There are some jobs you leave because the evening has ended.

The plates are cleared. The conversation has run its course. Someone is stacking chairs. It was good, perhaps even brilliant, but it’s clearly time to go home.

Leaving DWP has not felt like that.

The image I’ve had has been of walking out of a busy kitchen in the middle of service. Everything simmering away nicely on the stove. The ingredients good. The chefs knowing what they are doing. The smell promising. And yet, for reasons that are right and good and exciting, and of God, leaving before dinner was served.

Except that, last Monday, the metaphor shifted slightly.

Because some of the first plates are now leaving the pass.

The Prime Minister used his London Tech Week speech to announce a new AI jobs tool to help people out of work find the right jobs, create their CVs and get back into work. He was referring to the new experimental Work Hub, with its AI work assistant, CV builder, action plan, jobs board, and support for employers.

Some of this is new. Some of it is remaking things government has had around it for a long time. Some of it is very much at the start of its life. That’s why the experimental badge is important. It’s also why it’s brave. DWP has made something visible early enough for people to use it, test it, misunderstand it, criticise it and improve it.

So have a play, test it to the limits of your amusement, and maybe you’ll be pleasantly surprised, maybe you’ll learn something actively helpful.

Still, given what is being imagined for the eventual menu, this is very much the hors d’oeuvres.

But I’m no longer in the kitchen.

I’ve taken off the apron, stepped into a different kind of work, and now find myself watching the first plates appear with a strange mixture of joy, pride and longing.

More prosaically, I’m a couple of weeks into a new job at the Tony Blair Institute after 18 months working with the jobs and careers service team at DWP, latterly split with the work Alan Milburn is leading on young people and employment. I took a couple of weeks in between to stop and breathe, including a few wonderful days with God at Penhurst Retreat Centre. 

I wanted to take stock of this moment, especially because this has been a different kind of work transition.

This is the fourth major work transition in my career. In the past, those moves have come when I was quite happy to be calling time. The work had been good, in most cases brilliant, and while the next opportunity was always exciting, there was a sense that one season had reached its natural end, and I wasn’t doing much looking back over my shoulder.

The next opportunity is exciting, as the next opportunity has always been exciting. The difference this time is that I was still deeply excited by the work I was leaving.

That’s an odd thing to feel. A good odd thing, but an odd thing nevertheless. I’m grateful for the time and excited for what’s next. I’m delighted to be working internationally again, and doing so with my friend Barbara Ubaldi who was also my boss from my time at the OECD.

But I’ve still left DWP with a faint ache. Not regret, more the particular feeling of stepping away from something you’ve believed in and invested in just as it’s beginning to come into the light. 

A season of side quests

When I blogged about my return to work after leaving the OECD, I described the next season as a mix of thinking, advising and doing.

That was what I wanted. I knew I wanted to keep working on the things that had occupied so much of my attention internationally: institutions and systems, digital maturity, public sector capability, and the foundations that make transformation possible at scale. But I’d also missed the gnarliness of delivery. I’d missed the moment when an idea has to survive contact with reality.

The last 18 months gave me both, sometimes in the same week.

There was work with NHS England, supporting Rachel Hope’s excellently put together pocket of brilliant digital practices with Richard Pope and James Plunkett (who has been writing up excellent windows into that and have just published operating patterns we were seeding last year). There was work with GIZ and the Government of Kosovo on how to measure the progress and impact of digital transformation. A good reminder that measurement can so easily become theatre, but when it’s done well it’s about asking better questions. Are services improving? Are people experiencing something different? Is trust changing?

There was a paper for the World Bank about the building blocks low- and middle-income countries might need to consider if they are to harness AI successfully. I’d love to link to it, but unfortunately the paper isn’t going to be published, though I understand some of the material is going to find another home. Its argument was that the foundations for AI are broadly the same as the foundations for more general transformation in government: maturity, governance, institutional capability, responsible boundaries, and public value. The focus can’t be on the shininess of the tools themselves.

There were smaller but no less enjoyable pieces of work too, from an invitation from the UN University in Guimarães to deliver training on interoperability and service design, through to working with Digital Nation as part of their advisory work to the Thai government on digital transformation.

All of that was brilliant.

But it was also, in the nicest possible way, a set of side quests.

The main event was jobs and careers.

The unexpected hopefulness of DWP

DWP isn’t necessarily the department people reach for when looking for signs of public-sector imagination and optimism.

As an institution it carries a heavy load and a weighty responsibility. List the obvious associations and it’s quite the laundry list: conditionality, economic inactivity, sanctions, fraud and error, health assessments, work, benefits, etc. The whole organisation sits in the midst of intense political pressure, media scrutiny and big debates about human worth. No one who has seen I, Daniel Blake can fail to be affected by it. It’s basically impossible not to arrive at DWP without having priors already set.

And yet, the jobs and careers service has to be one of the most exciting and hope-filled things happening in government right now.

The labour market reaches deeply into people’s lives. Work is not the whole of life, and it is always a mistake – in welfare policy, migration policy or anywhere else – to treat employment as the measure of someone’s worth. But work often shapes a lot: our household income, obviously, but also our personal confidence, our sense of belonging, our status in society, our mental health, and our aspirations about what life looks like.

So a service that helps people navigate work is not simply another transactional service.

It is not only a better job board.

It is not only a more modern Jobcentre in your pocket and, hopefully, on the high street.

At its best, it’s a trusted touch point at some of the most difficult and formative moments of life: leaving education, losing a job, returning after illness, changing career, rebuilding confidence, hiring someone for the first time, or trying to work out whether the next step is even possible.

That’s the challenge I have loved being close to.

A year ago I mused about whether we could ever speak of public institutions with something like gratitude. I wondered whether one day someone might say of the jobs and careers service:

The Jobs and Careers Service walked with me. Through the despair of losing a job and the joy of promotion. In my dead ends and fresh starts. It was there. It helped. It did me good. It didn’t save me. But it didn’t fail me.

I hope that’s what this work achieves.

There’s an opportunity here for DWP to reimagine its role and unlock growth in the economy. But more than that, there’s an opportunity to build an institution that understands work not merely as an economic category, but as a human one.

That’s why it was such a privilege to make a small contribution to the work Alan Milburn is leading on young people and work. The interim report does a great job of getting under the skin of a challenge that too readily attracts one of the oldest and laziest arguments: that young people are feckless, entitled, badly mannered, insufficiently grateful, and probably proof that civilisation has entered its final decline.

The harder question is what it takes to build routes into work that are credible, human and matched by effective support. I hope everyone involved in that work is proud of the vision they have set out and of the solutions that now need to follow. And, as I wrote in Strong branches and good shade, I hope that the church can step up and play its part.

The impossible dream

At DWP my job title was “Transformation Lead” and on a good number of occasions people would ask what that actually meant.

The answer I landed on, only partly as a joke, was that I brought the vibes.

That could sound flippant. It’s not meant to be.

Not the vibes that transformation is easy. Not the vibes that everything becomes magically simpler if we say “AI”, “data-led” or “user-centred” often enough. Certainly not the vibes that a prototype is the same thing as a service.

The useful vibes are more stubborn than that.

They are the belief that a better future is possible and worth aiming for. The willingness to imagine an impossible dream, recognise that it is probably impossible, and keep trying to make it real anyway.

That is what public-service transformation so often requires. Not fantasy. Not hype. Not denial about constraints. But the refusal to let today’s operating reality become the limit of tomorrow’s ambition.

AI has made that tension sharper because it has changed the texture of imagination.

Which leads me to the other vibes too. The coding ones.

I have had a frankly ridiculous amount of fun using AI tools to make things, test ideas, break things, rebuild them, and collapse “what if?” into “have a look”. That started as a practical excitement in vibe coding, fireworks and the mortar of government, then became a much bigger set of thoughts about public-sector product management in a vibe-coded world, and what it might mean to go beyond the vibes. There’s still more to share about all of that.

But I wouldn’t argue that the lesson is for everyone to become a developer, or that prototypes are magic, or that public services can be transformed by enthusiastic tinkering alone. But I would argue that more people can now make something tangible, and that has huge potential for good.

Ideas that would once have stayed as hard-to-communicate abstractions can become something colleagues can see, touch, argue with and improve. That should make organisations more adventurous and imaginative.

But it also means they need more discipline, not less. The fact that it’s easier to make something doesn’t mean it’s easier to make something good. Or safe. Or accessible. Or sustainable. That’s why I think we should be talking about AI-assisted delivery.

And it’s where the impossible dream has to meet the mortar of government.

The jobs and careers service has been such a fascinating place to work through that tension because the problem space is both broad and intimate. It stretches from labour-market strategy to local Jobcentre conversations, from AI-assisted tools to the vulnerability of someone wondering whether anyone will ever hire them again.

That is why last Monday changed this kitchen dining metaphor. The Prime Minister celebrating the work in a speech does not prove the service is done. Launching experimental tools does not carry the full weight of the ambition. An AI work assistant does not fix the labour market. Of course it does not.

So while these tools are useful and important, they’re only the first things coming out of the kitchen.

The meal is bigger than the starters.

So what’s on that bigger menu?

Jobseekers, employers and colleagues

The obvious place for a jobs and careers service to start is with the person looking for work.

But even that category is less simple than it sounds.

It includes people who are out of work and desperate to find something. People already in work but needing better work. People whose health has interrupted their plans. People leaving education and trying to get their first foothold in the labour market. People who have skills but not networks. People who need flexibility, adjustments, childcare, transport, proximity, or the right person giving them belief and confidence about their next step.

The visible tools begin to speak to some of that. A CV. An action plan. A way to ask questions. A way to look for jobs. A way to begin when you are not sure where to begin.

For someone who is stuck, that first bit of traction can count for a lot.

But those next steps only become real if there is an employer on the other side of them.

That’s why the service cannot only treat employers as a source of vacancies to satisfy the needs of jobseekers. It would be easy to make the ambition too small: list jobs, improve matching, make the transaction better. That would be useful, of course. But there is a bigger prize available.

I wrote at greater length about this when I explored the idea of DWP as the HR department of the growth mission. What would it mean for DWP to really lean into being the ‘Department for Work’? To make those infamously increased NICs feel like government was offering some value in return? To be HR for businesses that do not have HR? To help an employer write a better advert? To offer a simple way to verify the right to work? To help someone be more confident in how they run interviews fairly? To make onboarding someone just a little less daunting?

Because employing someone is not a light thing.

A business puts trust in a person. A person puts trust in a business. We should all care about that relationship going well, because the health of those relationships is part of the health of the labour market and reflects the wider state of trust in society.

And then there are DWP’s own staff.

Work coaches are unsung heroes. They often have to contain important and complex conversations within appointments that are too short, in a system where the dominant external frame is justification for paying a benefit rather than helping someone take their next step forward. The job is harder, more relational and more morally serious than the public argument usually allows.

There is a huge opportunity for technology to pick up some of the formative activity around those conversations so that coaches can have interactions that are more useful, more human and more hopeful.

That’s easier to write than it is to execute. None of this is easy.

The labour market is under strain. Young people are struggling to find secure routes in. Health-related inactivity is a live and difficult challenge. The welfare bill is big, and getting bigger. Employers are facing their own pressures. Public trust is badly eroded. The technology is changing faster than our institutions are used to changing.

I recently reflected on whether there can be optimism about a pessimistic future of AI-induced work-lessness. Whatever the impact of AI on the labour market, whether the future is more abundant or more disruptive or a mixture of the two, the country will need better labour-market institutions.

If work changes, and if the entry points into work become narrower or less predictable, then a jobs and careers service that is exceptional in how it uses technology so that it can invest more in what it does for jobseekers, employers and frontline coaches is not a nice-to-have, it is critical national infrastructure.

That’s why I’m so glad some of it is now visible.

Not because the first visible version can possibly carry the full weight of the ambition. It cannot. No first version can. But because it gives the work the chance to be used, and misunderstood, and criticised and maybe even praised. It makes the service more exposed, more vulnerable, and more accountable.

And making things open, because it makes them better, is where public-service transformation should always be headed.

Back outwards

Which brings me to the other half of this transition: I have not only left the kitchen. I have gone back outwards.

Joining TBI is a return to the work I was doing after leaving GDS and joining the OECD: working internationally, thinking with governments, and picking away at the hidden machinery that sits between good ambitions and actually changing people’s lives.

I’ve missed that international dimension. There’s nothing quite like bumping up against reality in a different context. A fix that looks obvious in Westminster looks different in Pristina, Rabat or Santiago. The same questions keep appearing, but never in quite the same shape. That mix of recognition and disorientation is one of the things I love about the work.

Already, in the first fortnight, conversations have bounced between Latin America, Central Asia, Africa and Asia-Pacific. Not as abstract regions on a map, but through people close to particular governments, particular leaders and particular contexts. That country team-led model is one of the things I am most excited about: the chance to work with teams who are close enough to reality that the work cannot float forever as nice global theory.

So this is a change that is also a return.

But not just a return.

When I decided to leave the OECD it wasn’t because I’d stopped believing in the organisation. I still believe in it. I would like the UK government to make more of it. At its best, the OECD is a genuinely valuable thing: a place where civil servants can be challenged and encouraged, where countries can learn from one another, and where good evidence and effective arguments give momentum, with ballast, to innovation and reform.

But there is a ceiling.

Sooner or later, the work reaches the place where political appetite lives. The place where a good idea is no longer just a recommendation in a report, but something someone has to own and fund and defend and spend political capital on. Ideas don’t fly because they’re correct. They fly, or die, because someone with authority decides they’re worth the trouble.

That is a big part of my interest in TBI.

Its work sits closer to that altitude. Closer to the people who decide what’s worth tackling, where priorities lie, and how much difficulty they’re willing to accept in pursuit of change. That does not magically solve anything. Proximity to power does strange things to people, and distance from delivery does stranger ones. But ideas that never touch political leadership are elegant things that do not happen.

So I’m grateful for these last 18 months. The nitty gritty of delivery sends me back into advisory work with recent experience of what advice is for. Frameworks and comparison and maturity models all have their place (I’m pretty much duty bound to say that having written several). But their value is tested when someone tries to use them.

Delivery is where theory becomes accountable.

That’s one of the gifts of having spent time with the team at DWP. 

I also grant you that TBI is not a shy organisation. It has earned a reputation for advocating for a technology-led and technology-laden future, and for doing so with a level of confidence or provocation that can make people wince. Sometimes I have winced. I assume there will be future wincing. Perhaps I will even cause some wincing of my own. After all, I’d like to think I’m a team player.

But influence matters. If an institution has influence, the question for me cannot be whether I would have written every sentence, chosen every emphasis, or framed every provocation in the same way. That would be an unbearable test for everyone involved, and particularly if you had to work with me.

The better question is whether I can bring something useful to the table: some delivery scar tissue; some deep-seated affection for the boring things that sit beneath the shiny things; a bias towards user needs, not government needs.

And perhaps also a belief that multi-disciplinary work is not just designers, researchers, engineers and product people, wonderful though they are, but policy people, operational people, commercial people, legal people and frontline colleagues alongside those UCD professionals working in concert to make the same thing real.

TBI sits at the optimistic end of the spectrum. It often asks whether all this technology can do good things for society and for each other. That is an optimism I can share. But I also understand why it attracts criticism. I have made some of those criticisms myself. So part of the contribution I hope I can make from inside that optimism is to keep asking a fairly boring question: if we want all that shiny, can we talk about the plumbing first?

Services. Skills. Data. Governance. Institutions. Infrastructure. Delivery habits. Trust. The stuff that rarely makes the speech, but decides whether the promise survives contact with the people it is meant to serve.

There is something in that optimism that is attractive to me too. If there has been a thread running through my blogging since I dusted this off and started writing again, I think it is hope. Not chirpy hope. Not “everything will be fine” hope. More a refusal to let hopelessness decide the limits of what is worth trying. Those vibes of transformation.

Ultimately, my faith is the source of that. My trust is in God, not in the state. I do not believe government brings salvation or creates utopia or is the Kingdom of God brought to earth via policy papers. But I do believe there is something deeply sacred about the responsibility carried by governments, because the lives affected by policy, services, systems and institutions are not abstract but lived.

This is why I remain so interested in the calibre of those who lead us.

It is an absurd thought, really, but the number of people who sit at the very top of national governments is vanishingly small. You could fit them into a football stadium. And yet the fates of billions of people are shaped by a tiny proportion of us all.

That is part of why I will spend the rest of my life circling the ideas behind Kingdom Democracy: what it means to love those who govern, without getting weird about power; to pray for leaders as children loved by God before they are ever avatars of a particular policy programme; to seek the good of public institutions without confusing them with God’s Kingdom.

Prayer as the engine, not an ornament.

I do not know how much of that is likely to ever come up in my work. Probably very little. Perhaps more than I expect. But I am excited by the possibility of meeting more people around the world who are trying to inhabit similar tensions: close enough to power to be useful, wary enough of power not to be taken in by it, and hopeful enough not to spend their lives rolling their eyes from the sidelines.

Mostly, I’m excited to learn more, meet more and hope more.

That’s not a bad prayer for a new job, really.

So yes, this is a move back outwards.

But I hope it’s not a move away from what DWP has brought home. I hope it’s a move carrying those lessons somewhere else.

Because the point of advice is not advice.

The point of advice is that, somewhere, somehow, someone is better able to make something good happen.

And after 18 months in the kitchen, that feels clearer to me than ever.

Cheering for the kitchen

So I’ve left grateful. Grateful to have been invited into the team at all, and grateful for the work I got to be near. But I’ve also left with mixed emotions, and with the full meal still unfinished.

But I’m still cheering for the people in the kitchen.

I hope NHS England colleagues pull off the changes they are working towards. I hope the work with Kosovo creates useful ways of seeing impact. I hope the World Bank’s AI material finds its way into something that helps governments make better choices. I hope the training and advisory work in Portugal and Thailand nudges good work forward.

But I especially hope DWP serves up something worthy of the people it exists to support.

For jobseekers. For employers. For work coaches. For people who have been left on the shelf for too long. For young people trying to get into the labour market. For people whose confidence has been knocked out of them. For businesses that want to do the right thing but need help.

I do have a couple more jobs and careers posts that I’ll finish writing, so this is not quite the end of my going on about it. But it is the end of this more formal chapter. And, honestly, this has been such a wonderful 18 months. 

So, to my former colleagues at DWP: keep going. Keep aiming at that impossible dream of making the service better than people expect government to be. Keep making the tools better. Keep making the service more human. Keep iterating and learning and imagining new ways to help jobseekers, employers and your colleagues.

And, obviously, keep bringing the vibes.