Introduction: a rollout that revealed the real challenge

One of the biggest things that I first noticed when I started at MindStone, so working with Josh, he's like, right, we've got a 250 -person company. I need you out there next week. We need to deploy Rebel, a new operating system.

And I walked into Athens, where the company was, and I had my decks ready. I had all my use cases ready to go. I'd been deep into the product because I was like this is all about the product I need to

know the product inside out and then when I got there and I sat down the first person up I was like okay great I'm ready I sit down I automated about 30 percent of this person's role in six minutes and I was expecting to light on their face instead I saw grief I saw fear I saw someone that had been an individual in that business who had been

the go -to person for this task that I'd automated and it really made me stop in my tracks and think to myself this is not about the product this is not about the technology this is about identity shift so I'd like to ask all of you in the room before I get into the presentation who here has been made

redundant has lost their job has maybe even been through divorce or has lost something really important to them please raise your hand and how did it it feel it's pretty scary and it actually takes time for human beings to process this before we can come to the other side so thank you there's three things that we like to talk

A practical framework for AI transformation

Tools, training, and enablement

about at mindstone to really help you with a full ai transformation and that is tools so agreement and alignment on what the tools are that the company is going to use

so there's clarity around that there's also training which Stefano had talked about before so actually making sure that everyone in your company and all the teams actually have the skills to be able to use these new teams and then

lastly it's enablement and the enablement is the human part so when I first started working for MindStone I thought it was going to be more of about the product and training but actually largely what my role has been is about guiding companies through different changes so the things that we've seen

Why champion-led rollouts beat mandates

that have worked really well are not mandating no back no mandates we started to create with this company out in Athens a small group of champions and it was an opt -in so we picked people who had been through the training program and who had actually got really great results they'd been working on other

products they were cross -function they were different levels but these people were not only enthusiastic they wanted to opt -in they were curious and then having that group of champions we were able to do peer rollouts so it wasn't me and an outsider coming in with limited context of their business they were able able to sit down with their teams.

1And I genuinely believe when we mandate things, we're forcing it on people. So if we're forcing it on individuals where they are fearful, they're worried about

their job loss, potential job losses, they're worried about this identity shift, my role is changing and the company is not helping me change that role, mandating is absolutely the worst way to go. You will have even more resistance with that.

Leadership modeling: making AI collaboration visible

One of the other biggest things that we've have seen is the leadership team has to model this behavior openly so one of the things that I started doing because I'm like well if I'm going to tell other companies to do this then

I better be doing it myself so I started signing all of my work that I had worked with in collaboration with the AI in collaboration with Emma my favorite human because of course I'm my rebel's favorite human and executed by rebel it's cute it's sweet but actually i want people to know that when i am collaborating with ai i'm not hiding behind that because one of our coaches

had said to me recently someone had stood up and it was about 150 people on a call i feel like i'm cheating when i'm using ai and if leadership isn't modeling this behavior then no one else is going going to get behind and do it. Because we all know we can say whatever we want, but we will follow what our leaders are doing and the leadership team.

So I had to sit with the leadership team and say, actually, you have to talk about when you're using products as well. You have to model it, whether it's in your signature or whether you're having sessions to talk about this and openly talk about what you're doing in the business where you are collaborating with your AI tools.

Really important.

Permission to fail (and to test)

Permission to fail. Things will go wrong. Outputs won't always be perfect at first. Lots of the technology that we're using is cutting edge. It's all new.

One of the biggest things that we talk about within training is if you had a really smart intern turn up to start working with you and you gave it very little context and you said, hey go do my email triage for me they'd probably not do a great job but if we were giving more context around that then actually get better outputs so to enable people to actually start getting the outputs that they want they really need to be given explicit

Explicit permission to fail. And explicit permission to test. And without that, people are fearful.

They're like, well, I really want to use the AI tool. I do. I am curious. But actually, what happens if it goes wrong? What happens if I send an email to my boss and I've attached the wrong documentation?

And you really have to say it out loud. You can't just hope that it's implied with what you're saying.

Celebration and psychological safety during change

we've seen with the companies that we've helped transform through using rebel and other ai tools is that there's a lot of cultural shifts that need to happen we need to start celebrating when someone has automated their role so rather than there being fear like that first person that i sat with the grief actually having a celebration saying hey i managed to automate nearly all of

of my inbox responses this week. I've automated 20 % of my tasks, 30 % of my tasks, 40 % of my tasks and celebrating those individuals is so powerful and it's reinforcing all of the other behaviours that we've talked about as well.

And psychological safety. 1You have to tell the team that it is okay and maybe we don't really know what the future looks like yet. We don't really know what our

role designs really look like and psychological safety to even hold space for them and those moments of grief and fear actually really go a long way and if the leadership team aren't endorsing this then we go back to the very beginning where grief and fear and then there's restriction people are reserved they won't jump into it they won't lean in so we talked a bit

From identity loss to agency: redesigning work with reclaimed time

about identity loss at the beginning one of the things we were working with the company out in Athens is okay if we're going through this identity loss maybe there needs to be a little bit of space held but also maybe it's giving our teams permission and actually recognizing that if we did automate 30 40 percent of their roles how about we give them agency to think about what they

would spend that time doing so when I sat in a room with a couple of people at Nathans I said said to someone so obviously after the first session where I thought okay that was a bit of a car crash that didn't really go where I was expecting it to go I started saying to people well what would you do with 30 % of your time back and I have had some very different answers to that

question I have had answers um one guy had said to me actually I'm a new father and I have so much work. I never get home for bath time and bedtime. So having 30 % of my time back is now going to enable me to both do my work, do a good job and have family time.

I had another woman say to me, I have an entire team and I haven't had any time to focus on training and upskilling them. So now I'm going to spend that time building the training and the upskilling. I mean, that's amazing.

amazing and it's those moments that change that conversation around grief because all of a sudden we have agency to start thinking about what are we going to do with that extra time that we're going to get that is a powerful mindset shift so rather than something that's being taken away from them they're now given the capacity to dream up what else there is and really I think in that

that positive note it's thinking about reinventing work for humans sometimes we don't like change sometimes we're really messy sometimes we're really fearful but by being given permission and agency to think about the future we all have ideas of what we want to do and how we want to do

better work here's my qr code if anyone's interested to continue this conversation please please feel free to connect with me on LinkedIn.

Rebel demo: turning methodology into a shareable one-pager

And I'm now going to do a very short Rebel demo because I couldn't end this session without giving a little demo.

Hey Rebel, I'm at Mindstone London Meetup and I would like to pick one thing that I could demo to the audience.

Maybe we could extract methodology from my IP to just give a quick one pager to the audience.

A familiar interface designed for easy tool switching

Okay, so what Rebel will be doing, and as you can see, Rebel is a really great simple interface. You'll notice that many of the tools, whether it's Claude, whether it's ChatGPT or any of the others,

we've really made a mess of this. I can hold it. I can hold it. It's all good.

All of them are very similar and that's actually by design. It's designed deliberately for us humans to be able to move efficiently between any of them and just pick it up really naturally. So as the team, as Stefano had alluded to before, Rebel workspace docs. Let's do a HTML in Rebel.

Memory and clarifying questions for better outputs

Rebel has memory. So I know it can be really frustrating sometimes when you come into your AR tool and you're like, hey, let's pick up that project we're working on before. or the conversation that I was having earlier about the Word documents that I need to send out.

It feels, it's like, sorry, which project? I'm like, okay, this is frustrating. One of the most beautiful things about Rebel is it's actually storing all of my memories.

So I feel like I have a digital coworker that works with me. Every time I want to pull a previous task that we've worked on before, I just say to Rebel, hey, can you go and find that document

that we're working on around our transformation methodology? methodology and Rebel will pull in all of the history from that. So it means every time I come back I'm building on something and it's getting richer and richer documentation.

So as you saw before Rebel had popped up a couple of questions. This is by design so a lot of the prompting behind this rather than just giving Rebel a really basic prompt it's going to ask me clarifying questions

to try and gather more context from me to give me a better output the gamma deck that I was demoing on before that was obviously built by rebel I said I'm in an audience at mind stone for the community event and I want to pick some high -level stuff not too many words and Rob will produce that deck in about five

minutes so rebel sometimes can take a couple of minutes I have something earlier that it absolutely does so one of the really incredible things about

Connectors (MCP) and privacy-first architecture

rebel is the team I think at the moment we have about 126 to be precise

different connectors so it's an MCP which is a model context protocol which is essentially a connector between the different tools as you can see there's a a ton of them through here and it's super simple you click on it I don't have an Airtable account but it would bring you up and actually walk you through the guidelines so it's built to be as intuitive as possible to actually get you connected and then off you go it's also privacy first so everything is stored locally and let's come back to see where rebel has got to

The generated output: transformation guidance in minutes

here we go okay the human side of transformation 70 % of AI rollouts fail to reach sustained adoption that is true 50 % of changes of failures caused by inadequate readiness I would totally agree that's BCG and there's two times adoption rate when a leader opens with personal vulnerability absolutely spot -on so as you can see from this

rebel has pulled a lot of what I've talked about because this already sits in my memory it's turned into a HTML canvas so if I wanted to share this with the customer I could just essentially save it as a PDF and attach it to an email it's given us five moments to determine where the transformation takes takes root. I'm not going to read all of this out but I will scroll down.

I don't know about you but I find that pretty impressive that that popped up in about three minutes.

Conclusion: automating the routine to do more human work

I think that I've probably, I have automated almost about 45 % of my role and what I'm spending the time on now is things that are really human.

Relationship building with my customers, relationship building with the team more strategic work the stuff that takes a lot of thought and deep thinking to figure things out where i need capacity and i shouldn't be context switching

and then one of the other things as well that we do as well within mindstone is those rituals so we still have company stand -ups and the things where we're bringing the team together the things that are really human that ai can't take away so actually by being more ai enabled

within MindStone we are able to connect better as a team and do more strategic work and automate a lot of the stuff that actually I'd rather not do anyway

and at the end of the session Peter's going to talk about open sourcing but I'll let you do that afterwards any questions come and see me at the end

thanks very much