It has been a few weeks since I’ve done a weeknotes in this style so here we go.
What I’ve been doing
Mainly I’ve been wrapping up projects and standing up a new one that kicked off on 1 April (though it is phase two of a previous project). I’ve provided our accountant with all the information for our end of year accounts and have been drafting the annual report. We had another good community of practice session yesterday and I’ve been writing up those notes.
What I need to take care of
I’m going to be trying to do a bit more in terms of write ups. There’s a culture in CONNECTED BY DATA of using writing as a reflective and learning process - so there’ll be some write ups that I’m not best placed to take on but I want to do more of those that I can. I’m working on an op ed now. I’m also going to put some bits into the strategic roadmap that Jeni is drafting for 2025-26 and do more of the annual report (for subject leads to then review and amend).
I’m also planning to solo work on a small funding bid - for some booster funds. It is an amount that doesn’t warrant lots of time from other team members and the likelihood of success is low too. But it is good for me to keep stretching that muscle and I can do so comfortably when an in depth knowledge of the subject (to create a convincing argument) is not needed.
Through some Trustee work I came across Fair Collective and want to note them here in case that is of use for others and so I can find them again at a time I need them. They are a social enterprise, fixed price model, collective of freelancers covering a huge range of skills. You contact them with a request and they send it out to everyone on their books that meets your interest area and those that want to respond.
What I’ve been inspired or challenged or moved by
You can take the girl out of the library, but you can’t take the library out of the girl. I love this video - with real people - sharing a poem about a library (this is my kind of library - people and connection).
What I’ve been reading
I’ve had some time for catching up on work related reading - some of which I’ve had bookmarked for ages.
Operations related
- The UK Government’s February Employer Bulletin - mainly getting up to speed on the imminent changes to National Insurance.
- A website for thinking about ‘stewarding loss’ in the context of winding up an organisation. I liked how they referenced grief, and specifically the work of Kathryn Mannix who I’ve come across lot in a previous role when supporting ‘Death Positive Libraries’ (her book ‘With the end in mind’ is excellent). That work has then spun out into The Decelerator which translates into practical support and tools for winding up an organisation with care and intent.
Data and AI related
With all the recent work I’m been supporting around public engagement, and particularly what I’m hearing discussed in the community of practice, I’ve been interested in reading more about occasions where the general public have been involved to make sense of data and help communicate about it more broadly. In the recent Public Voices in AI programme (we’ve been convening a People’s Advisory Panel for it) they’ve started work on a webtoon - an online cartoon - to convey complex issues in an accessible way. Sense about Science is an organisation that includes a public engagement team and they have done a piece of work with the general public in understanding the role of DNA in the criminal justice system. The team then worked with those members of public to create a guide that was informative and understandable, including cartoon type illustration. It seems to me this is a good example of how detailed information can be communicated without deep academic language.
I’d saved some links about work the ONS are doing - as the Data Science Campus - (largely in the open) about developing and testing tools using AI. One is to support the Living Costs and Food survey. As part of that they receive tens of thousands of receipts submitted by the public from which the data about items purchased and costs etc has to be extracted. They are testing a tool that uses multimodal generative AI, combined with automated text classification, to improve the efficiency of that process. Another tool is seeking to improve search engines particularly around statistics. By combining a search engine with an LLM they aim to interpret the semantic meaning of queries and documents. Early testing has found that this tool is working well with very specific questions such as “how many people watched the King’s Coronation?” as opposed to “are we in recession?”
Jeni recommended to the team a podcast about a new book ‘Abundance’ by Ezra Klein and Derek Thompson. The podcast talks about YIMBY-ism (an acronym I hadn’t come across before but is, as one may guess, a flip on “not in my back yard - NIMBY). There was some interesting and timely discussion in the episode discussing the current American context and the impact of tariffs. I’ve been most left with an understanding of how YIMBY want lots of house building, but tariffs on Canada and Mexico will significantly increase the cost of house building (due to the need to import essential raw materials) and so how one policy ideal is in friction with another (as I write this President Trump is pausing tariffs so this observation is very much of the moment and may not be so true when you read this).
I also very much enjoyed reading an article called ‘Data Relations’ by Salomé Viljoen. It explores datafication, the impact, risks and possible responses. It touched on some of the criminological theories I learnt about at University including the concept of the panopticon suggested by Jeremy Bentham and the analysis of that by Foucault. Most interestingly (for me) Viljoen demonstrates how datafication is not only about the collection of data and using it but how it modifies our behaviour. One such example was Netflix and how they use our watching data to determine the optimal way to keep us watching their content. They have worked out that autoplaying “next episode” (even when this is at the end of a series and bridging us into something similar that they think we’ll like) keeps us watching. Their analysis is even deep enough to work out the precise number of seconds to wait before starting the next episode - how even a few more seconds would mean we had time to actually decide what we wanted and perhaps stop watching. Viljoen also considers how we can address the exploitation of us through receiving payment or fighting back against group data analysis. This reminded me of fairly recent concerns about period tracking apps and anti-abortion lawmakers/states in the USA. The data collected can be used to identify when someone is pregnant - and when they’re no longer pregnant. Due to concerns about how this data may be obtained and misused some people were adding false data as a way of obfuscation and resistance. It was noted that this approach is only effective in undermining the data analysis though if the false data is close enough to be a real pattern so as to not be discounted as an anomaly (logging 15 days of consecutive bleeding for example). This weaponising of data people are logging as part of their lifestyle is not a hypothetical. An earlier example of a third party selling movement data to the US military from a range of apps including ‘Muslim Pro’ (an app designed to help Muslims track prayer times and identify the direction of Mecca - and mapping this geographically therefore creating a clear picture of where Muslims were and where they were praying) should also be of concern.
As well as an urgent need to raise awareness of how our anonymous data is used (anonymity does not equal safety or no impact) there is a need to have wider conversations about what kind of world we want to live in. Do we want a world where Netflix (and they aren’t unique in this of course) are changing our behaviour? Or where we can be targeted for tracking things in an app that previously we may have scribbled in personal paper diaries. These examples aren’t AI specific, but AI will increase the capacity of data analysis and cross referencing and exacerbate these challenges. This is where the work of CONNECTED BY DATA is again so important - we need to be bringing public voice in/amplifying their views on what we want and having a say in how companies are using our data.