Classic 'smell the fart' acting. Nailed it.

I was reading about dual-track agile yesterday off the back of a xgov chat.

I'd be interested to know how these approaches are discussed in service assessments. Your post helps shape the required narrative

I liked the way Ha yesterday distinguished between:

Build to learn
Build. Then learn.

❤️ Great timing Steve, I've been struggling to articulate this for the last couple of weeks. Thank you.

We've used similar template prompts in a tool we've been testing. We haven't spotted any incorrect responses or hallucinations/Bullshit in sessions yet. But it still hasn't stopped inconsistent respones

This is lovely. Also, 'GovCrush' is amazing. I think I have a few of these haha

Haha! Mad skillz

I think I speak on behalf of all User Researchers in saying listening to your own voice in a session is the WORST.

I didnt know her well but followed her insightful thoughts and ideas for years. The civil service and the general public have, and will continue to, benefit from her generousity. Gutted to lose such an awesome leader x

That's the strapline for 2024

This comes back to having Assurance gates near the end of each phase to prevent this stuff from progressing

So many other competing demands on consultancies so I can see why this happens.Looking to secure furthering funding just to stay alive in this market but also constantly looking to build their portfolio/org CV to secure other contracts.Result is pet projects they can point to, e.g. a shiny AI thing

Also, download thr citymapper app. It's loads better than Google maps at directing you around London

Nice one babs

I like these questions at the end, and would add 'level' to frequency, impact and cost. How wrong could the tool be? What happens if there's an absence of info for the tool to find? Lots of internal testing to do