The modern office - where we read out what copilot tells us to think

[LorenzoCalzone]

I attended a meeting today with senior people in my company. People were clearly just reading out their AI drafted thought pieces. I thought "Is this the future of office work now? Listening to someone's copilot results, maybe countering it with another AI generated point of view". These perfectly scripted perspectives are all over linkedin too.

Of course I use it too. The toothpaste is out of the tube!

I wonder if mathematicians felt like this when the pocket calculator came along. Its cheating!

If you can use AI, you should use AI.

Why?
AI doesn't know my subject as well as I do.
It only knows what's already out there, so it can't teach me anything new.
It can't do the in-depth research that I can do.
It can't write in my style or voice.

Where we are.. a young generation who are porn addicted and think choking is a form of affection, reduced concentration span, high anxiety, obesity, 9 yr olds with skin care routines, obsession with consumerism and image, i could go on. It has been amazing in many ways but when you weigh up the bad especially SM it's questionable. Yet anyone who raises these concerns is a luddite, an old one out of touch, a nay sayer. People forget when a tech company is promoting new tech it's to increase profit, that is the one and only goal. It's beyond foolish to embrace something that has an impact on society without concerns or debate or regulation.

We’re no longer allowed to do online-only interviews for this very reason.

I am also anti-generative AI due to environmental and cognitive concerns

what a waste of water

You have to be careful with AI but I find it useful for some specific things. If I write a report. I can put it in copilot and ask it to tell me the weaknesses or gaps jn the report, to create a set of questions I might be asked or to summarise it.

It gives me a good gauge of where my thinking had gaps. For instance if the summary isnt what i though i was saying, i can go back and look my original and think why didnt my key point get picked out. Sometimes its AI being poor at reading and sometimes it me not being clear in my structure or language.

"It’s very useful to rephrase angry emails"

Yay... let's drive up everyone's energy bills because people are too lazy to write sensible emails.

I work in financial services. AI is being heavily pushed. We have been told if we dont embrace it, we will be left behind. There is redundancies ongoing at the moment.

I used to do a lot of technical work but people just seem to use copilot now instead.

I used it on Monday to ask if 'x' is a requirement under 'y' guideline and it said yes it is.
Then I asked it to tell me the piece of the guideline that states its a requirement.
It couldnt do this and had to backtracking and change answer.
Yet people are placing far too much trust in it.

id use all three copilot, grok, chatgpt, or copy and paste the analysis on each of them to expand it, eg im doing my own historical research essays eg egypt and i start with chat then use grok to expand the details , but i have to tell it to answer this but research the internet for answers etc

dh and I both use copilot to support our work. Dh has an exec who blindly reads out copilot responses in meetings with no critical thought or open opinion and gets paid £250k for it. Arrrgh!

Not all centres use closed loop systems, and those that do require even more electricity which in itself uses up a lot of resources, it absolutely is an issue, and and issue that a significant proportion of date centres try to hide by not tracking water usage.

im doing my own historical research essays eg egypt and i start with chat then use grok to expand the details , but i have to tell it to answer this but research the internet for answers etc

That's not historical research.

Here's another instance of AI gone rogue:

https://www.bbc.co.uk/news/articles/cz7dl7w8y7po

We are actively encouraged to use AI in my job

I use AI a lot for management bollocks that I use to do myself but found hard and boring. I’m cross with Copilot though because twice yesterday it told me to take the Victoria Line to Farringdon when I was planning a route to Heathrow. I mean surely it should be able to read a tube map.

I’m cross with Copilot though because twice yesterday it told me to take the Victoria Line to Farringdon when I was planning a route to Heathrow. I mean surely it should be able to read a tube map.

Why ask Copilot when you can look at the map yourself? Or use the tfl journey planner?

Are you certain Copilot knows when the Piccadilly line isn’t running, for instance?

Because I was modelling different journey times and lift availability at different stations as I will have heavy luggage. I know that the Piccadilly line isn’t running this weekend so the Elizabeth line will be more crowded but I certainly know for a fact without looking that the Victoria line doesn’t go to Farringdon.

i usually tell the ai to research the net for latest updates etc

did you tell it to research the latest updates via the net ? when asking it

The Victoria line has never gone anywhere near Farringdon so that wouldn’t have helped.

AI is complete shit and the sooner it crashes and burns, the better https://www.nbcnews.com/politics/trump-administration/us-government-map-africa-mislabels-every-country-global-conference-rcna590153

is a calculator demonstrably wrong half the time, while engaging in mass plagiarism and destroying the planet?

There[s been plenty of research to show that whatever paltry time savings person A makes using AI are lost down the line when persons B, C, and D have to clean up its mess.

And asking it to research the latest updates via the net definitely wouldn’t have helped.

Someone told me AI was great at planning trips. I input arrival times, airports, dates, hotels etc. It gave me an 8 day itinerary for a 6 day trip. And even when I pointed that out it maintained the 8 day itinerary. Which had been zig zagging around the place so much I’d have done very little in 8 days even if I had them! Utter horse shit. Not intelligent by any definition - it’s basically text prediction on a throne.

That makes two of us.