Year note: 2023

Gavin Freeguard
15 min readFeb 8, 2024

--

A belated Happy New Year to you. If not quite as belated as last year’s year note on the long 2022, which covered some bits and pieces from early 2023.

A few additional reflections before the ‘look upon my works’ bit…

Time to reflect: chairing a justice data special of Data Bites (me on the left of the timer showing 7 minutes 47 seconds left, with the brilliant Anna Powell-Smith to the right).

Sausages

I was at a barbecue over the summer where a friend of a friend, having heard what(ever it is) I do for work, looked at me and said: ‘I hope you don’t mind me asking, but isn’t it… depressing?’

On the one hand, and in the main, the answer is ‘no’. I really enjoyed 2023 (too mad a start to the year, the potential to perjure myself — more on that below — and context switching not being a billable client aside). I got to work with great people (thank you all) on a lot of great projects and, with an election looming, it felt like some Policy is starting to happen and there is a positive difference to be made.

On the other hand… there are good reasons to be depressed with the state of everything. It’s an uphill struggle trying to push an exhausted, gaslighting government, that sees political gameplaying as the end rather than the means, to a better (less bad, rather than necessarily good) place. Working around the Data Protection and Digital Information Bill has also been one reminder of just how outdated and insufficient our systems are — it can still be a shock seeing the sausage get made, even for someone largely inured to the gristle and ghastly delights of the public administration abattoir.*

Also: long-term, gradual, behind the scenes, sometimes intangible soft power and influencing work really matters and is often undervalued. And when it’s good, it feels very *very* good (as it did at various points last year). But I missed having some easier dopamine hits of getting satisfying shorter-term tangibles — e.g. research reports — out the door. I’m also conscious a fair amount of my work last year was summarising and connecting, rather than thinking and doing my own thing. Again, the former is vital — but I might need some more of the latter.

Hot takes

I’ve found myself reaching for a particular George Orwell quote on several occasions this past year:

‘So much of left-wing thought is a kind of playing with fire by people who don’t even know that fire is hot.’

But replace ‘left-wing thought’ with ‘discussion about AI’.

The hot takes and bad journalism (political correspondents, I’m mainly looking at you) were overwhelming at times. Knowing there are understandable reasons why our media ecosystem is where it is doesn’t make it less annoying — or any more acceptable. Words have consequences. Narratives matter. And while working to shape the narrative is part of the campaigner’s lot, it doesn’t absolve others of needing to do their jobs properly.

Mind the gap

I’m lucky to work with lots of organisations that do some great work around digital and data in government. But I’m increasingly convinced that we’re missing a thinktank, research institution or similar that focuses on these issues.

I’ve started thinking of it as four pillars: data and information in government, digital government, open government and the future of government (the blue sky thinking, cultural and institutional stuff that doesn’t quite fit elsewhere). I don’t think we can make our governmental and political systems better (see sausages, above) without engaging with these topics. I’m currently jotting down some ideas — if nothing else, it gives me a list of projects I’d like to pitch. Please get in touch if you’d like to chat further.

Relatedly, various conversations with Dave and Derek led us to try organising some monthly drinks for people interested in public sector reform. It came from a sense that there isn’t really a space for it — things like GovCamp are vital and brilliant but once a year, while OneTeamGov feels like it has withered. We could definitely have organised things better, but it was difficult to build momentum. Random observations: there are a lot of ex- (or never-) government people interested in the issue; I wonder if we’re missing grassroots reform movements in the public sector; perhaps everyone is just too exhausted to be able to do anything about it. We ran a session at GovCamp, and we’re thinking of having another go — survey here if you’d like to share your thoughts or get involved.

Anyway. Here’s what I got up to, workwise, in 2023.

The Covid Inquiry

Inquire within: appearing at the Covid Inquiry

It was a privilege to give evidence to the Covid Inquiry — you can read my written evidence (all 30,000+ words of it…) on data and digital government, or watch my oral evidence session from October and read the transcript. openDemocracy did a short piece on my evidence (what a URL that is) and medConfidential made some useful notes on my evidence, too. Matthew Somerville has a site making all the transcripts searchable.**

The Inquiry team approached me in late 2022 seeking an expert witness on data, data sharing and digital government. It would be fair to say that my imposter syndrome kicked in: I’d published a few things on data and digital government during the pandemic, but surely — SURELY — someone, somewhere had done more in-depth work? (Not least since much of my output was for an IfG deprioritising those topics.) Apparently not. It did bring to mind one of my favourite quotes about government and politics:

‘You might think somewhere there must be a quiet calm centre like in a James Bond movie where you open the door and there is where the ninjas are who actually know what they are doing. There are no ninjas. There is no door.’

A quote from one of the higher-profile Inquiry witnesses, one Dominic Cummings. The thing he misses is the more terrifying realisation: people think there *is* a door, and that you’re on the ninja side of it.

I can’t fault the Inquiry team — they looked after me very well, and removed a lot of the terror from the occasion; I think I avoided perjuring myself. (A fellow witness compared it to going for a medical procedure, where you’re given a nice room, asked if you’d like a nice cup of tea, taken into theatre, brought back to the nice room and asked again if you’d like a nice cup of tea.) And I don’t envy the Inquiry — one member of the team likened it to running more than ten public inquiries simultaneously. But I do wonder if there might be a more productive way of doing things. It felt stilted, slow (running at the pace of the stenographer), stage-managed and short — inevitable, given how little time there is to convey key messages to the primary audience of one, the chair, and the huge range of important issues to cover. I suspect we already know the main contours of what needs to change to be ready for next time. And I understand the real need for both the substance and the theatre of putting the main political players on the stand. But I wonder if there might be better ways to get into the detail, too.

It was weird writing my evidence based on work I’d done and what was already in the public domain — I didn’t have access to the testimony of those working on data inside government, apart from the odd extract immediately before my oral evidence (some of which looks well worth a longer read). That didn’t stop one government department irritably inquiring why I hadn’t covered certain issues in more detail — for one such subject, I searched GOV.UK and returned just two glancing mentions. That might be why. Government transparency, eh?

It also turns out you can’t rely on parliamentary proceedings — including very useful select committee evidence — thanks to the Bill of Rights. (Yes, the 1689 one.)

Wearing my Connected by Data hat, in December I gave evidence to PACAC’s inquiry on transforming the UK’s evidence base with an emphasis on data ethics, alongside Reema Patel from Ipsos, and managed to get a Jurassic Park reference in nice and early.

Connected by Data

It’s been a joy working with Connected by Data — with both the brilliant team and a varied group of great people and organisations from across civil society.

The Data Protection and Digital Information Bill continues to wend its way through parliament. In 2023, I published a write up of our December 2022 event (in parliament) about it, chaired lots of meetings including a March online workshop, live-tweeted Second Reading of the Bill in both Commons and Lords (on the last day of term before Christmas, what a joy), wrote up Report/Third Reading in the Commons (and Second Reading in the Lords), and maintained our list of useful resources on the Bill.

Then there was AI policy Christmas, as the AI Safety Summit sucked up (far too much of) everyone’s energy — I helped run our workshop at the AI and Society Forum convened by Promising Trouble.

Indeed, trying to keep up with the various data and AI policy developments in a pretty crazy year felt like a full time job. Having ditched my dataviz and data link compilation newsletter, I found myself writing a new Data Policy Digest (one, two, three, four, five, six, seven, eight, nine, ten in 2023, subscribe here). Colleagues not unreasonably joked that ‘Digest’ is slightly misleading given the length of them — hopefully more regularity and ruthless editing (and please god, less news) might make them shorter this year.

Despite having so much to react to, we also tried to get proactive ahead of the election, convening and consulting civil society voices on what data and AI policy could (and should) look like. Our Progressive Vision — a short pamphlet aimed at Labour and launched with a great event at Labour conference, alongside a longer, less Labour-focused version — seeks to kickstart that conversation, with framing, values and some policy ideas. If it achieves nothing else, hopefully it makes one key point: people and politics can shape the future of data, AI and tech and there are choices that we can and should make. (I also wrote up our event at my first ever Green Party conference.)

Eesti is East: inside the Linnahall in Tallinn. I could have used a picture of the beautiful old town but went for a decaying Soviet relic instead.

Connected by Data also took me to Helsinki, for the excellent myData conference (I presented on collective data futures), and Tallinn, to help facilitate a design lab on open government commitments ahead of the Open Government Summit (it’s only taken a decade of working on digital government, but I finally made it to Estonia!). Closer to home, I headed to Leeds for a techUK/dxw discussion on data and digital transformation for the Leeds Digital Festival, and Blackfriars for a Nesta event on data capacity in local government.

Open Data Institute

I started the year as interim head of public policy — as I said in my 2022 year note, my first interim position and quite a challenge on one and a half days a week. That didn’t allow time to do much other than keep the seat warm for (and help appoint) my (excellent) successor.

As interim, I spoke on Power & Diplomacy: When Open Data and Realpolitik Collide at the OpenUK conference, contributed to a preview of DPDIB upon its parliamentary return, co-authored a post on data access and the Online Safety Bill, looked over work including a report on data assurance, and briefed and wrote up ODI chair, Sir Nigel Shadbolt, at the science and tech committee on AI.

Back as special adviser, I’ve done a fair bit in the background (briefings for events, internal sessions, etc) but also contributed some thoughts to another piece on DPDIB and supported the ODI response to various consultations, including generative AI in education (DfE), the future of transport data (transport select committee), open communications and Smart Data (DBT), transforming the UK’s evidence base (PACAC), LLMs (Lords communications and digital committee) and the big one — the AI white paper (DSIT), helping facilitate and writing up a roundtable with ODI members and summarising our final response.

There was sadly no Data Game at the Summit, but I must try to get the excellent 2022 episode online. (If you have no idea what I’m talking about, check out the 2021 and the Turing AI UK episodes.)

Institute for Government

August marked a decade for me at IfG.

We ran eleven Data Bites in 2023: our first ever in January (on social mobility and data gaps), February on defence, March on health, an April justice special followed by another in May, and unsponsored smorgasbords in May, June, July, September, October and November. Thanks to the Nuffield Foundation, we complemented the two justice specials with some further research and a short report on improving the use of data in the justice system — those of us working on it remarked we’d never had such a good hit rate of research interviews, every single one adding some fresh insight.

Content-wise, I thought it was a successful Data Bites year, with real variety and a high standard of presentation. Sponsor-wise, it was less so (we have to hit certain targets to keep the series going, and we cancelled the planned December event). That break gave IfG the chance to ask for a review of the series. Feedback from an attendee survey was overwhelmingly positive (thank you — and to the increasing number of people telling me in person how much they enjoy it!). We have a cracker in prospect for the end of February and another event in mid-March.

Then — from April, which will be event number 50 and mark five years — we’ll be moving from an event once a month to one every six weeks. You can also expect some more themed events tying in more closely with the IfG’s wider work programme — I hope that will take the power of data to a wider audience. But I also don’t want to lose the data/digital community we’ve built around the series, and (see gap in the thinktank market, above) I don’t think there is a sufficiently high level of data and digital adoption or knowledge in government that we can lose that focus altogether.

My other substantial bit of IfG work was our report on data sharing during the pandemic, which drew on individual case study roundtables on legislation; GPDPR; NHS Covid-19 Data Store and NHS National Data Platform; sharing between central, local and devolved government; counter fraud activities; and the Clinically Extremely Vulnerable People Service. We had a launch event back in February (everything to do with the project is collected here), and Computer Weekly covered some of the findings. I’ll repeat what I said in my 2022 yearnote: I’m pleased with some things we surfaced, I still think there’s a missing milestone project on data sharing in government, and the feedback on our launch event was instructive (some thought much of what was said was helpful and new, others that they’d heard it all before).

Other bits and pieces for IfG included:

  • Chairing public events on AI at both Conservative and Labour party conferences and helping organise a private roundtable on AI at each
  • Running an internal workshop on how to think about using AI (a shout-out to Demos for their excellent guidance), and helping run a skill share on using FOI
  • Recycling another blogpost on take out the trash day, and taking part in a Twitter Space on it (my first, I think)
  • Helping deliver some IfG Academy sessions to academics on policymaking, particularly with a data/AI bent, and presenting to an international delegation about data and digital government in the UK
  • Looking over or otherwise contributing to some work on Oflog, AI and the use of data at the centre of government
  • Keeping a safe distance from the multiple reshuffles (Data Bites intros aside), apart from contributing some thoughts on DSIT (and a surprisingly popular tweet on David Cameron).

Everything else

At the start of the year, my seven-projects-for-six-employers also included:

  • A strategy review for Defend Digital Me. A really enjoyable and hopefully helpful bit of work — all credit to Jen for doing such a brilliant job and — as a few people I spoke to said — inventing an ecosystem it’s difficult to imagine not existing
  • Finishing off TICTeC Labs for mySociety, concluding with an event reflecting on the experience. (It was a joy to be at the mySociety 20th birthday party — I’m proud to have played even a miniscule part in the story)
  • Some work for Public Digital on digital public infrastructure and building on their thinking about ‘data as a service’, which definitely helped me think through things
  • And I spoke to Dr. Iulia Cioroianu’s students at Bath about working in data.

Since then:

  • I’ve been doing a tiny bit for the University of Winchester, helping them think through their work on data and education
  • I contributed some thinking on data in government to Labour Together
  • I’ve co-chaired two more Think Data for Government conferences for Think Digital Partners (always a really enjoyable day — someone from the ONS wrote up the most recent one in December, as well as Think Digital publishing one, two, three, four, five session summaries). I also chaired part of the GovNet Tech conference for the first time
  • I’ve delivered some sessions on dataviz for Apolitical and Smart Thinking (any excuse, as ever, to plug my previous piece for them on how to approach dataviz as a thinktank), and also helped deliver a ‘public service 101’ test session for the former.
If only it had been a Bar: a genuinely unintended visit to Chart Street

Also:

And:

Writing

There was no further progress on The Book, though I can at least point to the Covid Inquiry as to why there was no time for that. But I’m delighted that the Royal Statistical Society’s Significance magazine is serialising my long read on the R Number (originally planned to be the sample chapter). A big thank you to Understanding Patient Data, including Natalie Banner who commissioned it when she was there. Preview of what’s to come on the Wellcome Trust blog.

Yet again I find myself typing that I wanted to write more but actually wrote less. I need to set a target and actually do something about that, rather than writing the same thing next year. I must make time for it — but also just have the confidence to do it, too.

My reading also hasn’t recovered from falling off a cliff during Covid and the lack of commuting, either.

The year ahead

Work with Connected by Data, the ODI and IfG continues (with a bit more on AI for the latter), with a few other projects here and there, too… there’s the usual personal stuff about eating healthier and getting fitter (I’ve got into the Parkrun habit, which is useful having booked the Hackney half marathon), and braving the carnival of algorithmic harm that is dating apps… I still need to sort out my social media habits (cheers Elon) and work out how to better read, engage and promote in a world where Twitter is withering away… and…

I was going to make a prediction that this would be the year the Post Office Horizon scandal finally broke through and made people question blind faith in technology, but ITV got there before me.

*As Jeni puts it, ‘It’s pretty shocking that our only hope for actual democratic scrutiny is from unelected Peers (though I am very thankful for them)’.

**I’ll avoid a lengthy detour on how the inaccessibility of documents from the Inquiry (though I appreciate there are good reasons) and parliament (oh god, amendments!) show our public institutions are still not set up for the digital/information age.

--

--

Gavin Freeguard
Gavin Freeguard

Written by Gavin Freeguard

Freelance, gavinfreeguard.com. All things (usually government) data. Dataviz etc newsletter at twitter.com/WarningGraphicC.

No responses yet