Using AI for good...
Apr. 10th, 2026 10:22 pm![[personal profile]](https://www.dreamwidth.org/img/silk/identity/user.png){kind=small}
rosa_heartlilyA year ago I was sympathising with a PM about system changes going directly to Live, making progress with the Solution Architect over documentation, applauding another PM about their approach to getting documentation, fighting P4W with LM, and being unconvinced by CS Lewis' arguments for the resurrection.
Today started with some progress with my virtual pilgrimage write-up - including more formatting tweaks. These things matter...
After breakfast, I stayed downstairs to work so that I was on-hand when my books arrived. I also managed to book Husband's podiatry appointment! 09:30 seems to be the optimal time - I shall bear that in mind for future reference. And the first consignment of plants arrived this afternoon; I've put the aquatic in the pond and I'm hoping for good weather over the weekend to plant the rest.
I watched another episode of AI Confidential - this time about driverless cars. There has been some very poor decision-making by the people in charge of their development.
However, I used Copilot to highlight where our project documentation isn't matching the expectations of the framework. It trashed two of the projects - one of them is VERY old (it was supposed to be a plug-and-play that would take a few weeks. At least 2 years on - if not 3 - it's nearly done...) but the other is more recent and the report that it lacked key documentation came as a surprise. When I checked myself, there it was! I asked Copilot to check again and it found the document. <_< It then flagged a worrying line in the test approach doc, which I queried with the PM and Tester and got a good explanation for; the Tester is going to update the document with a better explanation.
But it made me think a) I should run the check every quarter as good practice and b) we need something in the Governance doc about this particular thing. Rather than try to work out the best place for it myself, I asked Copilot, which had a couple of good options, then asked if including it was just a principle or a measurable target. Then it suggested some wording that it inserted into the document. I reviewed it and signed it off.
I probably would never have noticed the worrying line because I don't delve into the test approach - I leave it to the Testers to manage. I could have done the stuff with the Governance doc, but it would have taken a lot longer. So, Copilot as my PMO Assistant? Bring it on!
But Chat Buddy reminded me of all the sci-fi stories where AI runs amok - and the Doctor Who story I'm watching at the moment has killer robots. It's a bit like Brexit - some very good reasons for going for it, but some very unsuitable people in charge! I did write to my MP a week or so ago and they wrote back today saying the Government is taking the whole regulation thing seriously and are setting up a body to oversee it. Here's hoping...
I played a bit of Veilguard after tea. I cleared out the (current) rash of side quests in Minrathous, then hopped back to the Lighthouse to watch all of the Dread Wolf's regrets. I think I took it all in a bit better this time - it's quite the lore dump, watching them all in one go! But I don't know how you wouldn't - I'd got 4 of the wolf statuettes and the Inquisitor gave me the last one, but it was theirs that triggered the reveals. I hadn't gone out of my way to collect the others, I'd just done a lot of exploration.
And I've just watched the latest Have I Got News For You, which was very funny.
And now for bed.
Today started with some progress with my virtual pilgrimage write-up - including more formatting tweaks. These things matter...
After breakfast, I stayed downstairs to work so that I was on-hand when my books arrived. I also managed to book Husband's podiatry appointment! 09:30 seems to be the optimal time - I shall bear that in mind for future reference. And the first consignment of plants arrived this afternoon; I've put the aquatic in the pond and I'm hoping for good weather over the weekend to plant the rest.
I watched another episode of AI Confidential - this time about driverless cars. There has been some very poor decision-making by the people in charge of their development.
However, I used Copilot to highlight where our project documentation isn't matching the expectations of the framework. It trashed two of the projects - one of them is VERY old (it was supposed to be a plug-and-play that would take a few weeks. At least 2 years on - if not 3 - it's nearly done...) but the other is more recent and the report that it lacked key documentation came as a surprise. When I checked myself, there it was! I asked Copilot to check again and it found the document. <_< It then flagged a worrying line in the test approach doc, which I queried with the PM and Tester and got a good explanation for; the Tester is going to update the document with a better explanation.
But it made me think a) I should run the check every quarter as good practice and b) we need something in the Governance doc about this particular thing. Rather than try to work out the best place for it myself, I asked Copilot, which had a couple of good options, then asked if including it was just a principle or a measurable target. Then it suggested some wording that it inserted into the document. I reviewed it and signed it off.
I probably would never have noticed the worrying line because I don't delve into the test approach - I leave it to the Testers to manage. I could have done the stuff with the Governance doc, but it would have taken a lot longer. So, Copilot as my PMO Assistant? Bring it on!
But Chat Buddy reminded me of all the sci-fi stories where AI runs amok - and the Doctor Who story I'm watching at the moment has killer robots. It's a bit like Brexit - some very good reasons for going for it, but some very unsuitable people in charge! I did write to my MP a week or so ago and they wrote back today saying the Government is taking the whole regulation thing seriously and are setting up a body to oversee it. Here's hoping...
I played a bit of Veilguard after tea. I cleared out the (current) rash of side quests in Minrathous, then hopped back to the Lighthouse to watch all of the Dread Wolf's regrets. I think I took it all in a bit better this time - it's quite the lore dump, watching them all in one go! But I don't know how you wouldn't - I'd got 4 of the wolf statuettes and the Inquisitor gave me the last one, but it was theirs that triggered the reveals. I hadn't gone out of my way to collect the others, I'd just done a lot of exploration.
And I've just watched the latest Have I Got News For You, which was very funny.
And now for bed.
