, and staying human as you navigate it. If this episode resonated, you’re very welcome here.

END

The career ladder is being dismantled. Not by corporate restructuring, but by the quiet competence of AI agents doing work that used to belong to people. And almost nobody in leadership is being honest about what this means.

The implicit deal was simple. Start at the bottom. Do the unglamorous work: the screening, the drafting, the scheduling, the research. Prove yourself. Move up. But those foundational tasks were never just busywork. They were the proving grounds where you learned the craft, built relationships, developed judgement, and got noticed. They were how mentors spotted potential, not through competency frameworks, but by watching someone handle the mundane with care.

Harvard researchers tracking 62 million workers across 285,000 US firms found junior positions shrinking at companies integrating AI. LinkedIn’s own data shows entry-level hiring rates down 23% compared to pre-pandemic levels, a steeper fall than the 18% decline in overall hiring.

So if the entry-level work disappears, how does anyone become senior?

The technology-optimist answer is one I’ve made myself - that AI frees people for higher-value work. It isn’t wrong, but it’s dangerously incomplete. When we automate the tasks, we don’t just remove work. We remove the context in which professional identity was formed.

What Jobscaping Gets Right, and Where It Falls Short

Jobscaping asks professionals to stop thinking of careers as ladders and start treating them as landscapes. I coined the term ‘Jobscaping™’ when I started thinking about how we as individuals can navigate the evolving landscape of work in the age of AI.

You scan your terrain continuously: what to prune (delegate to AI), what to plant (new capabilities), what to protect (distinctly human strengths). It’s grounded in three pillars (agency, ethics, and the practical discipline of reshaping your role) and it gives individuals a structured way to respond when the ground shifts.

This matters because the institutional responses, while necessary, are not arriving fast enough. Researchers and policymakers are converging on solutions. Bruno Manno, writing in the Stanford Social Innovation Review, argues that employers should restructure entry-level jobs so they teach as well as produce, with apprenticeship models providing the template. The Washington Monthly has made the case for treating the rebuilding of career entry points as civic infrastructure. Brookings and Opportunity@Work have mapped how nearly 11 million workers without degrees sit in “gateway” occupations highly exposed to AI, with almost half the pathways from those roles to higher-paying work now at risk.

This is vital work. But it operates at the level of policy, institutions, and systems. And systems move slowly.

Here is where Jobscaping claims its distinctive ground. Almost all existing research focuses on either the macro-economic picture or the organisational response. Very little addresses what the individual professional should actually do while waiting for institutions to catch up. That is the gap Jobscaping™ fills: a practical, ethically grounded framework for the person standing in the landscape right now, today, asking what to do next.

It is not a replacement for structural change. But it is the only framework I’m aware of that gives an individual professional a coherent method for navigating the transition in real time, grounded not just in productivity but in personal agency and conscience.

That said, here’s the polemic: individual agency cannot be the whole answer.

Jobscaping™ helps one person navigate the landscape. It doesn’t rebuild the paths for an entire profession. It doesn’t answer how a 22-year-old develops judgement that used to come from years of hands-on work. And it becomes complicit if it lets organisations pretend career development is now purely the employee’s problem.

If Jobscaping is going to be more than a personal coping mechanism, it needs to extend into the structural. David Autor and James Manyika at MIT have emphasised that whether technology is deployed to automate or augment human labour is ultimately a matter of choice, with many different futures possible at this stage. That framing sits at the very heart of Jobscaping’s philosophy: this is a choice, not an inevitability, and professionals have the right and the responsibility to make it consciously.

Six Moves Toward a Practical Response

1. Map where the rungs have actually gone. In your field, which entry and mid-level tasks are AI absorbing right now? Not in theory. Right now. Categorise honestly: fully automatable, partially automatable, stubbornly human.

2. Redefine what progression means. If vertical climbing is no longer the primary mode, what replaces it? Deepening expertise, expanding relational networks, developing ethical judgement, learning to orchestrate AI tools. The World Economic Forum has argued that early-career talent is becoming more, not less, critical in an AI-first world because digital natives ramp up quickly with AI tools. But this requires organisations to recognise and reward lateral growth, which most currently don’t.

3. Identify the new proving grounds. Some firms are already experimenting. Agency leaders have redefined junior roles so that young professionals handle strategic and conceptual work while AI drafts first versions, effectively flipping the old model. The emerging foundational competencies include evaluating AI output, navigating ambiguity, facilitating trust-building conversations, and designing the oversight systems that make AI useful rather than dangerous. These aren’t lesser skills. But they require entirely different training approaches.

4. Confront the organisational obligation. If an organisation automates the developmental infrastructure (the tasks that trained people, the roles that built judgement) it has an obligation to build something in its place.

5. Build practical tools. This is where Jobscaping becomes actionable. Self-assessments, reflection templates, regular check-in questions: What skills am I building that AI cannot replicate? Where am I creating visibility for my human contribution? Am I developing judgement, or outsourcing my learning along with my tasks?

Jobscaping is designed to provide the scaffold for becoming the former rather than the latter.

6. Test it in public. Nobody has this figured out. Share how you’re navigating the shift. Every professional who does contributes a data point. Enough data points and we start to see patterns.

The Real Stakes

The career ladder, for all its flaws, was a system of hope. It told people that effort would be rewarded and grunt work was going somewhere.

If we dissolve that system without building something in its place, we don’t just create a skills gap. We create a meaning gap. We tell a generation of professionals that the entry point to their field no longer exists, and offer nothing but vague reassurances about higher-value work that nobody has defined.

‘Jobscaping™’ is an honest attempt to fill the gap of entry point to mid career: it’s based on a practice of continuous, ethically grounded career shaping that puts the professional back in control. The research confirms the problem is real and accelerating. The policy responses are emerging but slow. What has been missing is a framework for the individual who cannot wait for institutions to act first.

That is what Jobscaping offers. But it needs to grow beyond individual agency into collective action. It needs to demand that organisations do their part. And it needs professionals who are willing to share what they’re learning as they go.

Are the traditional career rungs disappearing in your field? How are you navigating it? Hit reply. I’d genuinely like to know.

If this resonated, subscribe to The Sovereign Career Podcast for conversations with professionals who are actively Jobscaping their way through the AI transition.

Sources

1. Brynjolfsson, Chandar & Chen, “Canaries in the Coal Mine? Six Facts about the Recent Employment Effects of Artificial Intelligence” (2025) — cited in Economic Innovation Group analysis

2. Hosseini Maasoum & Lichtinger, Harvard University study on AI and junior hiring (2026) — cited in St. John’s University overview

3. LinkedIn entry-level hiring data (July 2025) — cited in Economic Innovation Group, “Looking for the Ladder”

4. SHRM, “AI’s Disruption of Entry-Level Work: What HR Leaders Need to Know” (November 2024)

5. Ahuja, S., “AI Is Reshaping Career Ladders”, Fast Company (March 2026)

6. Manno, B.V., “How AI Is Changing Entry-Level Jobs”, Stanford Social Innovation Review (October 2025)

7. Manno, B.V., “Making Apprenticeships Part of Civic Infrastructure”, Washington Monthly (April 2026)

8. Heck, J. & Muro, M., “How AI May Reshape Career Pathways to Better Jobs”, Brookings Institution & Opportunity@Work (April 2026)

9. World Economic Forum, “How AI Is Changing the Nature of Entry-Level Work” (March 2026)

10. SAP SuccessFactors & Wakefield Research, “AI Is Causing Entry-Level Roles to Evolve, Not Vanish” (April 2026)

11. Kim, K., Mester, N. & Sun, G., “AI and the Labor Market: A Worker’s-Eye View”, Washington University in St. Louis (2026) — reported in STL Today

12. Autor, D. & Manyika, J., on automation as choice (2025) — cited in Economic Innovation Group analysis

Subscribe now

Every day at 4:30pm, before I close my laptop, something happens automatically.

Claude Cowork opens my Outlook calendar, looks at what’s on for tomorrow, checks my inbox for anything that needs a reply, and drops a clean summary into a file I can open with one click.

No logging in. No scrolling. No forgetting to prep for the 9.00 am call I completely missed.

It looks something like this excerpt:

TOMORROW’S APPOINTMENTS — Monday 30 March 2026

1. 9:00 AM – 9:30 AM | [Redacted for Substack] Prep: Content for the 2-day workshop is in your inbox under Inbox > Speaking Events > [Confidential]. Look for emails from [Redacted for Substack].com

2. 9:45 AM – 11:00 AM | [Details redacted] Suggested prep: ERA updates for P&P

   • What is the purpose of this call? Note your key ask or outcome

   • Any background reading or previous email thread to review?

   • What do you want [name redacted] to walk away knowing or doing?

3. 12:45 PM – 1:45 PM | [Details redacted] (60-min booking) Prep: Online meeting — you are the organiser. Join link in invite. Suggested prep:

   • What prompted [Details redacted] to book this into your online diary? Review any prior exchanges

   • What’s your agenda for the 60 minutes?

   • Any materials, proposals or links to share in advance?

EMAILS NEEDING A REPLY — Sunday 29 March 2026

No emails requiring a reply today.

That last bit - the “suggested prep” headings - is something I asked Claude to add whenever an event doesn’t have notes already written in it. Instead of just saying “None noted” and leaving me staring into the void, it nudges me with the right questions to get started.

How I set it up

I use a desktop app called Claude Cowork - it’s made by Anthropic, the same company behind Claude. Think of it like having Claude available on your computer as a proper assistant, not just a chatbot you visit in a browser tab.

See bonus videos for help setting up Claude Cowork.

Inside Cowork, there’s a Scheduled Tasks feature. You describe what you want Claude to do and when, and it runs automatically in the background. Watch the video for an on screen walkthrough.

Here’s what I told it to do, in plain English:

Every day at 4:30pm, open my Outlook calendar, look at tomorrow’s appointments, and note any prep that’s already written down. If there’s nothing written, suggest some preparation headings to get me started. Then check my inbox for any emails that arrived today and need a reply - newsletters, notifications and automated emails don’t count. Save the whole thing as a text file I can open.

That’s genuinely it. I didn’t write any code. I didn’t configure any APIs. I just described what I wanted in conversational language, told it when to run, and it figured out the rest.

TBH there were some set up glitches and the whole thing took me 3 hours, but I overcame the challenges and it was well worth it because not only will it save me time, it will also make me better prepared rather than besieged throughout the day by dozens of random Alexa reminders!

The one setup step that does matter: Claude needs to be able to see your Outlook. The way that works is you have a Chrome extension installed (it’s called Claude in Chrome - free from the Chrome Web Store) and you stay signed into Outlook in your browser. (Just don’t sign out deliberately). When the task runs at the assigned time, Claude opens a tab in your already-signed-in Chrome, reads what it needs, and closes it. Your login details never pass through Claude, it just uses the session that’s already open, the same way you would if you opened the tab yourself.

Why this matters more than it sounds

I’ve been thinking about AI productivity tools for a while now, and the honest truth is that most of them require you to remember to use them.

You have to open the app. You have to ask the question. You have to remember it exists.

A scheduled task flips that. It runs whether you think about it or not. It’s the difference between a tool and a habit, except you don’t have to build the habit. It just happens.

For me, the 4:30pm briefing has become the thing that closes out my working day. I glance at the file, I know what’s coming tomorrow, and I close the laptop with my head clear.

That’s worth a lot.

What you need to get started:

• The Claude desktop app (Cowork is inside it — look for Scheduled Tasks in the sidebar)

• The Claude in Chrome extension (free, from the Chrome Web Store)

• Chrome open and Outlook signed in at the time your task is set to run — you don’t need to be sitting at your desk, just don’t close the laptop

The Claude subscription starts at $20/month approx £17 including VAT. You can be set up and running your first scheduled task in under 15 minutes.

One more thing worth knowing: you don’t need to be sitting at your computer when the task runs. You can be working in another browser tab and it won’t interrupt you Claude opens its own tab, does its work in Outlook, and closes it quietly in the background. The only thing to avoid is being inside Outlook web yourself at exactly the time the task fires up, as you’d both be in there at the same time and it could get confusing. Otherwise, carry on with whatever you’re doing.

I’ve shared the exact instructions I used below and put a spotlight on risk management walking you step by step through the details, security measures and prudent approaches.

Read more

Before we start - a few words of encouragement.

I’ve spent the last three years helping more than 20,000 HR professionals get to grips with generative AI - through my LinkedIn Learning course, through Emmeline.AI, and now here. And in that time I’ve worked with practitioners at every level - from those who’ve never typed a prompt in their lives to those already building automated workflows on their desktops.

This guide is for all Indie HRs.

If you’re new to AI and feeling a little anxious about what’s ahead - that’s completely normal. We start simple. Just you, a laptop, and Claude’s chat window. Nothing to install. Nothing to configure. Just a conversation. And if the Beginner level is as far as you go today, that’s enough - you’ll still save hours and serve your clients better.

If you’re already comfortable with AI and want to go deeper - it’s in there. The Intermediate and Advanced levels cover Claude Cowork and Claude Code workflows that will genuinely stretch what you thought was possible. Paid subscribers get all three levels in full every week.

And if you’re genuinely unsure where to start - I’d suggest beginning with Claude Chat. Get into the swing of it first. The other levels will still be here when you’re ready. 👑

Wherever you’re starting from - you’re in the right place. Let’s go.

What’s in this Edition

In this guide I’m going to show you exactly how to use Claude to transform one of the most common spreadsheet tasks in independent HR practice - employee data and workforce tracking.

For independent HR practitioners, a well-built employee tracker is one of your most valuable client assets. It’s the single source of truth for everything that matters - who’s employed, on what terms, when their right-to-work expires, what they’re paid, and when they started. Small business clients rarely have an HRIS, which means they’re relying on you to build and maintain this. Get it right and you become indispensable. Get it wrong - or let it fall behind - and the consequences for your client can be serious.

The problem is that building and maintaining these trackers from scratch, for every client, is genuinely time-consuming. AI changes that completely.

A quick guide to the three levels - and what each one actually does for you:

Each level does more than the last - but they tackle the work differently. Here’s the summary so you can navigate straight to yours:

👑 Beginner - Claude Chat builds your tracker structure and helps you prepare and clean incoming data. You still input the data yourself - but Claude makes every part of the process faster.

👑👑 Intermediate - Claude Cowork goes further. Once you have the data ready, Claude Cowork reads your actual files, inputs the data, updates records, and saves changes directly on your desktop. Critically, your files stay on your local computer throughout - nothing is uploaded to the cloud or shared externally. You handle the client relationship and data collection. Claude handles the rest.

👑👑👑 Advanced - Claude Code automates the entire pipeline. Data arrives, Claude processes it, the tracker updates, and you receive a summary - without you touching a thing. This level requires a dedicated secure environment - the Mac mini is intended to be exactly that, a safe space for your AI work. More on that in a future post.

Jump to your level. Or read all three to see where you’re headed. And if you’re genuinely unsure where to start - I’d suggest beginning with Claude Chat. Get into the swing of it first. The other levels will still be here when you’re ready.

One thing worth knowing straight away: one Claude Pro subscription covers all three levels. At approximately £16 per month (at time of writing), Claude Pro gives you access to Claude Chat, Claude Cowork and Claude Code - all in one desktop app. You don't need three separate subscriptions. You start at Beginner, and when you're ready to go further, everything you need is already there waiting for you. 👑

Subscribe now

Sovereign Workflow #1 -
Employee Data & Workforce Tracking

The problem: Small business clients rarely have an HRIS. So you end up building - and maintaining - employee trackers from scratch. Headcounts, contract types, start dates, salary bandings, right-to-work status. It’s fiddly, time-consuming, and one wrong formula breaks everything.

👑 Beginner - Claude Chat

What Claude does at this level: Claude builds you the complete structure of your tracker - column headers, data validation, conditional formatting formulas, and layout. It also helps you at every stage of the data collection and preparation process. It will draft your data request email to the client. It will create a reusable intake form. And if the client sends you information in a messy format - a Word document, a series of emails, a photo of a spreadsheet - you paste it into Claude and ask it to clean and reformat it ready to drop into your tracker.

What Claude does not do at this level is input the data itself. That job is still yours. But with Claude handling the structure and the preparation, what used to take a day can take an hour.

And let’s be honest - that data inputting is no small task. It can take hours. And before you can even start, you’re waiting on the client to pull the information together. Which they delay. And delay again. And sometimes you end up going in and doing it for them.

What you do: Go to claude.ai. Type your request in plain English. Copy the result into your spreadsheet. No installation. No setup. Just a conversation.

Try this prompt:

“I manage HR for a 25-person business. Build me an employee tracker with columns for: full name, job title, contract type (permanent/fixed term/contractor), start date, salary band, and right-to-work expiry date. Include a formula that flags any right-to-work expiry dates within the next 60 days in red. Format it so I can use it in Google Sheets or Excel.”

What you’ll get: A complete tracker layout with all columns, data validation suggestions, and a working conditional formatting formula - ready to copy straight into your spreadsheet.

See it in Action

Watch me run this exact prompt in Claude and see what comes back in real time. It takes a couple of minutes.

Download the template:

Click here to download Employee Tracker

I’ve also attached the finished employee tracker as a ready-made download. Use it as is, or adapt it for your clients. It’s yours - no strings attached.

Paid subscribers get a downloadable template with every future guide. This one is free for everyone as a thank you for being here at the start. 👑

Then iterate: Ask Claude to add columns, change the formatting, add a summary tab, or create a dashboard. Each request builds on the last. You’re in a conversation, not filling in a form.

Paid subscribers - here's what's waiting for you below:

👑👑 Intermediate - Claude Cowork - how to point Claude directly at your existing client files so it reads, updates, and saves your tracker without you touching a formula. Your data stays on your local computer throughout.

👑👑👑 Advanced - Claude Code - how to build a fully automated pipeline on your Mac mini that ingests new starter data, updates the tracker, and sends you a weekly summary email every Monday morning. Without you touching a thing.

Plus the full explanation of why the Mac mini is intended to be a safe space for your AI work - and what that means for your clients' data.

This is where the free preview ends. 👑

Paid subscribers - the Intermediate and Advanced levels are below, showing you how to take this further with Claude Cowork and Claude Code.

Read more

Today I unpacked my Mac mini.

I picked up the parcel from the Amazon driver, walked through to my kitchen, cut it open with scissors, and kept the boxes - just in case I needed to send it back. Because I genuinely wasn’t sure I knew what I was doing.

I carried it through to my snug. My home office. The room I do my best thinking in.

I cleared a space, set everything out, and started joining it all up with wires.

And then I tried to plug in the keyboard.

It didn’t fit.

Something about USB-A versus USB-C. I didn’t know which adapter I needed. I didn’t even know what I was looking at.

I had to ask an AI assistant to explain it to me - standing in my own office, holding a cable, feeling like I’d signed up for something much bigger than I’d bargained for.

And I had. I really had. I couldn’t even get the keyboard connected!

Here’s what I want you to know.

I’m number 12 on HR Magazine’s Most Influential Thinkers list. I’m on the AI 100 UK Leaders list. Over 20,000 HR professionals have completed my LinkedIn Learning course on generative AI. I founded Emmeline.AI. I’ve spent years helping business leaders understand and harness AI and yet, I still have one of those drawers full of wires and electrical thingies that never fit anything new!

I tell you this because you need to know it’s new for both of us.

Not just for you. For me too.

And then something shifted.

I was sitting there, surrounded by cables and boxes, when my phone pinged. It was Ruben Hassid - a Substacker I’d reached out to replying to a question I’d asked him about artefacts.

I’d asked because I genuinely didn’t know what the word meant in this context. I’m a qualified member of the Agile HR Community. I know ‘artefacts’ in that sense. But in Claude? In this workflow? I wasn’t sure.

And then I looked at what I’d actually built - with Claude’s help - to answer my own question.

A fully interactive infographic. Explaining artefacts. With working examples. In my brand colours. With a paywall. Ready for Substack!

I’d built it while figuring out what it was.

That’s when it hit me. I knew more than I thought I did. I just needed to keep going long enough to find out.

Subscribe now

I want to share that feeling with you.

Not the polished version. Not the “here’s what I’ve mastered” version. The real one - the cables-everywhere, wrong-adapter, what-even-is-a-USB-C version.

Because I think that’s what you actually need from me.

You already know I know HR. You don’t need me to prove that again. What you need is someone to walk this road with you - honestly - and show you that the confusion doesn’t mean you’re failing. It means you’re starting.

We can navigate it together…

• The difference between Claude Chat, Claude Cowork and Claude Code.

• What an artefact actually is in this context.

• How to set up a Mac mini to get stuff done while we sleep.

• What the security implications are for independent HR practitioners handling client data and can we trust AI agents.

• How to build a Substack that helps me earn a living. A girl’s got to eat!

I’m working all of this out. Right now. In real time.

So that’s what The Sovereign Career Hub is going to be.

Not me two chapters ahead, talking down to you.

Me, one step ahead at most - sometimes alongside you - figuring it out and sharing everything I learn as I learn it.

The AI tools. The workflows. The prompts. The mistakes. The moments when the penny drops and you think ‘Oh. Oh, I get it now!’

Those moments are coming for you. I promise.

But you have to keep reading the sentences that don’t make sense yet. You have to keep going when you don’t know which adapter to buy.

The penny will drop.

And when it does, you will be so glad you kept at it.

Subscribe now

So…

A very warm welcome to The Sovereign Career Hub. I’m Carolyn. And I’m figuring this out right alongside you. 👑

If this resonated - please share it with one HR colleague who needs to hear it today.

And if you’re ready to go deeper, the full AI workflow guides - broken down for Beginner, Intermediate and Advanced - are waiting for paid subscribers below.

And here is where the trap opens.

Because AI makes the administrative response to legislative change genuinely effortless. You can have a compliant SSP clause drafted, reviewed against ACAS guidance and formatted for your client’s handbook before the coffee goes cold. That is a real capability. It is also, if it becomes your primary value proposition, a quiet form of professional self-erasure.

The client who watches you update their handbook at speed learns one thing: that you are very good at updating handbooks at speed. That is not the lesson you want them to take from April 2026.

The independent HR professional who uses this legislative moment differently ~ who moves through the policy update and into the conversation about what these changes mean for how their client hires, onboards, exits and manages risk ~ is operating in an entirely different professional register.

This is the Delegate pillar of Jobscaping™ made visible in real time. The documentation belongs in Zone 1. Hand it to the AI. The architectural conversation ~ the one about systemic risk, onboarding design, and what Day One rights mean for this specific business with this specific workforce ~ belongs in Zone 3. That is yours.

The legislation does not distinguish between the two. Your clients will.

What matters is not how quickly you can process the change. It is what you do with the professional capacity that processing it no longer requires.

That is the Sovereign Practitioner’s advantage in April 2026. Not speed. Relevance.

The Expert’s Collaboration

To demonstrate this shift, I recently applied a Sovereign Workflow to the definitive ERA 2025 HR Matrix produced by Caroline Daniel Owner/Director of Clear HR in Sussex.

This is the matrix that sovereign HRs integrate into an AI workflow to reduce the admin tax and position themselves as architects rather than doers.

Subscribe now

By treating AI as a Junior Researcher, we didn’t just “note” the priorities we interrogated them for systemic failure points. We moved from simply acknowledging a rule change to securing a client’s entire onboarding architecture in a fraction of the time. This is how we liberate expert judgment from administrative noise.

🎙️ Early April: The Podcast Launch on Spotify

I am thrilled to announce that my new podcast launches on Spotify in early April. My first guest will be the architect of this legislative plan, Caroline Daniel. We will be looking under the bonnet of the Clear HR tech stack to see whether Caroline automates the mundane to protect her expert focus and her plans for leveraging AI in 2026.

If you would like to see this implementation plan in action, and witness the “Sovereign Magic” when practical AI is applied to expert resources, subscribe to this Substack and I’ll keep you updated.

Subscribe now

In the meantime, you can download Caroline’s plan directly from the following landing page: ERA2025 hr IMPLEMENTATION PLAN for sme’s

About this Sovereign Workflow

This is a practical component of Jobscaping™ ~ the framework for HR professionals to design a career that balances high-level impact with true operational autonomy. Sovereign Workflows integrate specific AI toolkits to push back admin creep, ensuring you remain the architect of your time, and your expertise, not its administrator.

What are you doing to position yourself as a Sovereign HR Architect?

Best wishes

Carolyn Shepherd, Sovereign Architect

As we navigate 2026, the pressure to “use AI” to keep up with the pace of work is immense. But for the HR independent, a poorly chosen “AI hack” isn’t just a shortcut; it’s a professional liability.

The Illusion of the “Free” Assistant

We’ve all seen the breathless LinkedIn posts promising to “automate your HR consultancy in an afternoon.” But as an independent, you don’t have a corporate IT department to vet the fine print.

If you are feeding sensitive client meeting notes or employee data into a standard, consumer-grade LLM, you aren’t just being efficient; you’re potentially leaking your client’s intellectual property into a public training set. In our profession, that isn’t just a “tech glitch.” It’s a breach of the fundamental trust that allows us to operate.

Sovereignty via the “Digital Vault”

To maintain Human Sovereignty, we must change how we view our tools. I advise my peers to stop looking for “hacks” and start building a Digital Vault.

A Digital Vault approach to AI means:

• Zero-Training Environments: Ensuring the data you input stays within your silo and is never used to train global models.

• The Junior Researcher Protocol: Treating AI as a highly capable, yet walled-off assistant that processes data under your strict supervision.

• The Architecture of Privacy: Setting up workflows where the AI helps you find the patterns, but you remain the sole gatekeeper of the final output.

The Efficiency is Digital; the Authority is Yours

Reclaiming your agency doesn’t mean avoiding technology; it means mastering it so thoroughly that it becomes invisible to the client, yet impenetrable to risk.

When you use a Sovereign Workflow, you aren’t offloading your brain to a machine. You are using a secure stack to strip away the admin creep—the formatting, the initial data sorting, the drafting of standard clauses—so you have the headspace to provide the nuanced, human judgment your clients actually pay for.

In 2026, the most successful HR independents won’t be the ones with the most “tools.” They will be the ones who have protected their credibility by ensuring that while the efficiency is digital, the authority remains entirely theirs.

About Sovereign Workflows

This is a practical component of Jobscaping™ which is the trademarked framework for HR professionals to design a career that balances high-level impact with true independence. Sovereign Workflows are designed to push back admin creep and support human sovereignty, ensuring you remain the architect of your time, not its administrator.

Personally I have handed 6 hours a week to AI and it’s making a big difference to the enjoyment I’m getting out of life actually. For example I have more time to attend live events and meet people. I was in London yesterday for a learning event hosted by 360Learning and was treated to an unexpected trip on the London Eye courtesy of their generous client, Merlin, whose L&D team were demonstrating the stickiness of storytelling and QED: I won’t forget Merlin in a hurry!

What would you do with an extra 6 hours a week if it was yours to do whatever you wanted?

Me and the wonderfully talented Anna Davenport from Good Habitz.

On the other side are the Sovereign Architects.

These professionals have realised that their value doesn’t lie in the “prompting,” but in the infrastructure. They’re not just using a tool; they’re building a digital engine that runs their Jobscaping™ frameworks. This allows them to focus on the high-level advisory work they were actually hired for.

A Note on Frameworks

I often speak about Jobscaping™ it’s the framework I use to help senior practitioners move from being ‘qualified hands on deck’ to becoming truly Sovereign HR Professionals. It is about shifting the focus from billable hours to high-level impact.

To support this, I share Sovereign Workflows here in the Hub. These are pragmatic, AI-augmented components designed to handle the heavy lifting of your practice, leaving you with less clutter and more headspace.

The Three Milestones of Sovereignty

• The Bot (Efficiency): You use AI to help you draft. You might save an hour, but you often lose it again to the “mental switching” required to manage the tool.

• The Project (Consistency): You build persistent workspaces. The system now understands your professional voice and your client’s specific culture. You stop explaining and start executing.

• The Agent (Autonomy): You move toward local frameworks where work is triggered by a folder, not a conversation. You start your day with an onboarding pack that is already 80% complete.

Why the Leap Matters

The shift from “Bot” to “Agent” isn’t a technical hobby; it’s a deliberate protection of your most valuable asset: your attention.

When you move your workflows into a structured, autonomous environment, you regain your Sovereignty. Your data remains secure. Your routine shifts from a series of “tasks” to a series of “reviews.” You move from being the administrator of your time to being its architect.

It’s the difference between using the tool and building the engine. Less clutter, more headspace.

A Moment for Reflection

Before we move into the practical “how-to” of these workflows in the coming weeks, I invite you to look at your diary for the past seven days.

How much of your time was spent in The Bot phase ~ performing high-speed digital admin, rather than acting as the Architect? It is a useful metric for how much “admin creep” has surfaced in your practice. We will be exploring how to bridge that gap very soon and we will explore the risks as well as the opportunities.

Launching The Sovereign Career Hub for independent HR professionals. Practical AI guidance that reduces admin fast, without compromising judgement, confidentiality, or your value. The efficiency is digital; the authority remains entirely yours.

Carolyn Shepherd

Founder, The Sovereign Career Hub

Subscribe now

Download Free Guide

A practical operating system to underpin independent HR work. Designed to free up attention time and amplify sovereignty.

Yes we use AI when it’s appropriate.
But this is Not an endless list of AI tools.
Not “how to prompt”.
Not “how to get clients”

Why me, why this, why now

It’s my combination of legal training, senior HR leadership, and deep practical AI expertise that places me in a unique position to help experienced HR professionals build sovereign careers.

Over 20,000 professionals have completed my LinkedIn Learning course on AI for non-techies which is a huge boost but being recognised on HR Magazine’s Most Influential Thinkers List was the endorsement I needed.

Since then I’ve designed my Jobscaping™ framework to help experts articulate and protect their irreplaceable human value as AI accelerates and threatens to reduce all knowledge workers to dust.

In a nutshell my work focuses on human-led adoption that is sustainable, ethical, and commercially sensible and HR is my tribe.

Because the shift is real. More HR professionals are becoming consultants, interims, fractionals and independents. The freedom is exciting, but the workload can quickly become chaotic.

This shift isn’t slowing down. As more HR professionals move independent, the market will get noisier and more competitive. That’s why getting your Sovereign Career system in place now matters, while you can still set your standards rather than chase them. AI can make that chaos better or worse depending on how you set it up.

This hub exists to make sure AI frees up your attention time and amplifies your sovereignty.

The problem: independent work creates an attention crisis

Most experienced HR professionals don’t struggle with capability. They struggle with volume.

When you go independent, you inherit everything: admin, scheduling, drafting, updates, client comms, documents, notes, follow-ups, and the constant background hum of keeping work moving.

Then AI arrives.

Used well, it becomes leverage.
Used badly, it becomes more noise.

The risk is not that AI replaces you. It’s that it drags you into low-value tasks and slowly shifts how clients perceive your value, from judgement and outcomes to output and turnaround.

That is how sovereignty gets lost.

Subscribe now

Download Free Guide

The solution: a system you govern, not tools you chase

Sovereignty, for me, is very specific:

• you protect your attention

• you set boundaries around what you do and do not deliver

• you package your expertise into clear offers

• you use AI to remove clutter, not to compete on speed

This is the key move:

Stop “using AI” and start governing a working system.

A system that reliably handles the repetitive work and creates space for the work only you can do.

The 90/10 principle

This Substack is built around a simple idea:

90% human attention on high-value HR work.
Judgement. relationships. decision-making. strategy. complex cases. trust.

10% managing your AI stack.
Drafting. triage. formatting. first passes. coordination. admin.

If AI is taking more than 10% of your time to manage, the system is wrong.

I’ll help you fix it.

What you’ll get here

To start strong, I’m making everything free for a limited time.

As a subscriber, you’ll receive:

The Sovereign Audit (welcome gift)

A practical tool to calculate your Clutter Cost and identify exactly where your attention is being drained. Download when you subscribe.

The 90/10 Newsletter

Bite-sized guidance designed for time-poor HR professionals who are building independent working lives. Practical, skimmable, and immediately usable, using my proprietary framework, Jobscaping™, so your work stays sovereign as the tech shifts around you.

No jargon. No overwhelm. Just a working model that holds up in real life.

Live working sessions (from time to time)

From time to time, I’ll run live working sessions (often Lunch & Learns, but not tied to a strict timetable). Some will be practical walkthroughs where we build parts of the Sovereign system together. Others will be mindmap-style explorations, where we map a familiar HR situation and trace what’s really creating the workload.

For example: an employee relations case that spirals into endless email, policy interpretation, meeting notes, stakeholder updates and document versions. We’ll map the moving parts, decide what should stay human-led, and identify what can be systemised or AI-supported without compromising judgement or confidentiality.

And crucially, we won’t just talk it through. We’ll capture the output so it becomes part of your working system. That might mean turning the map into:

a simple case workflow
a set of standard prompts and checklists
a triage script for first calls
a clean document structure
a draft pack for the most common comms (emails, meeting notes, stakeholder updates).

The goal is always the same: reduce repeat effort while protecting the parts that require human judgement.

I’ll announce sessions in advance, and I’ll share a replay where possible.

Start here

The first step is identifying your Clutter Cost.

Click the link in your welcome email to download the Sovereign Audit PDF, and let’s begin.

Because a Sovereign Career isn’t built by working harder.

It’s built by protecting what you’re worth, and designing a system that keeps you in control.

Download Free Guide

Subscribe now