Notes 2021, episode 2
The shortest month of the year
February is the shortest month of the year, and it always flies past very quickly. But it’s also a bit of a dark month for me, locked in between the Christmas vibe and the promises of spring. We had a few days of freezing cold and snow in London, and it was beautiful to see Alexandra Palace all white; a bit less was to see people forgetting that there’s a pandemic and bunching up on the top of the hill to come down with their toboggans — lockdown fatigue is surely happening and we must really think before doing silly things. In more positive news, vaccines are being rolled out pretty fast and they seem to be working really well.
Work in February was busy and challenging, but satisfying. This contributed to my mental health getting much better in comparison with the lows of January. I have a small but dedicated team. My project officer is someone I’m increasingly trusting and being able to delegate to, while working with ACE has been very positive. Meanwhile, we have fixed the starting date for both our Senior Data Scientists vacancies and recruitment for our two Data/Technology Leads received a lot of interest, with a few candidates asking to have a chat (in doubt, please do ask for a chat, always!). Recruitment is as usual a long and complex process, but it’s progressing.
The Dragons’ Den-style competition which we launched last month also received a lot of interest, and I spent part of the month engaging with several NHS networks. Just yesterday I gave a quick talk to the Ageing Well network about the AI Skunkworks, AI in general, and the Den, and I was really happy with the questions I received.
Why are we doing this? Running a competition might look like a gimmick, but it’s a great tool to start engaging with people, triggering curiosity, and begin conversations about AI. Crowdsourcing problems — of which there is plenty of! — is a way to get a conversation about AI started and, as a consequence, help the healthcare system acquire knowledge and skills while developing proof-of-concepts. This is fundamentally the Skunkworks’ vision: increasing knowledge by doing; create proof-of-concepts quickly, so that others can take them and build full-fledged products; engage with people who work in the NHS and have pain points that they think might be solved with AI, and work together to test their understanding of AI, discuss what AI can do and what it can’t do.
I keep repeating as a mantra that I want to see some projects succeed, but also some projects fail, because that’s how we will collectively learn where AI really works and how to deploy it safely AI. As my team grows we’ll also launch other programmes (not everything will require Dragons!), so please stay tuned and get in touch with me if you have ideas.
I spent some time to provide my policy and strategy colleagues with feedback on the work they are doing. Something I’ve learned in previous jobs, and especially at the DfT, is that I enjoy working close to policy and strategy teams (the strategy is the delivery, right?): I believe that, especially in the context of AI, technologists play a big role in defining the discourse, but there’s also another benefit that comes from looking at how different teams and functions engage with the development of policy and strategy — it really helps to create, iteratively, a common language that strengthens the way we engage and deliver our work, acting as a catalyst and a motivator.
I recovered some of my reading mojo and I finished reading Jordan Mechner’s “The making of Prince of Persia”. Jordan is the software developer behind the saga, which he coded in the 80s when he was in his twenties. It’s a brilliant book, which not only gave me a burst of nostalgia of a time when having a home computer could fill your days (without any connection to the Internet), but also offered a lot of reflection on the highs and lows of a long-term project, the love and hate relationship with a job and a company, the meaning of success and failure, the need for recognition, and the importance of resilience and self-motivation. It’s a little book that I strongly recommend.
Also, I had a fanboi moment:
I moved my newsletter to Buttondown. I did this for two reasons. The first is purely financial: I was getting close to the paid-for threshold of 2000 subscribers that Mailchimp applies, making me all of a sudden have to pay $50/month, while Buttondown has a gentler slope: $5/month every 1000 subscribers; but most importantly, I love the ethos of Buttondown: it’s not VC-funded, it’s simple, it charges from day one and does so reasonably, and, interestingly, it’s a one-man company, which is pretty much my dream job (well, I mean, if I wasn’t already working in my dream job… :P). I also updated my subscribe page to the look and feel of my website. I hope you like it.
In home news, we started some garden work, which came with a shock… literally: a hidden 50,000V cable was in our garden, just 10 cm below ground level. The builder who was helping us cut through it, triggering a minor explosion and power cut to the block. Luckily no one was injured and the damage was fixed pretty quickly by the power company. The cable was unknown to us and the surveys that were done when we bought the house; I find it a little troubling that something like this could happen in the 21st century, but my googling shows it’s sadly not a rare event!
I’m helping the local community campaign for better broadband, as we only receive standard ADSL. We are three roads with about 3–4 Mbps at max, surrounded by ultrafast and superfast Internet (because reasons). Regardless of the specifics of the campaign which I won’t mention here because there are ongoing conversations, something pretty nice came off it: I’ve got to know about 40 neighbours, there’s a lot of support, and we now have a thriving Whatsapp group. In a diverse community like ours with all its challenges, it is very heart-warming to see neighbours working together towards a common goal.
Oh, I was about to forget! As we can’t run Open Data Camp, we’ve launched Open Data Cafe. It’s a shorter event, literally bring your own coffee and be ready to talk open data on April 10th between 12 and 1330 :) Tickets are still available here.
Have a great March, and speak in spring!
