bowsy.co.uk
I've been a long-time supporter of the government technology and transformation communities since 2009. I've worked for mySociety, the GLA, HackIT and GDS. I'm on the committees for GovCamp, TransformGov Talks and the UK Polyamory Association.
1,163 posts
589 followers
294 following
Regular Contributor
Active Commenter
comment in response to
post
Yikes, thanks. Will check!
comment in response to
post
There was govtech before GDS and some of it was good actually :-)
comment in response to
post
One piece covering the event. bsky.app/profile/bows...
comment in response to
post
Yes. Which is why trying to address it with policy is a mistake. You cannot apply a real solution to a fictional problem.
comment in response to
post
www.whatdotheyknow.com/help/request...
Has a "contact us" option if you get stuck.
comment in response to
post
Same thing for protest marches. No-one thinks they fix things but they do increase a cause's profile and help keep people involved motivated.
comment in response to
post
I think an argument can be made that the main benefit of the HoC and similar petition sites is not to drive direct change but to provide a focus point for awareness raising.
comment in response to
post
I wrote about government data standards and digital systems following the Supreme Court decision.
medium.com/desiderium-s...
comment in response to
post
@derekalton.bsky.social ☝️
comment in response to
post
Hypothesis. The actual migration numbers don't matter *at all*. Instead plot "dissatisfaction with immigration" vs. "how much immigration is headlined in the media". People aren't riled up by new neighbours - but by what the press tells them to be angry about.
comment in response to
post
I was at school with two siblings whose middle names were Thorin and Galadriel...
comment in response to
post
I had a chat with a government comms professional the other day. They genuinely seemed to struggle with the concept of a blog. Really, anything that wasn't going into national media via The Grid.
comment in response to
post
Sheesh - talk about having your own floors / flaws 🙄
comment in response to
post
No political party in the UK in my lifetime has said "these are the hard choices" they've always said "you can have everything". It baffles me that they could be on YouTube explaining detailed difficult choices every week but instead still focus on legacy media and twitter...
comment in response to
post
I'm currently reading @rory-stewart.bsky.social 's book which has it's own floors but talks a great deal of sense about this. In-person voting seems the most crazy and easily fixable thing.
comment in response to
post
I'm only being semi-facetious with that last part. Things like ChatGPT have a surface veneer of bothsideism but a core of fixed values (slavery is bad but trans rights is *shrug*). There should be laws forcing model makers to use standardised prompts to surface their values.
comment in response to
post
Once again, I strongly encourage people to read this short story : marshallbrain.com/manna1.
Meanwhile, the critical piece of work feels like creating public "moral value" tests that when the AIs are in charge at least they'll be "good people".
comment in response to
post
At 3am, obviously.
comment in response to
post
We also have a friend called Thurstan. Are there secretly a lot British Thurstan's out there keeping quiet?
comment in response to
post
No great words to say at such a time except that you continue to honor her memory every day by the way you interact with the world.
comment in response to
post
Of course, if they really wanted to do an analysis of how to save money I'd look into why each government organisation builds digital services in isolation on their own infrastructure rather than collaborating on a broad whole government platform.
comment in response to
post
I've very very much not an expert in benefit analysis but I'd love to see some professionals give feedback on all of this.
I really really want GDS to continue to be a success but this, especially the AI part, just smells bad. At the very least they could make the underlying data available.
comment in response to
post
"a Large Language Model (LLM) was used to assess the content of each GOV.UK service page and categorise
the transaction channels available."
"Where online channels were available, each service was assigned a random online uptake percentage."
comment in response to
post
"To estimate these savings comprehensively, the CDDO employed a Monte Carlo-style simulation approach. This approach involves generating thousands of scenarios (5,000
simulations) by randomly assigning costs, transaction volumes, and digital adoption rates" <truncated>
What?
comment in response to
post
"To enhance the credibility and reliability of the analysis, additional layers of external validation were implemented."
All of the validating organisations are part of government. Was there any external business or academic review of these models?
comment in response to
post
"Productivity gains are calculated using the assumption that 100% of routine tasks and 10% of non-routine tasks can be automated."
What?
comment in response to
post
Or do we just make a bunch of civil servants pause the good work they were doing to make up some ridiculously huge numbers that no-one believes and will never be refered back to again?
comment in response to
post
Finally found some time to read this (thanks @ben-shave.bsky.social). All I have to say is WOW this is some Tony Blair Institute level of absolute insanity. I am and always have been a very big supporter of GDS but does anyone think any of this is based in reality?
comment in response to
post
Also, I'm very disappointed, but not remotely surprised, that both the Government Digital and AI Roadmap survey and the Government Digital & Artificial Intelligence Roadmap webinar on May 20th are for civil servants only.
Why wouldn't you want feedback from motivated 3rd parties?