What I'm focused on right now.

Last update: August 2026. Based mainly in Greece, Valtellina and the Dolomites, working across to the UK. The diary is full until 21 September.

Practising

Fab Campaigns is my advisory practice, and the platform through which the partnership methodology gets applied. Board-level governance, partnership architecture and licensed frameworks deployed with regulated institutions in the UK and Italy.

This month the practice is being consolidated. One shape instead of a list of things I am capable of. It began as tidying and turned into a decision, because putting the services into a single structure forced a question I had been walking past for a year: which parts of this work do I actually want more of.

The answer came back faster than I expected, and it is not the one a business development plan would have written for me. I want partnership architecture, and I want the work of moving research towards something that can be manufactured, sold and lived in. I want less of business development for its own sake, less of grant applications, less of the administrative machinery around both. That is not a complaint about paperwork. Somebody has to do it and do it properly. It is an observation about where thirty years of evidence says the energy goes, and it goes into the structure of a relationship and into the distance between a laboratory and a building site. So the practice narrows towards those two, and the rest gets partnered out to people who are genuinely better at it than I am.

Three things have to land before 12 August, because the week after that I am up in the mountains and on remote sea cliffs, and anything unfinished stays unfinished.

The first is a set of board-facing deliverables on AI governance for the UK. The task there is not the argument, which I have been making in public for months. It is turning the argument into something a director can act on in a meeting: classification rather than obfuscation as the honest fix for data in AI systems, supplier access and the difference between capacity sitting inside your borders and control sitting in your hands, and the three questions I now put to any AI number I am shown. Who funded it. Could the funder also be the customer. Would you know from the accounts.

The second is for the Veneto, and it is the timber work. The forest-to-construction supply chain written up properly rather than gestured at: which businesses actually occupy each stage, from the forest consortia through first transformation to the panel and prefabrication companies; what building in timber costs against masonry and concrete once you compare like for like; and what the lifecycle numbers say when you take them from peer-reviewed studies instead of brochures. There is also the question almost nobody funds, which is how buildings made of wood affect the people who live and work inside them. Minor forest supply chains nobody has organised. Local timber in public construction. The persistent gap between what the universities prove and what the valleys build. That gap is the business, and this is the month it stops being a conference conversation and becomes a document with a region’s name on it.

The third is the consolidation itself, which carries the same deadline for the same reason.

Meanwhile the frameworks keep proving themselves by being used, not described. The Stage Gate, the partnerships framework, the FabCube, the Information Classification Protocol and the fortnightly one-to-one are all in live service across the Italian engagements. The one-to-one framework in particular has earned its place: thirty minutes, hard stop, written pre-read, one Main Thing per call. Built for my own relationships first, now handed to others and holding.

Building

FabCampaigns turned ten years old last week. What an achievement. It is still a one-man-band organisation, intentionally so. I will write soon about everything the practice has covered, and frankly I am struck by the scale of the journey: the path, the topics, the evolution. The recent years have been in constant acceleration, with positive results year on year. I have built and invested a lot in seeding other ventures when I should have been focusing more on FabCampaigns itself. That becomes a priority again, probably from October.

Smart Mountains is in its seeding phase. Environmental intelligence infrastructure for Alpine territories. Governance and cap table are formalised, and the work now is investor conversations and territorial stakeholders, in that order rather than the flattering one. Fabiano and I cannot make this happen alone, for several reasons, so we will decide whether to go ahead in October, once we have settled the new company structure and the team we are going to announce. We need new talent and energy on board, not only on strategy and governance, but on the technical architecture and authority, and on local pilot deployments.

Imprese Favolose, the Alpine incubator/accelerator co-developed with Centro Consorzi in the Dolomites, has moved from framework to formation. Pre-incorporation governance is drafted under both English and Italian law, and the cross-border structure is sitting with counsel before anything gets signed. Slower than enthusiasm would like, exactly as fast as it should be. I need to see which stakeholders are actually on board: partners are now asking for execution, and I will not commit more time and resources unless the senior stakeholder at CC is substantially on board too.

The operational side does not wait: the first dedicated resource is onboarded in Belluno, the fab lab network is taking shape, and the pipeline is full of real industrial material rather than intentions. The framework agreement underneath it all was rewritten this summer, from ten legal articles into nine that name outcomes, dates and who is accountable. That change tells you more about how the year is going than any deck would.

Writing

Separation, Not Ownership went out on 29 July, and it is the longest and hardest thing I have published here. Circular financing in AI infrastructure: Alphabet invests in Anthropic on condition the money is spent on Google compute, Amazon does the same and books the revaluation as profit, and two customers now account for more than half of a two-trillion-dollar cloud backlog recorded by the providers who helped pay for it. The piece matters to me less for the finance than for what happened inside it. I set out five conditions a non-extractive programme would have to meet, ran my own test, and watched it fail, because it measured endowment rather than intent. Publishing the failure was the point.

The Ideas strand is now clear: data and AI governance told through live cases rather than theory. The NHS Federated Data Platform piece, then classification over obfuscation, then Garmin, then this. The next one is forming around Brussels against London, and which regulatory geography frontier AI actually answers to.

Monday Momentum continues. Friday Field Notes when the trail or the crag insists. The book club has a two-part autumn arc in planning: Ed Conway’s Material World and the E and S of ESG in October, then Brussels against London in December, when the regulatory picture will have moved again.

Moving

Physiotherapy has restarted properly, and the season is officially a base volume phase. Easy mileage, conservative loading, no heroics. The objective sits in 2027, and everything this year is in service of it.

The base is mainly in the mountains until September. Dolomites and Valgerola: family, easy rock with people who matter, long approaches at recovery pace.

Then a Greek interlude across three islands that each ask for something different. Kalymnos for the limestone, Chalki for the fact that there is no point, Karpathos for the meltemi and the villages the tourism never quite reached. The shoulder decides the grades. I decide everything else.

Exploring

Last quarter I wrote that I was already doing board work without the title. Nothing since has changed that, and the consolidation has sharpened why it matters. The two things I have just admitted I want more of, partnership architecture and bridging research into industry, are exactly the two things a board asks for and rarely gets from an executive already inside the business. So the search stands, and the filter stands with it.

My first proper UK NED appointment: data privacy and ethical tech, ESG and climate intelligence, or bridging R&D between universities, private equity and industry. Fit, not availability. The appointment has to sit close to something I care about, and what I care about is written across this whole page. I would rather wait.

On timing: I am full between now and 20 September, and saying so plainly is more useful to both of us than a diary of half-attention. Write to me anyway, because I read everything and nothing gets lost. I will simply not start anything new until then. If it can wait five weeks, it gets a better version of me.