What's annoying me today

Microsoft outage meaning my relaxing Friday morning coffee ride being rudely interrupted. Still there are worse places to have to help draft comms than sit by the bay with a coffee.

Will you be sittin' when the evening come?

Did said team member learn some new swear words shortly after review?

My stupid car, detected an engine fault and decided to breakdown in the entrance to a car park so I blocked everyone else form getting in/out!

TBF most were able to squeeze past with a little bump onto the pavement, but precious BMW drivers were getting all agitated at me.
I had to drive one little old ladies 20 yr old hatchback out, as she was too scared to do it herself.

AI is largely useful if you have the training and guard rails with which to use it.
Any kind of bid prep I would absolutely be using it.

Not everyone is thrilled about AI though.

www.theguardian.com/uk-news/2026/aug/14/scottish-police-expect-public-opposition-larbert-datacentre

It's very good for parsing large and disparate datasets e.g. my pricing tool. I have no use for it as a content creation tool.

I've just used it to create a tool that pulls letters together using set standard paragraphs. I had to put in the text as it wouldn't in case I didn't have the right to use it (I do) but it works well and saves over £1k per year as opposed to bought in software.

I use it for research into costs (eg hotels of a certain type) but I wouldn't use it to create anything that wasn't structured by me for work purposes and I certainly wouldn't send it's responses to clients at the moment

I've used it to produce board packs, with a prompt not fat off, iodate the Match slides to June. It scans all the files emails etc and does a damn good job. Bit of tweaking but honestly if that wasn't done noone would notice.

I've a couple of guys in India that do coding for me and I'm already at the point where if I added up all the hours I spend explaining the projects to them, update meetings, documentation, appraisal etc - I could just build the thing myself with the help of co-pilot. If one of them resigns for whatever reason I'll be suggesting to my boss that he gives me 20% of their salary as "AI budget" and doesn't re-hire.

I am of course aware that this very same food-chain crunch could happen to me at some point.

Co-pilot is excellent at generating complex Excel formula.

Also, I can input an email and ask co-pilot to generate a professional sounding email in response. We have an internal AI that is directed to use only the regulations and guidance we have directed it to, so it is quite reliable, but always needs a review and adjustment - the main issue is regulations are published every few years and the damn stupid bot AI can't figure out where one set of regulations starts and another ends, especially when politicians and the law have a tendency to backdate changes.

Then you arent using it right if you have any form of complex document to prepare

We have a pitch bit of AI. There is always Human in the Loop. Initial client contact is reforded, transcribed and summarised into a briefing document which is sent to the prospect for sign off. From there initial pitch documents are prepared based on the briefing, amends and library. Of we win this acts as the basis of tje client brain. All comms are recorded and added to the brain
We can pull metrics from the various tools we use for the brain and reporting.
Ive also got the basis of prospecting tool that will search various platforms for particular tech and identify various job titles and pull contact details.
Also use AI for the QA grunt work.
More to come.

Point of order: I'm not using it incorrectly, I'm not using it at all except for crunching large and diverse datasets to give me usable guidance upon which to make decisions.

My complaint was about people using it for things that it absolutely should not be used for, like writing a personal communication to a senior exec about a sensitive topic.

AI is of no use to me, except theoretically framing my emails in a nicer manner for the intended audience. Writing code is easier than working out what I want it to do.
One thing I dislike about it is the socialised expense that we are all paying for it through electricity prices, resource waste, and opportunity cost. Making it cheap enough to use gets us hooked on free drugs. Plus the IP theft. And all the other things.
And that senior management see that their own jobs could be replaced by it, which means so can everyone else's.

We're actually already at a point where the AI companies are so deeply leveraged and the shareholders impatient for returns that the cost of tokens is making businesses start to question whether free rein for employees to use it at will is a good idea. Saving employee costs to just replace it with AI consumable fees is a decision lots won't make.